Tropical Cyclones
[Index]
Main Index Home Page Stock Weather Photos Australian Severe Weather Forum Storm News and Storm Chasing Reports Tropical Cyclones / Hurricanes / Typhoons Weather Data and Links Wild Fires / Bushfires Weather Observation Techniques Weather Picture Catalogue Tornado Pictures and Reports Stock Video Footage and DVDs for sale
Monthly Global Tropical Cyclone Summary August 2004
[Summaries and Track Data] [Prepared by Gary Padgett]

                  MONTHLY GLOBAL TROPICAL CYCLONE SUMMARY

                               AUGUST, 2004

  (For general comments about the nature of these summaries, as well as
  information on how to download the tabular cyclone track files, see
  the Author's Note at the end of this summary.)

  *************************************************************************

         SPECIAL FEATURE - SOURCES OF TROPICAL CYCLONE INFORMATION

     The purpose of this section is to list some websites where many and
  varied types of tropical cyclone information are archived.  Many readers
  will know about these already, but for the benefit of those who don't,
  I wanted to include them.   After a couple of months, I will move this
  note to the ending section of the summary.

  (1) Aircraft Reconnaissance Information
  ---------------------------------------

     Various types of messages from reconnaissance aircraft may be
  retrieved from the following FTP site:

     ftp://ftp.nhc.noaa.gov/pub/products/nhc/recon/>

     Information regarding how to interpret the coded reconnaissance
  messages may be found at the following URL:

     http://www.nhc.noaa.gov/reconlist.shtml>

  Links are also included to websites with further information about the
  U. S. Air Force 53rd Weather Reconnaissance Squadron and the NOAA Air-
  craft Operations Center.

  (2) Archived Advisories
  -----------------------

     All the advisory products (public advisories, forecast/advisories,
  strike probabilities, discussions, various graphics) issued by TPC/NHC
  are archived on TPC's website.  For the current year (using 2004 as an
  example), the archived products can be found at:

     http://www.nhc.noaa.gov/archive/2004/index.shtml>

  Links to tropical products archives for earlier years are available at
  the following URL:

     http://www.nhc.noaa.gov/pastall.shtml>

     I am not aware at the moment of any other TCWC which archives all
  its tropical cyclone warning/advisory products for public access, but
  if I learn of any, I will add them to this list.

  (3) Satellite Imagery
  ---------------------

     Satellite images of tropical cyclones in various sensor bands are
  available on the NRL Monterrey and University of Wisconsin websites,
  courtesy of Jeff Hawkins and Chris Velden and their associates.  The
  links are:

     http://www.nrlmry.navy.mil/tc_pages/tc_home.html>

     http://cimss.ssec.wisc.edu/tropic/tropic.html>

  On the NRL site, the link to past years can be found in the upper left 
  corner of the screen.  For the CIMSS site, a link to data archives is 
  located in the lower left portion of the screen.

     I'm sure there are other sites with available imagery available, and
  as I learn of them, I will add the links to this list.

  *************************************************************************

                           AUGUST HIGHLIGHTS

  --> Destructive hurricane strikes southwestern Florida communities
  --> Large severe hurricane strikes Bahamas and southeastern Florida
  --> China struck by two typhoons--one very deadly
  --> Two intense typhoons pass through Marianas en route to Japan
  --> Typhoon grazes southeastern South Korea

  *************************************************************************

                 ***** Feature of the Month for August *****
                         
                 SOUTHERN HEMISPHERE TROPICAL CYCLONE NAMES
                            2004 - 2005 SEASON
                                   and
                  NORTH INDIAN OCEAN TROPICAL CYCLONE NAMES
                            

              TROPICAL CYCLONE NAMES for the AUSTRALIAN REGION
                          
     The Australian Bureau of Meteorology maintains three Tropical
  Cyclone Warning Centres (TCWC):  Perth, Western Australia; Darwin,
  Northern Territory; and Brisbane, Queensland.  Each centre is allotted
  a separate list of tropical cyclone names for tropical cyclones forming
  within its area of responsibility (AOR).  In addition a TCWC located at
  Port Moresby, Papua New Guinea (PNG)--a former Australian territory--
  maintains a list of native names to assign to the very rare tropical
  cyclones which form within its AOR.

     The AORs of the respective centres are:

  (1) Perth - 125E westward to 90E and south of 10S.  Currently, and for
      at least the next few years, the Perth TCWC will issue warnings for
      any systems north of 10S and south and west of the Indonesian islands.

  (2) Darwin - 125E eastward to 138E and extending northward to the
      equator.  There is a little irregularity with the eastern border
      in the Gulf of Carpentaria.  The Darwin TCWC issues High Seas
      Warnings for the entire Gulf of Carpentaria, but Brisbane issues
      Tropical Cyclone Advices and names cyclones in the eastern portion
      of the Gulf.  Also, currently, and for at least the next few years,
      the Darwin TCWC will issue warnings for any systems west of 125E
      and within the Indonesian archipelago in the Banda, Flores, and
      Java Seas.

  (3) Brisbane - 138E eastward to 160E and generally south of 10S.  The
      northern border with the Port Moresby AOR is somewhat irregular.

  (4) Port Moresby, PNG - immediate vicinity of the island of New Guinea
      and eastward to 160E generally north of 10S although the southern
      border is somewhat irregular.

     Names for the 2004-2005 season (** indicates name has already been
  assigned):

          Perth          Darwin        Brisbane        Port Moresby
  -----------------------------------------------------------------------

         Phoebe **      George          Harvey           Guba
         Raymond        Helen           Ingrid           Ila
         Sally          Ira             Jim              Kama
         Tim            Jasmine         Kate             Matere
         Vivienne       Kim             Larry            Rowe
         Willy          Laura           Monica           Tako
         Adeline        Matt            Nelson
         Bertie         Narelle         Odette
         Clare          Oswald          Pierre
         Daryl          Penny           Rebecca
         Emma                           Sandy
         Floyd                          Tania
         Glenda                         Vernon
         Hubert                         Wendy
         Isobel                         Alfred


           TROPICAL CYCLONE NAMES for the SOUTHWEST INDIAN OCEAN
                      and the SOUTH PACIFIC OCEAN

     The Tropical Cyclone Warning Centre (TCWC) at Nadi, Fiji, has
  tropical cyclone warning responsibility for the South Pacific east of
  160E and from the equator to 25S.   The Meteorological Service of New
  Zealand at Wellington has warning responsibility for waters south of
  25S, but almost all tropical cyclones in this basin form north of 25S.
  When a rare cyclone forms in the Wellington area of responsibility
  (AOR), it usually will be assigned a name from the Fiji list (such as
  was done for Tropical Cyclone Gita in February, 1999.)

     Tropical cyclone warning responsibility for South Indian waters west
  of 90E are shared by several TCWCs.       The Regional Specialty
  Meteorological Centre (RSMC) for the region is the office of Meteo
  France on the island of La Reunion.  However, following a long-standing
  practice, the sub-regional centres at Mauritius and Madagascar share
  the responsibility for actually naming tropical storms with Mauritius
  naming systems east of 55E and Madagascar covering the area west of
  55E.   RSMC La Reunion issues warnings for the basin independently of
  these sub-regional centres, but only advises regarding when or when not
  to assign a name to a developing cyclone.

     Names for the 2004-2005 season (** indicates name has already been
  assigned):

          Southwest Indian                       South Pacific
  -----------------------------------------------------------------------

     Arola **          Neddy                 Judy          Urmil
     Bento **          Ouledi                Kerry         Vaianu
     Chambo            Patricia              Lola          Wati
     Daren             Qiqita                Meena         Xavier
     Ernest            Ramon                 Nancy         Yani
     Felapi            Sopani                Olaf          Zita
     Gerard            Tina                  Percy         Arthur
     Hennie            Ula                   Rae           Becky
     Isang             Vera                  Sheila        Cliff
     Juliet            Willem                Tam           Daman
     Kalo              Xaoka
     Lilian            Yelda
     Madi              Zuze


            TROPICAL CYCLONE NAMES for the NORTH INDIAN OCEAN

     Several years ago a proposal was made at a meeting of the WMO/ESCAP
  Tropical Cyclone Committee to study the possibility of assigning names
  to the tropical cyclones of the Bay of Bengal and the Arabian Sea.  It
  took a few years to work out all the details and secure the consent of
  all the nations involved as well as to draft and approve a set of names,
  but by the summer of 2004 all the hurdles had been overcome and formal
  naming of tropical cyclones in this basin was set to begin on an
  experimental basis in September, 2004.   IMD didn't have long to wait
  to assign the first name--Severe Cyclonic Storm Onil formed very early
  in October in the Arabian Sea.

     The procedure for assigning names in the North Indian Ocean is
  similar to that in place for the Western North Pacific basin.  Eight
  countries--Bangladesh, India, Maldives, Myanmar, Oman, Pakistan,
  Sri Lanka, Thailand--contributed eight names each.  The names are
  arranged in eight columns with the nations listed in alphabetical
  order.  All the names in Column One will be used, followed by the
  names in Column Two, etc.

     The first eight names to be allotted in the North Indian Ocean basin
  are (** indicates name has already been assigned):

                           Onil **    (Bangladesh)
                           Agni **    (India)
                           Hibaru     (Maldives)
                           Pyarr      (Myanmar)
                           Baaz       (Oman)
                           Fanoos     (Pakistan)
                           Mala       (Sri Lanka)
                           Mukda      (Thailand)

  ***********************************************************************
  
                            ACTIVITY BY BASINS

  ATLANTIC (ATL) - North Atlantic Ocean, Caribbean Sea, Gulf of Mexico

  Activity for August:  4 tropical storms
                        1 hurricane
                        3 intense hurricanes


                        Sources of Information
                        ----------------------

     Most of the information presented below was obtained from the
  various tropical cyclone products issued by the Tropical Prediction
  Center/National Hurricane Center (TPC/NHC) in Miami, Florida:
  discussions, public advisories, forecast/advisories, tropical weather
  outlooks, special tropical disturbance statements, etc.    Some
  additional information may have been gleaned from the monthly
  summaries prepared by the hurricane specialists and available on
  TPC/NHC's website.     All references to sustained winds imply a
  1-minute averaging period unless otherwise noted.

     An interesting article discussing the atmospheric circulation
  patterns which led to the incredibly active and destructive Atlantic
  hurricane season, as well as information on the effects of individual
  cyclones, can be found on CSU's atmospheric sciences website at the
  following URL:

     http://tropical.atmos.colostate.edu/Forecasts/2004/oct2004/>


                   Atlantic Tropical Activity for August
                   -------------------------------------

     Following a stormless June and July, tropical cyclone activity
  exploded in August to produce the most active month of August on record.
  The average statistics for August (based on 1950-2003) are:  2.7 NS,
  1.5 H, and 0.6 IH with an NTC of 24%.  During August, 2004, a record
  eight tropical storms developed with four reaching hurricane intensity.
  Three of the hurricanes became intense hurricanes (Category 3 or higher
  on the Saffir/Simpson scale), and the NTC for August was a staggering
  84%.  And all of the cyclones except Hurricane Danielle had an impact on
  land.   Hurricane Alex passed a scant 10 miles off Cape Hatteras while
  a Category 2 hurricane, and later reached Category 3 status north of the
  38th parallel while moving northeastward over the warm Gulf Stream waters
  south of the Canadian Maritimes.   Tropical Storm Bonnie formed in the
  Gulf of Mexico and made landfall near St. Marks, FL, while Hurricane
  Charley was gathering steam in the Caribbean.  Charley struck western
  Cuba as a Category 3 hurricane, weakened slightly after crossing the
  island, then rapidly intensified into a strong Category 4 hurricane
  before making landfall near Punta Gorda and Port Charlotte, FL.  The
  storm crossed the Peninsula, maintaining hurricane intensity, and
  exited near Jacksonville.  Charley continued north-northeastward and
  made a second U. S. landfall in South Carolina.

     Hurricane Danielle formed in mid-month near the Cape Verdes but moved
  northward over the eastern Atlantic, becoming an impressive Category 2
  hurricane.  Tropical Storm Earl formed at the same time as Danielle and
  moved through the southern Windwards as a tropical storm, but lost its
  circulation and was downgraded to a tropical wave in the southeastern
  Caribbean.  Earl's remnants continued westward, reaching the Eastern
  Pacific several days later and ultimately redeveloping into Hurricane
  Frank.   Tropical Storms Gaston and Hermine were twins forming late in
  the month along an old frontal boundary stretching eastward into the
  Atlantic from the southeastern U. S. coast.  Gaston formed near the
  coast and struck South Carolina as a strong tropical storm near hurri-
  cane intensity.  Hermine formed from another LOW along the same frontal
  trough west of Bermuda and moved northward, eventually reaching south-
  eastern Massachusetts as a weak tropical storm.  (NOTE: Gaston has been
  upgraded to a Category 1 hurricane at landfall during a post-storm
  review of all the data.)

     Mighty Hurricane Frances formed on 25 August about midway between the
  Lesser Antilles and Africa.  Frances moved relentlessly on a west-
  northwesterly track which carried it north of the Leeward Islands and
  Puerto Rico as it became a large, severe Category 4 hurricane.  Frances
  slashed its way through the Bahamas, gradually weakening to Category 2
  levels during the process.  The storm made landfall in southeastern
  Florida as a very large, slow-moving Category 2 hurricane--not "major"
  by Saffir/Simpson standards, but still causing major damage as it crossed
  the Peninsula.  Frances emerged into the northeastern Gulf of Mexico and
  made a final landfall as a tropical storm near St. Marks, FL.

     A strong tropical wave moved off the west coast of Africa very late
  in August and passed through the Cape Verde Islands on the 31st.  The
  system was quite well-organized at one point, and SAB assigned a Dvorak
  classification of T2.5/2.5 at 31/1800 UTC.  However, the convection
  quickly began to diminish and the system was not classified as a
  tropical depression at the time.  The weak LOW persisted, however, and
  over a week later, on 9 September, was briefly upgraded to tropical
  depression status.  (This system will be mentioned in the September
  summary.)

     A special thanks to Kevin Boyle for writing the reports on Danielle,
  Earl, Gaston and Hermine.

  NOTE: The official TPC/NHC storm reports for Hurricane Gaston and
  Tropical Storm Hermine, as well as some earlier and later tropical
  cyclones, are now available online at the following link:

       http://www.nhc.noaa.gov/2004atlan.shtml>



                             HURRICANE ALEX
                                 (TC-01)
                           31 July - 7 August
                 --------------------------------------

  A. Storm Origins
  ----------------

     The first tropical cyclone of what has turned out to be one of the
  most active Atlantic seasons on record had its roots in an upper-level
  LOW which interacted with a surface trough--a fairly common mode of
  tropical cyclone formation in the subtropics.   On the afternoon of
  27 July an area of showers associated with a weak surface trough was
  located about 435 nm east of the northwestern Bahamas.  Over the next
  couple of days the elongated trough, interacting with an upper-level LOW,
  produced numerous showers and scattered thunderstorms over a wide area
  extending eastward from the Bahamas several hundred miles.  Convection
  had become more concentrated by the morning of the 29th, and upper-level
  winds were becoming a little more favorable for development.

     Visible satellite pictures on the morning of 30 July indicated that a
  LLCC was forming about 235 nm east-northeast of Freeport in the Bahamas.
  A U. S. Air Force Reserves reconnaissance aircraft flew into the area
  during the afternoon and found a broad circulation centered approximately
  240 nm east of Cape Canaveral, Florida.  Convection was still somewhat
  disorganized but the environment appeared favorable for further develop-
  ment.  However, overnight convection diminished and upper-level winds
  increased again, leading to a loss of organization.  At 0930 UTC on the
  31st the weak circulation was centered about 150 nm east of northern
  Florida and was not well-organized.   But during the morning hours the
  downhill trend reversed and the LOW appeared to be gaining in organiza-
  tion.  A reconnaissance mission during the morning did not find a well-
  defined surface circulation, but did report winds of 20-25 kts to the
  east of the LOW, which was located about 175 nm east of St. Augustine
  at 1530 UTC.    During the afternoon the LOW continued to exhibit
  increasing convective organization, and based on this plus a 30-kt
  intensity estimate from TAFB, the first advisory on Tropical Depression
  01 was issued at 2100 UTC.  The depression's center was located about
  150 nm south-southeast of Charleston, South Carolina, and drifting toward
  the northwest at 8 kts.


  B. Synoptic History
  -------------------

     A reconnaissance plane during the evening of the 31st could find no
  center, but a QuikScat pass just before the evening advisory confirmed
  that TD-01 consisted of a broad circulation with 20-25 kt winds.  Strong
  northeasterly flow was impinging on the depression, keeping the broad
  center on the northern edge of the deeper convection.  The MSW was upped
  to 30 kts at 1500 UTC on 1 August based on reconnaissance FLWs of 33 and
  36 kts and a visual surface estimate of 30 kts.   Another flight early
  in the afternoon found 41-kt winds at 300 m with a visual surface esti-
  mate of 35 kts.  On this basis, TD-01 was upgraded to Tropical Storm Alex
  at 1800 UTC, located about 70 nm south-southeast of Charleston, SC, and
  essentially stationary.  There were signs that the circulation was
  gradually getting better organized, although the aforementioned shear was
  preventing the well-defined mid-level circulation noted in radar imagery
  from aligning with any of the many low-level swirls that were continuing
  to pop out on the north side of the convection.

     Early on the morning of 2 August radar imagery indicated that the
  core of Alex was becoming much better organized, and a reconnaissance
  flight into the storm found that the CP had fallen to 992 mb and measured
  peak 850-mb winds of 62 kts.  A special advisory was issued at 1200 UTC,
  upping the winds to 50 kts.  The center was relocated to a position about
  105 nm south-southeast of Charleston, SC, and drifting eastward.  As the
  upper-level shear pattern changed, so Alex's structure evolved with the
  strong central convection shifting east of the center with a well-
  developed band in the southeast quadrant.  The drop in pressure leveled
  off during the afternoon for a time, concurrent with the formation of a
  20-mile wide eye that was open to the southeast.  Later in the evening
  another flight found that Alex had a 10-nm diameter eye with a surface
  pressure of 987 mb and FLWs of 69 kts.  Alex's MSW was upped to 60 kts in
  the 03/0300 UTC advisory, based primarily on the central pressure.  The
  storm at this time was moving northeastward at 8 kts, but an approaching
  mid-level trough over the Missouri River valley was forecast to dig a
  little more southward before pushing eastward, an event which would
  likely lead to a more south-southwesterly steering flow that would help
  to pull Alex northward toward the North Carolina coast.

     In an intermediate advisory at 03/0600 UTC, Alex was upgraded to the
  first hurricane of the Atlantic season.  Alex was then centered about
  65 nm south-southeast of Wilmington, NC, moving northeastward at 8 kts.
  A reconnaissance mission had found 850-mb winds of 81 kts along with a
  CP of 983 mb.  Afterward, the Wilmington and Newport WSR-88D radars had
  shown 85-kt winds at around 2500 metres.  The radar imagery revealed an
  eyewall in constant flux--occasionally fully-closed and occasionally
  open to the southwest, likely due to upper-level southwesterly flow
  impinging on the cyclone.  Alex's intensity took another jump, rather
  unexpected, at 1500 UTC up to Category 2 status.  The pressure had fallen
  to 972 mb, and a reconnaissance aircraft found peak FLWs of 105 kts at
  1135 UTC.  Around 1700 UTC the center of Hurricane Alex passed only
  10 miles southeast of Cape Hatteras, bringing sustained Category 1
  hurricane winds to the Outer Banks.  The maximum winds at the time were
  85 kts, but the strongest winds were in the eastern semicircle--away from
  the coast.  The storm by this time was embedded in the deep-layer south-
  westerly flow to the north of the subtropical ridge and was moving north-
  eastward at 15 kts.   By 0300 UTC on the 4th Alex was already 150 nm
  east-northeast of Cape Hatteras.  An evening reconnaissance flight had
  found a peak FLW of 87 kts, so the MSW was decreased to 80 kts.

     By 1500 UTC on 4 August Alex had reached a position about 350 nm
  east-northeast of Cape Hatteras and was accelerating out to sea at about
  17 kts.  Dvorak satellite intensity estimates had come down a bit, but
  Alex was still a healthy Category 1 hurricane with the MSW decreased to
  only 75 kts.  Six hours later it was obvious that Alex had
  intensified once more.  Based on a consensus of Dvorak estimates, the
  MSW was raised to 90 kts, making Alex a Category 2 hurricane once more.
  The storm was just south of the north wall of the Gulf Stream over SSTs
  between 26C and 27C.   But an even bigger surprise was in store.  Alex
  continued to intensify during the evening hours of 4 August with CI
  numbers from TAFB and SAB reaching 102 kts and a 3-hour AODT average
  of 105 kts.  Other Data-T and ODT numbers also supported increasing the
  MSW to 105 kts, making Alex the first intense hurricane (Category 3 or
  higher on the Saffir/Simpson scale) of the season.   Alex became the
  strongest intense hurricane on record to develop north of the 38th
  parallel, with Hurricane Ellen of 1973 coming in second at 100 kts.  Alex
  and Ellen are the only two known hurricanes to reach 100 kts at such a
  high latitude.  In the case of Alex, SSTs were running more than 2 Deg C
  above average in the area.   At the time of the 05/0300 UTC advisory Alex
  was located south of Nova Scotia, or about 700 nm southwest of Cape Race,
  Newfoundland, and moving east-northeastward at 22 kts.

     Alex held on to its peak intensity for a period of 18 hours.  Cloud
  tops continued to cool during the morning of 5 August, reaching temp-
  eratures of -65 to -75 C.   By 1500 UTC cloud tops were beginning to
  warm, but Dvorak classifications were unchanged at T5.5.  Alex was then
  located about 350 nm southwest of Cape Race, racing east-northeastward
  at 30 kts.  By afternoon the eye was becoming less distinct and cloud
  top temperatures were continuing to warm as the hurricane began to move
  over colder waters.   The MSW was reduced to 90 kts at 2100 UTC, and
  further to 75 kts at 06/0300 UTC.  Data from a 05/2210 UTC SSM/I overpass
  suggested that the LLCC was beginning to decouple from the mid and upper-
  level circulations.  Around 2200 UTC Alex's eye went almost directly over
  Canadian buoy 44140, which reported a minimum SLP of 979.3 mb.

     Deep convection decreased significantly during the early hours of the
  6th of August, and by 0900 UTC was confined to a few clusters in the
  eastern semicircle.   The intensity was set to 65 kts based on a average
  of satellite classifications and the rapid forward motion (45 kts).  The
  final NHC advisory was issued at 06/1500 UTC and placed the center of
  Alex about 750 nm east of Cape Race.  The MSW was reduced to 50 kts and
  the cyclone was rapidly becoming extratropical as it raced over the
  cold North Atlantic waters.  By 0000 UTC on 7 August Alex's remnants
  consisted of a 35-kt gale center located approximately 500 nm northeast
  of the northwesternmost Azores.   I received the following e-mail from
  Kevin Boyle regarding the "afterlife" of Alex:

     "Alex became a deep depression (975 mb) and has parked itself to the
  west of the British Isles where it will remain throughout the coming
  week.  At present it is advecting a lot of warm, humid air and temp-
  eratures failed to fall below 19 C (DP of 19 C) last night (night of
  8 August).  Bracknell surface charts indicate that the filling Alex will
  cross Britain as a weakening feature towards the weekend.  Torrential
  downpours are expected throughout the week and there is a concern for
  flooding in places, especially after the heavy rain and thunderstorms
  at the end of last week."


  C. Meteorological Observations
  ------------------------------

     A NOAA buoy located near Frying Pan Shoals reported a wind gust of
  43 kts shortly after midnight on 3 August.  Later the buoy reported
  sustained winds of 33 kts, gusting to 41 kts.  Around midday on the
  3rd there was an unofficial report of sustained winds to 43 kts with
  a peak gust of 73 kts from Okracoke Island.  Another unofficial report
  was received from Hatteras Village of sustained winds to 57 kts, gusting
  to 75 kts.

     Unofficial sustained winds of 63 kts and a peak gust of 89 kts were
  reported from near Cape Hatteras by two private firms, Weatherflow, Inc.
  and the Hurricane Intercept Research Team, respectively.  Additionally,
  Weatherflow, Inc. measured a 5-min avg wind of 65 kts at Avon Pier around
  1735 UTC.


  D. Casualties and Damage
  ------------------------

     There were no known casualties associated with Hurricane Alex.
  Estimates place the damage in the Outer Banks at around $2.5 million.

  (Report written by Gary Padgett)



                         TROPICAL STORM BONNIE
                                 (TC-02)
                             3 - 14 August
               -----------------------------------------

  A. Storm Origins
  ----------------

     Shower activity associated with a tropical wave which had moved off
  the western African coast in late July began to show signs of increased
  organization on the morning of 31 July several hundred miles southwest
  of the Cape Verde Islands.  Late in the afternoon the main action was
  located about 520 nm west-southwest of the islands and propagating
  westward at 17 kts.   The STWO from TPC/NHC indicated that some slow
  development was possible.   Subsequent days saw the wave continue to
  move rather rapidly across the tropical Atlantic, gradually increasing
  in organization.  By early morning on 3 August the wave had reached a
  point about 460 nm east of the Windward Islands.  Organization had
  improved to the point that advisories were initiated on Tropical
  Depression 02 at 03/1500 UTC.  The depression's center was estimated to
  be about 400 nm east of the Lesser Antilles and was moving westward
  at 18 kts.  The initial MSW was estimated at 25 kts.  There was some
  question as to whether the rapidly moving system had a closed surface
  circulation, but the case for this was made based on an analysis of
  QuikScat data.


  B. Synoptic History
  -------------------

     The depression continued rapidly westward toward the Windward Islands.
  The center was located about 40 nm north-northwest of Barbados at 0600
  UTC on 4 August and just east of St. Lucia three hours later.  Whether
  or not there was actually a closed LLCC continued to be a matter of
  uncertainty.  The MSW, initially set at 25 kts, was upped to 30 kts at
  04/0300 UTC based on CI estimates of 35 and 30 kts from TAFB and SAB,
  respectively, plus some near 30-kt uncontaminated wind vectors from a
  03/2208 UTC QuikScat overpass.  By early morning of the 4th convection
  had become more concentrated, but surface pressures suggested that there
  was not a closed circulation.  A reconnaissance plane reached the area
  during the morning and found a very sharp wind shift and some south-
  southwesterly winds at 760 m, but was not able to close off a circu-
  lation.  Since most dynamical model guidance was indicating that the
  system would soon slow down, it was carried as a tropical depression in
  the 04/1500 UTC advisory.   However, by afternoon the convection had
  become elongated in an east-west direction and less concentrated, so
  TD-02 was downgraded to a tropical wave at 2100 UTC about 330 nm south-
  southeast of San Juan, Puerto Rico.

     No sooner had TD-02 been downgraded than it looked like it might be
  about to stage a comeback.   During the evening hours of 4 August
  convection became more concentrated about 150 nm east of Bonaire, and
  the wave was moving slower at about 15 kts, the slower movement making
  reformation of a surface LOW possible.  However, no new circulation
  formed and by the next day upper-level winds had become less favorable
  for development.  Over the next few days the remnant tropical wave
  continued rapidly westward across the Caribbean, generating showers
  and thunderstorms over a wide area in the Greater Antilles and the
  central and western Caribbean Sea.   On the 7th the system appeared
  much weaker and wasn't even mentioned in the 07/2130 UTC STWO.  However,
  by the afternoon of the 8th disturbed weather had become better organized
  over the extreme northwestern Caribbean and some potential for develop-
  ment was noted in the afternoon STWO.

     During the evening a tight vortex spun up over the Yucatan Channel,
  as seen in radar imagery.  Likely this feature was at mid-levels, but was
  a harbinger of things to come.  By early morning of 9 August the system
  had entered the Gulf of Mexico and was located about 65 nm north of
  Cancun.   A reconnaissance plane visited the area around midday and found
  a very small surface circulation with a 450-m FLW of 56 kts.  Advisories
  were re-initiated on the system, which was christened Tropical Storm
  Bonnie, at 2100 UTC with the center located about 355 nm south of the
  mouth of the Mississippi River.  The initial warning intensity was set
  at 40 kts, and the lowest CP reported by the aircraft was 1007 mb.  Winds
  were increased to 45 kts at 10/0300 UTC after an evening reconnaissance
  mission found a 7-nm diameter closed eyewall.  However, the eyewall soon
  deteriorated somewhat and expanded to 17 nm.  It is unusual to see a
  closed eyewall in a storm of less than hurricane intensity.

     Bonnie's initial northwesterly motion became northerly on the 10th as
  a shortwave trough moving into the northwestern Gulf of Mexico weakened
  the ridge to the north.   The intensity reached a temporary peak of 50
  kts at 1500 UTC after a dropsonde reported 70-kt winds a couple hundred
  feet above the surface with surface winds of 51 kts.  However, central
  convection diminished somewhat during the afternoon and by 11/0300 UTC
  the MSW had been lowered to 40 kts.  Bonnie's intensity fluctuated up
  and down several times during its 3-day trek across the Gulf of Mexico--
  something rather common in very small tropical cyclones.   At 2100 UTC
  on 10 August Bonnie was centered about 275 nm south of the Mississippi
  River's mouth, moving slowly north at 5 kts.  Gales extended outward
  from the center 40 nm in the southern quadrants and only 25 nm to the
  north.

     Bonnie continued to experience ups and downs on 11 August.  During
  the morning the cyclone became much better organized with a burst of
  convection with cloud tops as cold as -83 C forming over the center.
  Buoy 42001, located about 45 nm northeast of Bonnie's center, reported
  a 10-min avg wind of 41 kts with a gust to 52 kts.    Gradient wind
  computations using a 10-mb pressure difference between the center and
  the buoy indicated that 50-kt winds were possible.  Based on this, the
  1500 UTC advisory MSW was conservatively increased to 45 kts.  A later
  report from the buoy reported a gust to 66 kts, so the MSW was increased
  to 55 kts in the 1800 UTC intermediate advisory.   In response to an
  unseasonably strong deep-layer trough over the central U. S., by 1800
  UTC Bonnie had made the expected sharp turn to the northeast.  The storm
  was then located about 180 nm south of the mouth of the Mississippi and
  moving northeast at about 10 kts.   After the tremendous burst of deep
  convection observed during the morning, Bonnie's convection leveled off.
  During the afternoon the tight inner wind core which had been present for
  the past three days collapsed, leaving the maximum winds in a band of
  convection removed from the center.  An evening reconnaissance flight
  found peak FLWs of only 30 kts, even though CI estimates were 45 and
  55 kts.   Since it was possible that the aircraft had not sampled the
  strongest winds, the MSW was reduced to only 50 kts at 12/0300 UTC,
  although the forecaster commented that this might be generous.

     Bonnie accelerated toward the northeast as the morning of 12 August
  progressed.  Winds were decreased to 45 kts at 0900 UTC, but were upped
  slightly to 50 kts in the 1200 UTC intermediate advisory.  Bonnie was
  then centered only about 70 nm southwest of Apalachicola, FL, and was
  moving northeastward at about 20 kts.   By 1500 UTC Bonnie was weakening
  once more and the cloud pattern was becoming elongated, suggesting that
  extratropical transition was underway.   The center of Bonnie hugged
  the coast from near Apalachicola east-northeastward to near the head of
  Apalachee Bay where the center finally made landfall.  At 1800 UTC the
  center of the weakening cyclone was located inland about 50 km south-
  east of Tallahassee, FL, moving toward the east-northeast at 24 kts.
  Maximum winds had decreased to 35 kts, and at 2100 UTC Bonnie was down-
  graded to a depression.     The cloud pattern was by then becoming
  associated with a frontal system.  This was the final advisory package
  issued by NHC as the responsibility for issuing advisories was handed
  over to HPC.  Bonnie's remnants sped northeastward across the coastal
  plain of Georgia and the Carolinas and by the afternoon of the 13th
  had moved out to sea off the coast of Virginia.  By 14/0000 UTC the
  LOW was off the New England coast, speeding north-northeastward, and
  apparently was soon absorbed by an extratropical LOW to the north.


  C. Meteorological Observations
  ------------------------------

     While Bonnie was over the Gulf of Mexico, ship H3GQ reported 40-kt
  southwest winds 40 nm south of the center at 1800 UTC on 10 August.  On
  the morning of the 12th, the center of Bonnie passed very near NOAA buoy
  42039, which reported gusts to 45 kts with a minimum SLP of 1002.7 mb.

     The following rainfall observations were gleaned from the HPC storm
  summaries.   Perry, FL, had a storm total of 79 mm, while Athens and
  Augusta, GA, both netted 53 mm in association with Bonnie.   During the
  six hours ending at 0600 UTC on 13 August, Hunter, GA, and North Myrtle
  Beach, SC, recorded 72 mm and 75 mm, respectively.

     The only location recording a 24-hour total exceeding 100 mm was
  North Myrtle Beach, SC, which measured 108 mm in the 24 hours ending
  at 1200 UTC on the 13th.   A little further north, Newport, NC, recorded
  a storm total of 65 mm, and Cherry Point, NC, measured 75 mm in the
  30-hour period ending at 13/1800 UTC.

  NOTE:  More detailed rainfall information can be found in the HPC
  advisories on Bonnie, which are archived on HPC's website:

     http://www.hpc.ncep.noaa.gov/>


  C. Damage and Casualties
  ------------------------

     Damage from Tropical Storm Bonnie was minimal.   The storm was
  responsible for three deaths in North Carolina due to tornadoes spawned
  by the weakening cyclone.

  (Report written by Gary Padgett)



                           HURRICANE CHARLEY
                                (TC-03)
                             9 - 16 August
                 -------------------------------------

  A. Storm Origins
  ----------------

     A tropical wave moved off the western coast of Africa on 4 August.
  The system moved rapidly westward with little sign of development until
  the 8th, when it was located some 750 nm east of the Windward Islands.
  Even then, associated thunderstorm activity was not well-organized.
  On the early morning of the 9th the wave was located about 130 nm east of
  the Windwards, moving west-northwestward at 22 kts with no signs of
  tropical cyclone formation.  However, later on during the morning showers
  and thunderstorms became much better organized and surface pressures had
  fallen significantly over the islands.  A Special Tropical Disturbance
  Statement was issued at 1345 UTC which noted that Barbados had reported
  wind gusts to 46 kts during the previous couple of hours.  Subsequent
  surface observations from Trinidad and Margarita showed west-southwest
  and west-northwest winds, respectively, of about 10 kts, establishing
  the fact that a surface circulation existed.  Hence, advisories were
  initiated at 1745 UTC on Tropical Depression 03, located about 45 nm
  southeast of Grenada and moving westward at 19 kts.  Visible satellite
  imagery depicted a very well-organized system with distinct banding
  features.  TD-03 was forecast to intensify to hurricane intensity in
  72 hours.


  B. Synoptic History
  -------------------

     TD-03 faired rather well as it traversed the typically unfavorable
  southeastern Caribbean Sea.  During the evening banding appeared less
  impressive than earlier, but new convection formed near the center of
  circulation.   At 0900 UTC on 10 August the depression was upgraded to
  Tropical Storm Charley, located about 390 nm south-southeast of Santo
  Domingo and moving west-northwestward at 21 kts.  The cyclone had become
  better organized overnight, and even though outer bands of deep
  convection were currently minimal, the circulation occupied a large
  envelope and outflow was excellent in all directions.  Dvorak classi-
  fications from all agencies were T2.5.  As the day progressed Charley
  slowly became better organized, and the first U. S. Air Force Reserves
  reconnaissance flight into the storm, around 2000 UTC, found a tight
  center with a CP of 999 mb and peak FLWs of 72 kts just to the northeast
  of the center.   Based on the aircraft data and a CI estimate of 55 kts
  from TAFB, Charley's MSW was increased to 55 kts at 11/0300 UTC.  The
  storm was still moving rapidly west-northwestward from a position about
  260 nm east-southeast of Kingston, Jamaica.

     The reconnaissance mission into the storm around 11/0600 UTC found
  that the CP had dropped to 995 mb with a closed eyewall present.  A
  flight around midday found peak FLWs of 80 kts at 850 mb; hence, Charley
  was upgraded to the season's second hurricane at 1800 UTC when located
  approximately 80 nm south of Kingston, Jamaica.  The young hurricane
  was moving west-northwestward at the slightly slower pace of 16 kts.
  During the evening Charley was sporting a small, closed eyewall of only
  8 nm as it began to turn slightly more toward the northwest.  The storm
  continued to strengthen during the night--winds were upped to 75 kts at
  12/0600 UTC.   An upper-level LOW to the west had been inhibiting outflow
  in that sector, but this feature began retreating westward away from the
  cyclone, leading to a lessening of shear and improved outflow.  By 1500
  UTC Charley was moving northwestward at 15 kts, and the track gradually
  became more north-northwesterly as the day progressed.  In an inter-
  mediate advisory at 1800 UTC, Charley's MSW was upped to 90 kts, making
  it a Category 2 hurricane on the Saffir/Simpson scale.  The hurricane was
  then located about 165 nm south-southeast of Havana, Cuba.  This upgrade
  was based on a peak FLW of 105 kts and a 92-kt dropsonde surface wind.
  The CP was only down to 980 mb, but the storm was moving into a region
  of higher-than-average surface pressures, implying that the usual
  pressure/wind relationship might not be valid.

     At 13/0300 UTC Hurricane Charley was moving north-northwestward toward
  the southern coast of Cuba, being located between the Isle of Youth and
  the Cuban mainland about 45 nm south of Havana.  By 0600 UTC the hurri-
  cane had almost crossed the island and was situated only 22 km west of
  downtown Havana.   Even though operationally Charley was a Category 2
  hurricane when it reached Cuba, meteorological observations from the
  island indicate that the hurricane had reached intense hurricane status,
  or Category 3 on the Saffir/Simpson scale.  The maximum 1-min avg wind
  recorded in Cuba was 103 kts with a minimum SLP of 966 mb.  A storm
  surge of 3.7 m was measured in Playa Cajio along the southern coast of
  Havana Province.  (More of the Cuban observations follow in Section C.)

     Reconnaissance observations revealed that Charley did not weaken very
  much at all after crossing Cuba.  Shortly after the eye had cleared the
  northern coast, a plane found a CP of 970 mb with peak 700-mb winds of
  104 kts.  Also, Key West radar showed Doppler winds higher than 100 kts.
  The MSW was upped to 95 kts at 0900 UTC.  Things began to happen very
  rapidly and dramatically during the morning of Friday, 13 August.  An
  eyewall dropsonde around 13/1200 UTC indicated that the MSW was still
  near 95 kts, but the CP had dropped 5 mb to 965 mb between 1200 and 1400
  UTC.  By 1500 UTC Charley's motion had become northerly at 16 kts, and
  two hours later the storm was moving north-northeastward at 18 kts.  The
  1700 UTC intermediate advisory upgraded Charley to a Category 3 hurricane
  with 110-kt winds, located about 60 nm south-southwest of Ft. Myers, FL.
  At about the same time, a reconnaissance aircraft found that the pressure
  had fallen further to 954 mb and measured a peak FLW of 141 kts.  On
  this basis, a special advisory was issued at 1800 UTC upgrading Charley
  to a Category 4 hurricane with a MSW of 125 kts.  

     Charley's pressure continued to fall, and the final reconnaissance
  fix at 1956 UTC, just before the eye reached the coast, found a CP of
  941 mb and a peak FLW of 148 kts.  The eye of Charley moved inland near
  Cayo Costa around 2000 UTC.  The operational MSW at the time of landfall
  was 125 kts, and this value has been widely reported in the media as the
  landfall intensity.  However, the 13/2100 UTC NHC discussion bulletin
  noted that the FLW of 148 kts corresponds to a surface MSW of 130 kts,
  and I have learned from a reliable source that in all probability 
  Charley's official landfall intensity will be 130 kts.  An unofficial
  gust of 110 kts was recorded near Punta Gorda along with a 943.6 mb
  surface pressure.  Charley began weakening as it moved quickly north-
  northeastward across the Florida Peninsula.  At 2300 UTC the center was
  about 100 km south-southwest of Orlando with the MSW estimated at 100
  kts, and by 14/0300 UTC had passed very near that city and was nearing
  the Atlantic coast near Daytona Beach.  The estimated MSW had decreased
  to 75 kts by this time, based on surface and WSR-88D Doppler wind data.

     By 0600 UTC on 14 August the center of Hurricane Charley had emerged
  into the Atlantic just off the Florida coast and was located about 165 nm
  south-southwest of Charleston, South Carolina.  A reconnaissance plane
  found peak FLWs of 88 kts in the eastern quadrant with a CP of 994 mb,
  so the MSW remained at 75 kts in the 14/0900 UTC advisory.  By 1200 UTC
  the storm was located about 30 nm south-southeast of Charleston and was
  racing to the north-northeast at 25 kts.  At 1500 UTC radar and surface
  observations indicated that the center of Charley was on the coast near
  Myrtle Beach, South Carolina, with the CP down to 989 mb.  The MSW by
  this time had dropped to 65 kts, and the former Category 4 hurricane
  was downgraded to a tropical storm three hours later when located over
  eastern North Carolina about 55 km north of Wilmington.  Wrightsville
  Beach reported a sustained wind of 61 kts at 1736 UTC, so the MSW was
  held at 60 kts for the 2100 UTC advisory.

     At 15/0000 UTC the center of Tropical Storm Charley was located just
  east of Virginia Beach, Virginia, and racing northeastward at 29 kts.
  Winds had decreased to near 45 kts and the storm was beginning to lose
  tropical characteristics.  By 15/0600 UTC Charley was passing about
  35 nm east of Atlantic City, New Jersey, with peak winds of only minimal
  tropical storm intensity.   Charley made a final landfall around 0900 UTC
  on Long Island near the town of Farmingville, and by 1200 UTC was located
  in the vicinity of Boston.   The final TPC/NHC advisory on Charley was
  issued at 15/1500 UTC with the system located east of Cape Cod and still
  moving quickly northeastward.  The extratropical remnant was forecast to
  lose its identity in a frontal zone, and apparently this happened, as
  OPC made its final reference to the post-Charley system at 16/0000 UTC.

     Hurricane Charley's recurvature at a fairly low latitude in the Gulf
  of Mexico was more typical of a late September or October hurricane.
  The rather unusual track was due to an unseasonably strong shortwave
  trough moving into the southeastern U. S.  This same trough had recurved
  Tropical Storm Bonnie into the eastern Florida Panhandle the day before
  Charley savagely struck the Charlotte Harbor communities.  Charley was
  the first of four destructive hurricanes to strike the state of Florida
  during the historic 2004 Atlantic hurricane season, and the first of
  three to bring hurricane-force winds to many of the inland counties
  located on the Peninsula.  Hurricane Charley was also the first major
  hurricane to strike the Charlotte Harbor since at least 1944.  Following
  is a list of the stronger hurricanes to strike this area of Florida
  since 1871, and none appears to have been anywhere near as intense as
  Charley:

  (1) Oct, 1873 - A major hurricane made landfall in the Charlotte Harbor
      area, destroying Punta Rassa (this from ATLANTIC HURRICANES by Dunn
      and Miller).  The reanalyzed Best Track file assigns a MSW of about
      100 kts and a SLP of 959 mb at landfall.  Records indicate that the
      surge reached a height of 4.3 metres.

  (2) Sep, 1894 - A Category 2 hurricane with winds of 90 kts made landfall
      in the Charlotte Harbor area.

  (3) Oct, 1910 - The famous "loop" hurricane made landfall a little to the
      south of Port Charlotte and to the north of Fort Myers.   The
      reanalyzed Best Track file assigns a MSW of 105 kts at landfall.

  (4) Oct, 1944 - The final hurricane of 1944 made landfall between the
      Charlotte Harbor area and Tampa.  Based on the old Best Track file
      (the ongoing reanalysis has not yet reached 1944) the MSW appears
      to have been about 105 kts.

     It is interesting to note that all the major hurricanes affecting this
  region occurred after the first of October except for Hurricane Charley,
  which illustrates the fact that Charley's track was most unusual for a
  mid-August hurricane.


  C. Meteorological Observations
  ------------------------------

  (1) Cuba
  --------

     As Charley crossed Cuba hurricane-force winds extended about 13 nm
  to the right of the center and 10 nm to the left of the track--a very
  small hurricane indeed.   As noted above, the strongest wind measured
  on the island during Charley's passage was a 1-min avg sustained wind
  of 103 kts at the Vaisala meteorological station located at the airport
  in Playa Baracoa, just west of Havana City, and just inside the eastern
  eyewall.  The station measured peak gusts of 130 kts and a minimum SLP
  of 974 mb.  Another Vaisala station at San Antonio de los Banos measured
  a maximum 1-min avg wind of 97 kts, gusting to 115 kts, before the
  sensors were blown away.  The station in Guira de Melena, also in the
  eastern eyewall, reported a MSW of 92 kts, peak gusts of 116 kts, and
  a minimum SLP of 971.6 mb.  The minimum SLP of 966 mb mentioned earlier
  was estimated, based on a consideration of the available data and the
  distance of the various stations from the center of the eye.

     Charley was a rather dry hurricane in Cuba--rainfall amounts of 100 to
  150 mm were reported only in restricted areas near the path of the eye.
  Eyewitnesses (no pun intended) reported that the stars could be clearly
  seen during the passage of Charley's eye, and there were some reports of
  a reddish-like appearance of the sky within the eye.  A storm surge of
  3.7 m above MSL was observed in Playa Cajio, where 360 houses were simply
  swept away.  The ocean spread up to 2.6 km inland at that point, and
  penetrated 1.5 km inland at Surgidero de Batabano, where the surge height
  was 2.8 m.

  (The information in this section comes from an excellent Preliminary
  Report on Hurricane Charley in Cuba prepared by Dr. Jose Rubiera, Dr.
  Maritza Ballester and Dr. Cecilia Gonzalez, National Forecasting Center,
  Instituto de Meteorologia, Cuba.)

  (2) Reconnaissance Aircraft
  ---------------------------

     A few comments from Rich Henning, a member of the U. S. Air Force's
  53rd Weather Reconnaissance Squadron:

     "In the case of Charley, the MSLP hung around 970 mb all night (with
  a 700-mb core temperature of only 12-14 C), then it dropped 29 mb (with
  a 10-mb drop in about 90 minutes) as the 700-mb core temperature spiked
  up to 20 C and the eye contracted to 5 miles.  The 13/1930 UTC fix
  included an astonishing description of 138-kt winds ONE MILE southeast
  of the eye center."  Charley provided a fascinating case study of meso-
  scale core processes gone amok.

     As noted above, the lowest CP and peak FLW measured by reconnaissance
  aircraft was 941 mb and 148 kts, respectively, at 13/1956 UTC--shortly
  before Charley's eye made landfall in Florida.

  (3) Florida
  -----------

     As noted above, one of the NHC advisories alluded to an unofficial
  gust (estimated) of 110 kts along with a 943.6 mb SLP measured at Punta
  Gorda.  Storm chaser Mike Theiss was in the Charlotte Harbor area during
  the passage of the eye (which lasted only 5 minutes) and recorded a
  minimum pressure of 944 mb.  Peak gusts of 90 kts and 95 kts were
  estimated at Arcadia and Wauchula, respectively, by Emergency Management
  personnel.  The peak gust recorded by a NWS station was 97 kts at Punta
  Gorda at 13/2035 UTC.

     The following table contains reports of sustained winds (2-min avg)
  exceeding storm force (48 kts):

  County      City           Dir (deg)   Spd (kts)        Date/Time (UTC)
  -----------------------------------------------------------------------
  Osceola     Kissimmee      ???         53               14 / 0035
  Orange      Orlando        160         67               14 / 0105
  Orange      Orlando        130         57               14 / 0129
  Seminole    Sanford        120         63               14 / 0210
  Volusia     Daytona Beach  200         48               14 / 0353
  Volusia     Ormond Beach   100         59               14 / 0315

     The following table contains reports of peak gusts exceeding hurricane
  force (64 kts):

  County      City           Dir (deg)   Spd (kts)        Date/Time (UTC)
  -----------------------------------------------------------------------
  Osceola     Kissimmee      ???         65               14 / 0035
  Orange      Orlando        160         91               14 / 0105
  Orange      Orlando        120         74               14 / 0129
  Seminole    Sanford        120         80               14 / 0210
  Volusia     Daytona Beach  200         72               14 / 0353
  Volusia     Ormond Beach   100         75               14 / 0315

  Note: The above values represent the highest found.  Many of the ASOS
  sites failed during the height of the hurricane.

     Two NASA wind towers near Cape Canaveral at 16.5 m above ground level
  reported sustained winds or 53 kts and 56 kts with both reporting peak
  gusts of 75 kts.  The Wind Shear Alert System (33.5-45.7 m above ground
  level) at Daytona Beach International Airport reported a maximum gust
  of 84 kts.

     As Hurricane Charley passed through Florida rather quickly, rainfall
  amounts were not particularly impressive.  The highest storm total amount
  was 141 mm at Sanford in Seminole County, recorded between 12/1200 and
  14/1200 UTC.  Kissimmee recorded 132 mm during the same 48-hour period.
  Apopka in Orange County recorded a 24-hour total of 102 mm between
  13/0400 and 14/0400 UTC.

     Estimates place the magnitude of the peak storm surge at 4-5 metres.

     More meteorological observations can be found at the following links:

     http://www.srh.noaa.gov/tbw/CharleyWebPage.shtml>

     http://www.srh.noaa.gov/mlb/charley/PSHMLB.txt>


  D. Damage and Casualties
  ------------------------

  (1) Cuba
  --------

     There was a lot of damage reported to housing, crops, trees, power and
  telephone lines and overall infrastructure in the western portion of
  Havana Province, in the west of Havana City and the eastern part of Pinar
  del Rio. However, only four fatalities were reported from this major
  Category 3 hurricane, the first of such an intensity to strike Havana
  Province since 1948.  Very severe damage was sustained by agriculture.
  Hundreds of hectares of banana plantations were destroyed by the winds.
  Even edible root plantations, such as yucca, were heavily damaged.
  Citrus plantations were stripped of fruit, and tens of thousands of
  mango, guava and avocado trees were also blown down and the fruit lost.
  Reports indicate that some 40,500 homes were heavily damaged and 8300
  houses totally destroyed.

     In addition to the four fatalities reported in Cuba, one death was
  attributed to the hurricane in Jamaica.

  (2) Florida
  -----------

     Hurricane Charley was the second most destructive hurricane ever to
  strike the state of Florida, the most destructive being Category 5
  Hurricane Andrew in 1992.  The most concentrated destruction was in
  the communities of Punta Gorda and Port Charlotte in the Charlotte
  Harbor area of southwest Florida.  Total dollar estimates of insured
  losses range from around $7.5 to $11 billion, with total damages likely
  to exceed $15 billion.  In the Charlotte Harbor area, approximately
  10,000 homes were destroyed with another 16,000 severely damaged.  Many
  of these were no doubt mobile homes, which are not designed to withstand
  the extreme winds encountered in intense tropical cyclones.   On the
  day after the passage of Charley an estimated 1.3 million persons were
  without electrical power.  One source estimated the number of persons
  initially without power at 2 million.

     An estimated 1.4 million people evacuated their homes prior to the
  approach of Hurricane Charley.  Some 2300 stayed in local shelters.
  A close call came at Arcadia, about 20 miles inland, where one wall
  collapsed at a civic center which was being utilized as an evacuation
  shelter for some 1200 people.  Very fortunately, only one person was
  injured as a result of the incident and her injuries were minor.
  Charley's winds played havoc with small aircraft at the Charlotte
  Airport with many planes ripped apart by the fierce winds.  An observer
  reported seeing one small plane flying down the runway as if were about
  to take off.

     The U. S. Labor Department reported that new applications for
  unemployment insurance increased by a seasonally adjusted 10,000 to
  343,000 for the week ending 21 August.  Half of this 10,000 rise was
  attributed to claims stemming from Hurricane Charley.

     Hurricane Charley was directly responsible for nine deaths in Florida
  with 16 more indirectly attributable to the storm.  Most of the indirect
  deaths occurred post-storm and were due to things like electrocution,
  heart problems, carbon monoxide poisoning, automobile accidents, and heat
  strokes.  One man died when he fell from a tree, and another died when
  a tree fell on him.  Many post-storm injuries were reported due to things
  like stepping on nails, tripping over debris, chainsaw accidents, auto-
  mobile accidents at intersections where traffic lights were inoperative,
  bruises, cut and strains.   Hurricane-related stress also caused an
  increase in heart attacks and respiratory problems, and many people
  became ill with diarrhea and vomiting after drinking contaminated tap
  water.

  (3) Carolinas and Northward
  ---------------------------

     The Property Claims Service reports that insured damages due to
  Charley were estimated at $25 million in North Carolina and $20 million
  in South Carolina.  In the state of Rhode Island, one death was reported
  as a direct casualty of Hurricane Charley.

  (4) Additional Information
  --------------------------

     More information on storm damage can be found in the reports prepared
  by the Tampa and Melbourne NWS offices referenced in Section C.3 above.
  In addition, many articles dealing with the impact of Hurricane Charley
  and subsequent storms may be found at the following URL:

     http://www.reliefweb.int/w/rwb.nsf/vLND>

     Following are some links supplied by John Wallace which contain
  additional information on the impact of Hurricane Charley:

     http://www.palmbeachpost.com/storm/content/weather/special/storm/2004/atlantic/charley/news.html>

     http://www.palmbeachpost.com/storm/content/business/epaper/2004/08/24/a1d_insure_0824.html>

     http://www.alertnet.org/thenews/fromthefield/109421485563.htm>


  E. Storm Chaser Reports
  -----------------------

     Following are short write-ups sent by storm chasers Mike Theiss and
  Jim Leonard, who were eyewitnesses to Charley's fury in the Charlotte
  Harbor area.

  (1) Mike Theiss
  ---------------

     "I documented Hurricane Charley at Charlotte Harbor and never imagined
  this storm would turn out to be like this.  I had a window blown out from
  a piece of a flying roof, a large telephone pole with a big transformer 
  landed behind my truck, and I had debris slamming against my car during
  the real intense eyewall which lasted for about 20 minutes.  All I can
  say is "UNREAL".  I had heard reports of this storm intensifying, but I
  had no idea it got as strong as it did.  I recorded a pressure of 944 mb
  and was in "Awe" in the eye of this beast.  The core was very small and
  tight and I was in the eye for around 5 minutes.  The damage in the area
  I was in reminded me of Andrew's aftermath, but in a smaller area.  I
  observed concrete structures blown apart, every building at least
  de-roofed, trees snapped in half, and large billboard signs with only
  the poles left twisted beyond belief.  I drove a few miles down the road
  and the damage was bad, but nowhere near the devastation in Charlotte
  Harbor.  I haven't seen any radar loops yet of it making landfall and
  can't wait.  I will post pics in a few days.  I have to focus on getting
  my truck in the shop to get repaired."

     More pictures and video clips may be found at Mike's website:

     http://www.eyeinthetropics.com/>

  (2) Jim Leonard
  ---------------

     "First of all, many thanks to Eric Blake for his frequent radar
  updates during the entire afternoon of the chase.  We started the morning
  at a friend's house at Cape Coral, a suburb of Ft. Myers.     By late
  morning I noticed the eye of the hurricane began to show a slight NNE
  heading, so the plan to head up to Sarasota was adjusted southward.  As
  we headed over the bridge from Cape Coral to the city of Ft. Myers, I saw
  a long rainfree cloud base approaching from the south.  I turned the van
  around and went back to the bridge to get in a better position.   The
  cloud line moved over our position without doing anything when all of a
  sudden overhead and to the NE I noticed rapid scud motion which I figured
  would lead to a potential tornado.  At that moment I saw a large spray
  ring on the bay about a half mile to the northeast.  We could only see
  this for a few seconds as there were too many trees in the way.  Then a
  heavy rain over us occurred and obscured the vortex.  Once the rain let
  up the cloud base circulation moved NW.  At that time I saw rapidly
  circulating scud tags for a couple minutes more before losing sight of
  it.

     "As the afternoon wore on we drove north and south between Ft. Myers
  and North Port like a yo-yo as the center of the hurricane wobbled NNE.
  An average-sized eyewall would have been much easier to place ourselves
  in its track.  As this storm was so small in size you had to be really
  accurate to get in its path.  Also, we had to find the safest structure
  and have the best tree scenes during the strongest winds.   We ended up
  in the ground-level parking garage of a hospital between Port Charlotte
  and Charlotte Harbor.   While searching for our spot the winds were
  averaging 40 to 50 kts in gusts when a sudden gust of 70 kts occurred.
  We had to take cover fast!  A second gust probably 60 kts or so occurred
  as a tree branch blasted out my left rear window of the van--this while
  video was rolling, great audio!  Minutes later we positioned ourselves
  in the parking garage as all hell broke loose!

     "The core of this hurricane was so small and its forward movement so
  rapid that the winds increased amazingly fast.  We were probably in 
  an area of better constructed buildings as just a few blocks either side 
  of us damage to structures was much worse.  During the first half winds 
  in my estimation were in the 120 to 130 mph range.  There were many 
  sections of roofing material flying around and I filmed the roof of a
  bank as it came off in pieces across the parking lot from us.  The wind
  speed dropped off dramatically as the northern portion of the eye moved 
  across--this took about two or three minutes--when all of a sudden the 
  winds shifted to the NNE, then eventually north with the rain really 
  blinding at this time.  I estimated the gusts at the height of this part 
  as high as 135 to 140 mph.  The strongest winds on the back side lasted
  about 15 to 20 minutes.  It was like going through a 15-mile-wide
  tornado!  About 30 minutes later we began a brief damage survey before we
  headed back home.   I noticed how the intense damage was definitely in
  streaks, which is typical in rapidly deepening hurricanes.  The eyewall 
  was characterized by very turbulent gusts, which is a common occurrence 
  when the eyewall convection is very intense.    This would account for 
  the streaks of intense damage.   This was first noted in Hurricane Celia 
  in August, 1970, in Corpus Christi, Texas, a storm with a similar central
  pressure and storm size at landfall."

     More information may be available on Jim's website:

     http://www.cyclonejim.com/>

  (Report written by Gary Padgett)



                         HURRICANE DANIELLE
                               (TC-04)
                           13 - 21 August
               --------------------------------------

  A. Storm Origins
  ----------------

     A strong and well-organized tropical wave, possibly already at
  tropical depression intensity, left the African coast on 12 August and
  was first mentioned in NHC/TPC's Tropical Weather Outlook at 2230 UTC
  later that day.  The wave was accompanied by a broad LOW, and moving in 
  a westerly direction, was centred around 260 nm southeast of the Cape 
  Verde Islands by early the next morning.  The disturbance proceeded to 
  develop quickly and was soon upgraded to Tropical Depression Four at 
  13/1500 UTC.  The system was designated Tropical Storm Danielle at 
  14/0300 UTC while passing south of the Cape Verde Islands.  Danielle's 
  MSW may have reached 35 kts six hours earlier, but the forecaster chose 
  to wait until satellite CI estimates from TAFB and SAB had both 
  reached a consensus of 2.5.


  B. Synoptic History
  -------------------

     Tropical Storm Danielle was heading westward through a low shear 
  environment and warm SSTs southeast of a mid-level ridge.  These 
  conditions encouraged gradual strengthening during the 14th.  Banding 
  features became better defined over the southern semicircle before 
  beginning to wrap around the centre at 2100 UTC on 14 August.   The 
  MSW had risen to 55 kts by this time, and a 14/2247 UTC SSM/I microwave
  pass detected a developing eye.  Danielle intensified into the first
  Cape Verde hurricane of the year at 15/0300 UTC when the MSW reached
  65 kts and CI estimates reached 4.0.  Continued strengthening on the
  15th brought the sustained winds up to 85 kts by 2100 UTC.  Just prior
  to this, the wind field surrounding Hurricane Danielle was reduced after
  a 15/2016 UTC QuikScat pass had indicated a smaller wind radii than
  previously analyzed.   After this amendment, hurricane-force winds
  extended 20 nm from the centre while gales reached out as far as 90 nm
  in the southeastern quadrant.

     Danielle's track gradually curved towards the northwest on the 16th
  as a large area of troughing began to develop over the western Atlantic.
  The hurricane was looking healthy on infrared satellite images with a
  circular CDO punctured by a small eye, and good outflow was present in
  all quadrants.  Danielle arrived at its peak strength of 90 kts at
  16/0300 UTC, and this intensity held constant for the rest of the day.
  A single enhanced BD curve infrared image at 16/1800 UTC suggested a
  stronger system with CI estimates of 5.5., so for a short time Danielle's
  MSW could possibly have reached 100 kts.  (It will be interesting to see
  if the winds are upped any during post-storm analysis.)   By 16/2100 UTC
  Hurricane Danielle had turned north-northwestwards with its CDO less
  symmetrical than earlier in the day, a sign that the environment was
  becoming more hostile.  The 10-nm eye persisted through the night, but
  this feature had faded somewhat by 17/0300 UTC.   However, it continued
  to make intermittent appearances throughout the day.  Veering northwards,
  Danielle remained a 90-kt hurricane until 17/2100 UTC when the MSW 
  started to drop off. 

     The intensity dropped a little further at 0300 UTC on 18 August as 
  Hurricane Danielle tracked northward.  The partially-exposed LLCC 
  had become difficult to locate by this time as a result of southwesterly 
  shear and dry air intrusion, but both microwave imagery and a 17/2100 
  UTC QuikScat pass indicated that the centre was located southwest of 
  the deep convection.  The cyclone began to weaken more rapidly and was 
  barely at hurricane strength at 18/0900 UTC.  It was downgraded to a 
  55-kt tropical storm six hours later as it turned towards the north-
  northeast and began to decelerate.  At 18/2100 UTC the centre of 
  Danielle passed near a drifting buoy which measured a CP of 1008.6 mb.
  Despite the shear Danielle continued to generate bursts of convection 
  on the 19th, mainly to the northeast of the LLCC.

     The storm was about to perform a rather jagged hairpin turn over the
  next few days within a low to mid-level ridge, south of a passing short-
  wave trough, and east of a strong mid to upper-level LOW near 33N/45W.   
  The hurricane was being steered in the flow between the mid to upper-
  level LOW and the subtropical ridge.  Danielle headed northeastward, 
  then eastward and finally wound up stationary at 19/2100 UTC.   For 
  the next day or two Danielle was expected to meander in the same general
  area.  This was a completely different scenario to what some of the
  forecast models had been initially forecasting.   Danielle was 
  originally predicted to undergo extratropical transition and pass 
  through the Azores island chain.  Instead, a HIGH built in the vicinity 
  of the Azores and blocked the cyclone's path in that direction.

     In defiance of hostile upper-level shear and high surface pressures, 
  sporadic bursts of deep convection continued to fire up northeast of the 
  well-defined centre.  Danielle was maintained as a 35-kt tropical storm 
  for the majority of the 20th.  It was downgraded to a tropical depression
  at 2100 UTC when most or all of the convective activity had subsided 
  and only the LLCC remained.  Movement continued to be very slow and 
  erratic but eventually a northwesterly crawl became established late on
  the 20th.  This heading persisted into the 21st before Danielle assumed
  a more westerly track.  The system was dropped by NHC at 21/1500 UTC when
  the final advisory was issued.  The depression was then located a little
  over 740 nm west-southwest of Azores.  The remnant circulation crept
  slowly westwards until the 22nd when movement became virtually static.
  As the ex-hurricane started to drift slowly north on the 23rd, convection
  redeveloped near to and east of the center, and the 23/1502 UTC STWO
  mentioned the possibility that Danielle could regenerate back into a
  tropical depression.  However, deep convection soon diminished and
  conditions started to become more unfavourable for redevelopment.  The
  LOW continued north to northwest at a quicker pace on the 24th.  Finally,
  whatever was left of Danielle was gradually incorporated into the warm 
  sector of a mid-latitude LOW on the 25th.


  C. Damage and Casualties
  ------------------------

     No damage or casualties are known as a result of Hurricane Danielle.

  (Report written by Kevin Boyle)



                         TROPICAL STORM EARL
                               (TC-05)
                           13 - 16 August
               ---------------------------------------

  A. Storm Origins
  ----------------

     The tropical wave that became Tropical Storm Earl exited the coast of 
  Africa around 6 August.  It was at first a rather dull, uninteresting 
  feature as it tracked its way westwards across the tropical Atlantic.
  The wave was initially mentioned in NHC/TPC's STWO at 1130 UTC on  
  12 August when it was centred about 1100 nm east of the Windward 
  Islands.  A slow development potential was noted in this statement.  The 
  disturbance gradually became a little better organized through the 12th 
  and 13th.  As environmental conditions appeared to be favourable for 
  cyclogenesis, the 13/1630 UTC STWO warned that a tropical depression
  could form within a day or two and that interests in the Lesser Antilles
  and the eastern Caribbean Sea should closely monitor the system over the
  next few days.


  B. Synoptic History
  -------------------

     Tropical Depression Five formed at 2100 UTC on the 13th roughly 910 nm
  or so east-southeast of the Windward Islands in an area historically 
  hostile for tropical cyclone development.  However, it was this same 
  general area that had spawned Hurricane Charley only a few days before.
  TD-05 was moving rather quickly westwards--perhaps too fast for its own
  good--through a favourable environment, so further strengthening seemed
  likely.    However, it took another 24 hours for the depression to reach
  tropical storm status.  TD-05 was named Earl at 14/2100 UTC, by which
  time its rapid forward speed of around 20 kts had taken it to within
  nearly 350 nm southeast of Barbados.     In response to the upgrade, the
  governments of the Windward Islands each issued a tropical storm warning.
  Tropical Storm Earl's motion was controlled by the steering flow provided
  by a strong mid-level ridge to its north, and this feature was forecast
  to persist and continue driving the tropical cyclone smartly westwards
  for the rest of its short Caribbean cruise.

     Some models gave the U. S. a fright by suggesting that in the long-
  term Earl could recurve northwards as a Category 2 hurricane towards the
  Gulf Coast or even Florida.    Thankfully, this scenario failed to 
  materialize, especially so soon after Charley's devastating landfall
  in Florida.  Even though the overall cloud pattern appeared slightly 
  elongated on the morning of the 15th, deep convection increased 
  significantly near the centre and there was a little strengthening. 
  Satellite intensity estimates had reached T3.0, but because banding 
  features were barely discernible, the MSW was increased to only 40 kts, 
  which turned out to be the maximum intensity of Earl.  While Earl's 
  satellite representation at 15/1500 UTC showed an organized system 
  with impressive outflow channels, the Hurricane Hunter reconnaissance
  aircraft investigating Earl struggled to find a well-defined closed LLCC.
  Meanwhile, the system was bombing through the Windward Islands and passed
  just south of Grenada at around 15/1800 UTC before entering the Caribbean
  Sea.

     Tropical Storm Earl deteriorated dramatically overnight, and by the
  next morning convection had weakened markedly and the impressive outflow
  pattern that the system possessed earlier had been replaced by a uniform
  easterly flow.  In fact, QuikScat data revealed that Earl had degenerated
  to the extent that it resembled an open wave.      The government of
  Venezuela refused to allow aircraft into their airspace for
  reconnaissance purposes so NHC decided to continue advisories for a
  little while longer, especially as Earl was in close proximity to land.
  Microwave and QuikScat imagery indicated that the poorly-defined centre
  or wave axis was racing ahead of the convection.  The MSW was kept at
  40 kts despite T-numbers having dropped to 2.0 (30 kts), and this was due
  to a burst of deep convection near the alleged centre.   An Air Force
  reconnaissance aircraft reached the storm the next morning and was unable
  to find any sign of a closed circulation associated with Earl.  A Quik-
  Scat pass also failed to locate the LLCC.  On this basis, NHC downgraded
  Earl to an open wave with 35-kt winds and issued the final advisory at
  16/1500 UTC.  The wave continued to generate winds of tropical storm
  force as it passed south of Jamaica.  Models continued to indicate the
  possibility that the remnant of Earl could re-instate itself as a
  tropical cyclone and reach hurricane strength as it tracked westward
  through the Caribbean.  However, a planned follow-up reconnaissance
  flight to investigate the tropical wave on the 17th was cancelled since
  the system failed to show any signs of re-organization.  Continuing its
  journey westwards, the remnants of Earl moved inland over Honduras/
  northern Nicaragua later that same day, ultimately re-emerging over
  Eastern Pacific waters around 18 or 19 August and redeveloping as
  Hurricane Frank.
 

  C. Damage and Casualties
  ------------------------

     According to the Caribbean Disaster Emergency Response Agency, 
  damages on Grenada were minor and amounted to no more than broken
  branches, a few fallen trees, and downed power lines.  Ten houses had
  their roofs blown off while one home suffered partial roof damage.
  There were reports of flooding in the St. George parish area of
  Grenada, and the Point Saline International airport was closed for a
  short time.  There were no reports of injuries or casualties.  Also,
  there were no reports of casualties or damages from any of the other
  Windward Islands due to Tropical Storm Earl.

  (Report written by Kevin Boyle)



                            HURRICANE FRANCES
                                 (TC-06)
                         25 August - 10 September
               --------------------------------------------

  A. Storm Origins
  ----------------

     A vigorous tropical wave moved off the African coast early on the
  21st of August.  Conditions downstream were considered favorable for
  tropical cyclone development, but the wave initially showed no signs
  of cyclogenesis.  By the next afternoon the wave was passing south of
  the Cape Verde Islands and convection had become slightly better
  organized.  The wave continued marching westward on the 23rd with little
  change, but by early morning of 24 August was again exhibiting increased
  convective organization.  The disturbance was then located approximately
  500 nm west-southwest of the Cape Verdes, moving westward at 13 kts.
  The convective appearance continued to improve into the evening hours,
  and at 0300 UTC on 25 August the first advisory on Tropical Depression 06
  was issued by TPC/NHC.


  B. Synoptic History
  -------------------

     The initial advisory placed the center of TD-06 about 750 nm west-
  southwest of the Cape Verde Islands, moving west-northwestward at 15 kts.
  The MSW was estimated at 25 kts and the system was forecast to reach
  tropical storm intensity within 12 hours.  During the morning of the 25th
  convection associated with the depression looked somewhat ragged, but
  during the afternoon the system showed renewed organization with a small
  CDO and a convective band wrapped around the western semicircle.  Since
  satellite intensity estimates were 35 kts from TAFB and SAB, TD-06 was
  upgraded to Tropical Storm Frances at 25/2100 UTC.  The new tropical
  storm was located about 1250 nm east of the Lesser Antilles and moving
  westward at 17 kts.  During the early morning hours of 26 August Frances'
  cloud pattern improved significantly.  Winds were upped to 50 kts in the
  26/0900 UTC advisory, and there were some indications that a banding-type
  eye was trying to form.  The intensification trend continued and at 2100
  UTC Frances was upgraded to hurricane status with 70-kt winds.  Upper-
  level outflow was good to excellent in all quadrants.

     Hurricane Frances continued to steadily strengthen on the 27th, and
  at 2100 UTC the MSW was upped to 100 kts, thereby making Frances the
  third intense hurricane (Category 3+ on the Saffir/Simpson Scale) of the
  season.  The hurricane was then centered about 700 nm east of the Lee-
  ward Islands.   Up to this point Frances had been moving on a west-
  northwesterly track, being guided by a strong subtropical ridge to the
  north.  For a couple of days a weakening of the ridge due to a higher-
  latitude shortwave trough had been forecast, and this verified very
  well.  On 27 August Hurricane Frances began a northwestward jog in
  response to the temporary weakening of the ridge.  And, just as had
  earlier been forecast, in a couple of days the cyclone returned to a
  west-northwestward heading as the trough lifted out and the ridge
  strengthened over the western Atlantic.  The intensity remained more
  or less constant late on the 27th and early on the 28th due to the first
  of several eyewall replacement cycles.  After completing this cycle,
  Frances' intensity began to climb again, reaching an initial peak of
  115 kts at 28/2100 UTC, based on an analysis of satellite intensity
  estimates, both manual and objective.  Frances was located about 600 nm
  east of the northern Leeward Islands as it became the season's second 
  Category 4 hurricane.  

     As forecast, the subtropical ridge began to build and expand north of
  the hurricane.  Consequently, on the 29th Frances began to move back to
  a heading of just slightly north of due west.  The first U. S. Air Force
  Reserves reconnaissance aircraft reached Frances during the afternoon
  of 29 August.  A minimum pressure of 949 mb was measured, and the peak
  FLW found was 112 kts.  Since the plane had not sampled the entire
  circulation by the 2100 UTC advisory time, the MSW was left at 115 kts,
  but this was reduced to 110 kts at 30/0300 UTC.  The pressure was up to
  954 mb at 2300 UTC and the highest 700-mb FLW reported was 118 kts at a
  position about 9 nm northwest of the center.  At 0900 UTC the MSW was
  reduced further to 105 kts.  The most recent peak FLW found by an
  aircraft was only 99 kts, but the cyclone appeared much better organized
  and satellite intensity estimates were ranging from 102 to 115 kts.
  Frances was located at 30/0900 UTC about 335 nm east of the northern
  Leeward Islands, moving west at around 8 kts.

     Frances' intensity began to edge upward again on 31 August as the
  pressure began to fall once more.  The MSW was upped to 110 kts at
  31/0000 UTC and to 115 kts at 1200 UTC.  The hurricane had completed
  an eyewall replacement cycle early in the morning, and the 1200 UTC
  reconnaissance mission reported a peak FLW of 138 kts at 700 mb and an
  extrapolated CP of 944 mb.   Six hours later the MSW was nudged up again
  to 120 kts.   A reconnaissance plane had found winds to 144 kts at
  700 mb with an extrapolated CP of 938 mb.  Also, the Stepped Frequency
  Microwave Radiometer (SFMR) on the NOAA P-3 plane measured surface winds
  of 118 kts, and T-numbers had reached 6.5 from the satellite agencies.
  Thus, Frances was a solid Category 4 hurricane on the Saffir/Simpson
  Scale located about 125 nm north of San Juan, Puerto Rico, moving west
  at 15 kts.  Hurricane force winds extended outward 60 nm from the center
  in the northwest quadrant and from 35 to 50 nm in the other quadrants
  while gales covered a zone almost 300 nm in diameter.

     Large, severe Hurricane Frances maintained its intensity through the
  first day of September.  The lowest CP in the storm's history--935 mb--
  was measured during the wee hours by a reconnaissance aircraft.   The
  storm was getting set to go through another eyewall replacement cycle.
  A reconnaissance crew very early on the 1st reported finding three
  concentric wind maxima, and the reconnaissance fix at 01/1719 UTC
  reported a double eyewall.  The discussion bulletin at 01/2100 UTC noted
  that data from the SFMR on the NOAA P-3 plane were used to decrease the
  wind radii estimates in the northwestern quadrant.  As a result of this,
  it was decided to delay the issuance of a hurricane watch for the Florida
  East Coast.  Very early on 2 September a dropsonde in the northeastern
  eyewall reported surface winds near 125 kts.  This, along with CI esti-
  mates of 127 kts from all three satellite agencies, was the basis for
  raising Frances' MSW to the peak value of 125 kts at 02/0900 UTC.  The
  attendant CP was measured at 936 mb.  The storm's center had passed only
  25 nm north of Grand Turk Island around 2100 UTC on 1 September, and at
  0900 UTC on the 2nd was centered a like distance north of Mayaguana
  Island in the Bahamas.  This position is also 450 nm east-southeast of
  Palm Beach, Florida.

     At 1500 UTC on 2 September hurricane warnings were ordered up from
  Florida City northward to Flagler Beach, including Lake Okeechobee.
  Hurricane Frances was located very near San Salvador Island in the
  Bahamas around 1800 UTC, still at its peak intensity of 125 kts.  A
  sustained wind of 99 kts was reported from the island.  By 2100 UTC data
  from a reconnaissance plane indicated that Frances' CP had risen to
  948 mb, so the wind was adjusted downward slightly to 120 kts.  It was
  initially thought that this might represent a minor fluctuation in
  intensity, but Frances continued to slowly weaken as it progressed west-
  northwestward through the Bahamas.  The MSW was lowered to 110 kts in
  the 09/0300 UTC advisory, placing Frances back at Category 3 status.
  At 0600 UTC the hurricane was centered near the northern end of Cat
  Island, moving slowly west-northwestward at 8 kts.  The MSW at this
  time was again decreased slightly to 105 kts.   Six hours later Frances
  was located over Eleuthera Island or about 225 nm east-southeast of the
  southeastern coast of Florida.  Frances was continuing to slowly weaken.
  While cold convective tops were still being generated over the center,
  Hurricane Hunter and Bahamian radar data showed that they were coming
  from eyewall fragments in the northeastern quadrant.  At 0900 UTC the
  CP had risen to 954 mb and the peak FLW was 98 kts.   Reconnaissance
  data during the later morning indicated that the inner core of Frances
  had deteriorated since the previous day, likely due to some upper-level
  southwesterly shear between 300-250 mb which was undercutting the
  impressive outflow layer.

     Shortly before 03/1500 UTC, Little Harbor in the Abacos group reported
  a wind gust of 100 kts.  At 1800 UTC Frances was centered very near the
  southern tip of Great Abaco Island.  Winds were down to 100 kts and the
  CP at 961 mb as the large hurricane continued to slowly weaken.   The
  MSW was decreased to 90 kts in the 04/0000 UTC intermediate advisory.
  This downgrade to Category 2 status was based upon data from a variety
  of sources:  radar, satellite, reconnaissance data, and surface reports.
  As the day progressed the inner core convection became a little more
  symmetric and reconnaissance observations indicated that an eyewall might
  be reforming, but Frances was to remain at 90 kts until its landfall on
  the Florida coast.  At 1200 UTC the hurricane was centered very near
  Freeport on Grand Bahama Island.   Settlement Point was reporting
  sustained winds to hurricane force at that hour with gusts to 70 kts.
  The SLP was 974.5 mb and falling rapidly.  Strong winds continued to
  buffet the island for much of the day as Frances remained stalled.  By 
  1500 UTC WSR-88D radar showed that Frances had developed a large eye
  about 70 nm in diameter.   During the evening hours of 4 September the
  large hurricane began once more to drift west-northwestward toward the
  southeastern coast of Florida.  The center was located at that hour
  approximately 45 nm east-northeast of Palm Beach with strong winds
  already affecting the coastline.

     The center of Hurricane Frances' large eye reached the coast around
  0500 UTC on Sunday, 5 September, near Sewall's Point, Florida.  This
  location is also in the vicinity of Stuart, Jensen Beach and Port
  Salerno.  The landfall intensity was 90 kts with the CP near 960 mb.
  Hurricane force winds extended outward from the center 75 nm in the
  northern semicircle and about 60 nm to the south.  Once inland Frances
  continued to move west-northwestward across the central Florida Peninsula
  while slowly weakening.  At 05/1500 UTC Frances' center was located by
  radar approximately 130 km east-southeast of Tampa; however, due to the
  large size of the circulation, Grand Bahama Island was still reporting
  sustained winds of tropical storm force while winds of 40-45 kts in
  feeder bands were being reported from the Keys.  At 2100 UTC Frances
  was downgraded to a 60-kt tropical storm, even though the CP was quite
  low at 975 mb.   As the large storm continued to weaken over central
  Florida, it maintained well-organized convective banding and tropical
  storm-force winds over a large area.   The St. Augustine C-MAN reported
  a sustained 56-kt wind observation at 05/2200 UTC, and around the same
  time Tampa reported a MSLP of 977.7 mb.

     By 0300 UTC on 6 September the center of Tropical Storm Frances was
  moving out over the warm waters of the northeastern Gulf of Mexico just
  northwest of Tampa with the MSW estimated at 55 kts.  The NHC discussion
  bulletin at 0900 UTC noted that thunderstorm activity had steadily
  increased near the center over the past few hours, and Doppler radar data
  from Tampa and Tallahassee, as well as observations from a reconnaissance
  mission, indicated that a ragged eye was trying to form.  The forecast
  for that hour indicated the possibility that Frances might regain
  hurricane intensity before making landfall in the eastern Florida
  Panhandle.   That forecast, however, failed to materialize.  The MSW
  remained at 55 kts until Frances made landfall near St. Marks, Florida,
  around 1800 UTC on the 6th (Labor Day).  The final TPC/NHC advisory on
  Frances, issued at 07/0300 UTC, downgraded the system to tropical depres-
  sion status with the center located about 50 km southwest of Albany,
  Georgia.  The depression was moving toward the north-northwest at 10 kts
  and was expected to make a northward turn during the next 24 hours.

     With Frances' winds winding down, the primary concern turned to inland
  flooding from the expected heavy rainfall.  As has been the procedure for
  the past several years, warning responsibility for the remnants of
  Frances was handed over to the Hydrometeorological Prediction Center
  (HPC) in Maryland.  Tropical Depression Frances continued northwestward
  through southwestern Georgia into extreme eastern Alabama before
  recurving to a general northeastward heading.  At 0300 UTC on 8 September
  the depression was located about 55 km northeast of Atlanta, Georgia.
  The system continued northward to northeastward, passing through extreme
  western North Carolina, the northeastern tip of Tennessee, extreme
  western Virginia and into West Virginia.  At 09/0300 UTC the remnant LOW
  was located over West Virginia about 280 km west of Washington, D. C.,
  and had accelerated some to 17 kts.  The heading subsequently became
  more northeasterly and 24 hours later, when HPC issued their final
  advisory, the LOW had crossed western Pennsylvania and New York and had
  moved into Canada, being located about 40 km south-southwest of Quebec
  City.  Frances' remnant LOW was then moving northeastward at 21 kts.
  The Ocean Prediction Center (OPC) issued a couple of gale warnings on
  the LOW, even though the center was inland, as it moved into northern
  Maine.  The final definite mention of the system was at 10/1200 UTC.
  Frances' remnant was the southwesternmost center of a complex LOW with
  the main center near 48N/65W, moving to the east-northeast at 40 kts.
  The post-Frances center was moving in the same direction at 30 kts.


  C. Meteorological Observations
  ------------------------------

     Many locations reported gusts well above hurricane intensity, and
  several reported sustained winds at or higher than hurricane force.
  The highest sustained wind/gust combination I could locate was at
  Ft. Pierce, which reported a 1-min avg wind of 70 kts with a peak gust
  of 94 kts around 0400 UTC on 5 September.   Port Mayaca Lock in Martin
  County reported a peak sustained wind of 74 kts at 05/0500 UTC.  There
  was a report from the public of a gust to 92 kts at 05/1145 UTC from
  a location in Martin County (27.12N/80.26W)--the anemometer height was
  8.5 meters.  Port Orange reported a 3-day rainfall total of 383 mm,
  while Deland Coop. in Volusia County reported a storm total of 286 mm.
  The Corp of Engineers reported a storm surge value at the St. Lucie
  Lock of 1.8 meters.  Other surge estimates ranged from 1.8 meters
  at Cocoa Beach to 2.4 meters at Vero Beach.

     Many, many more observations can be found in the storm report
  compiled by the Melbourne NWS office.  The link is:

     http://www.srh.noaa.gov/mlb/frances/products/PSH.txt>

     As of this writing, the official TPC/NHC storm report on Hurricane
  Frances is not yet available online, but it is highly likely that it
  will be within the next few weeks.  The report will contain additional
  observations, and interested persons should periodically check the
  following link:

     http://www.nhc.noaa.gov/2004atlan.shtml>

     Hurricane Frances produced heavy rainfalls from Florida to Canada.
  A significant area in central Florida plus western North Carolina
  recorded storm totals exceeding 15 inches (380 mm), while portions of
  Virginia and eastern Georgia measured totals exceeding 10 inches
  (255 mm).  A small zone in southeastern New York experienced a storm
  total in excess of 7 inches (178 mm).  The following link contains a map
  depicting storm total rainfall associated with Frances over the eastern
  U. S.:

     http://www.hpc.ncep.noaa.gov/tropical/rain/frances2004rain.gif>

     Following are a few selected storm totals gleaned from the HPC
  advisories on Frances.  These represent periods of various length ranging
  usually from 30 to 42 hours.

     Linville Falls, North Carolina                  459 mm
     Edgemont, North Carolina                        422 mm
     Lake Toxaway, North Carolina                    367 mm
     West Point, Ohio                                338 mm
     Tallulah Falls, Georgia                         255 mm
     Walhalla, South Carolina                        253 mm
     Lake City, Florida                              236 mm
     Ellamoore, West Virginia                        229 mm
     Nicholls Knobb, Virginia                        209 mm
     Busted Rock, Virginia                           187 mm
     White Plains, New York                          163 mm
     Hartwell Dam, Georgia                           157 mm
     Altoona, Pennsylvania                           148 mm
     Middlesboro, Kentucky                           129 mm
     Newfound Gap, Tennessee                         103 mm

     Amounts exceeding 50 mm were reported from many stations across New
  England.  More rainfall information can be found in the HPC advisories,
  which are archived at the following link:

     http://www.hpc.ncep.noaa.gov/tropical2004/2004storms.shtml>

     Frances brought significant rainfall to parts of Canada also.  Heavy
  rains were recorded just north of eastern Lake Erie and over south-
  eastern Ontario.  A few spots measured new record 24-hr totals:

     Cobourg               82.2 mm
     Kingston             137.0 mm
     Ottawa (Airport)     123.0 mm
     Trenton              111.8 mm

     Frances' remnants brought heavy rain also to the Maritime Provinces.
  Charlo, New Brunswick, recorded 98.3 mm in association with the storm,
  and Miscou Island and Miramichi, New Brunswick, each received over
  50 mm.  In Newfoundland, Deer Lake and Stephenville each set a new record
  for the date (10 September) with 71.9 mm and 65.2 mm, respectively.
  Twillingate measured 62.3 mm, while Badger and Corner Brook each recorded
  over 50 mm.   (A special thanks to Chris Fogarty for the Canadian data.)


  D. Damage and Casualties
  ------------------------

     Frances was not nearly as intense as the devastating Charley, but
  was much larger.   Charley's swath of hurricane-force winds was only
  40 nm, whereas sustained hurricane winds in Frances stretched over
  a diameter of 135 nm.   In areas where there are many trees in very
  close proximity to houses, sustained winds of storm force (>48 kts)
  with accompanying gusts to hurricane force can cause an amazing amount
  of destruction by trees being blown onto houses and cars.   The
  diameter of storm-force winds in association with Hurricane Charley
  was about 90 nm, whereas in Frances winds 50 kts or higher covered a
  zone 220 nm across.  Hence, Hurricane Frances did not produce the
  extreme structural damage caused by Charley over a small area, but did
  produce significant damage over a considerable portion of the entire
  Florida Peninsula.

     Beach erosion was moderate to severe from Cocoa Beach to Vero Beach
  with roads and boardwalks experiencing significant damage.   Street
  flooding was a problem in many cities, and a few tornadoes touched down,
  causing minor damage.   St. Lucie County was one of the hardest hit
  counties with initial reports indicating damage of $2.5 billion to
  public and private facilities.  The Ft. Pierce Municipal Marina was
  destroyed as were most of the hangars at Ft. Pierce Airport.  Thousands
  of residences experiences roof and tree damage, and many thousands of
  mobile homes were damaged or destroyed.

     Hurricane Frances caused more damage to the Kennedy Space Center at
  Cape Canaveral than any storm in history, according to a Fox News story.
  The storm blew an estimated 1000 exterior panels from a giant building
  where spaceships are assembled, creating about 40,000 square feet of
  "open window" on two sides of the building.   Sustained winds reached
  60 kts at the facility with gusts to 82 kts.   The shuttle hangars and
  spaceships were not damaged, but part of the roof came off the building
  where the shuttles' thermal tiles are made, resulting in a potential
  blow to NASA's return-to-flight effort.

     The storm summary on Frances in the latest release by Dr. Bill Gray
  and associates at Colorado State University indicates that insured losses
  from Frances are estimated at around $4 billion, which would suggest
  a total damage figure for the storm of around $8 billion.  The hurricane
  has also been blamed for 24 deaths.  All of the CSU forecast releases
  may be found at the following URL:

     http://tropical.atmos.colostate.edu/forecasts/index.html>

     The post-storm report prepared by the Melbourne NWS office, referenced
  in the above section, also contains further details regarding storm
  damage in the counties within its area of jurisdiction.  Also, the
  official TPC/NHC report, when it becomes available, should be consulted
  as the best available source of damage figures relating to Hurricane
  Frances.

  (Report written by Gary Padgett)



                           HURRICANE GASTON
                               (TC-07)
                       27 August - 2 September
             -------------------------------------------

  Editor's Note:  The report below, authored by Kevin Boyle, was written
  before the post-analysis upgrade of Tropical Storm Gaston to hurricane
  status.  However, it does point out the definite possibility of such
  happening.

  A. Storm Origins
  ----------------

     At 1818 UTC 27 August NHC issued a Special Tropical Disturbance 
  Statement on a LOW centred about 120 nm southeast of Charleston, 
  South Carolina.  This feature had formed from the remnants of an old 
  frontal system which extended from the United States East Coast and 
  out into the Atlantic. This front also instigated the development of 
  Tropical Storm Hermine on 29 August.  The disturbance that became 
  Gaston was first mentioned in NHC's Tropical Weather Outlooks on 26 
  August, but it wasn't until the next day that significant development 
  began.  The system then proceeded to organize quite quickly, resulting 
  in the Special Bulletin at 27/1818 UTC.  This statement announced that 
  advisories would be started at 2100 UTC and a tropical storm watch 
  would be put into place for portions of the Georgia and South Carolina 
  coasts.

     At the time of the first warning, Tropical Depression Seven was
  still loitering nearly 130 nm southeast of Charleston, South Carolina. 
  Steering currents controlling the system were very weak and continued 
  to be so during the 28th.  However, a ridge developing to the north and 
  east began to nudge the tropical cyclone westwards during the course of 
  the day.  In the days ahead, this heading was predicted to turn 
  northwards as the ridge was weakened by a shortwave trough.  Tropical 
  Depression Seven continued to quickly wind up and tropical storm 
  intensity was soon reached at 1500 UTC 28 August.  Visible satellite 
  imagery showed a curved convective band wrapping halfway around 
  the centre at this time.  The watch issued for South Carolina was 
  changed to a warning in response to the upgrade.


  B. Synoptic History
  -------------------

     An Air Force Reserve Hurricane Hunter aircraft was sent to investigate
  the tropical cyclone and discovered that Gaston was a stronger storm 
  than previously thought.  The plane found 59-kt 450-m flight-level 
  winds to the north of the centre and a CP of 996 mb.  The MSW was 
  raised to 45 kts at 28/2100 UTC on the basis of these measurements.  In 
  the event Gaston strengthened to a Category 1 hurricane before 
  landfall, which seemed likely, a hurricane watch was issued for the 
  South Carolina coast, and this was then raised to a warning at 28/2345 
  UTC.  Hurricane Hunter aircraft investigating Tropical Storm Gaston 
  at 29/0300 UTC found 60-kt flight-level winds to the southwest of the 
  centre, a CP of 994 mb, and reported winds of 35-kts extending about 
  50 nm or less in all quadrants.  In addition, the plane reported a 35-nm 
  diameter eye with a partial eyewall, and this observation was backed up 
  by radar from both Charleston and Wilmington which showed that the 
  eye was open to the south.  At 29/1000 UTC Gaston was edging closer to 
  land and was about 35 nm east-southeast of Charleston, South Carolina, 
  at this time.  Its associated MSW had risen to 60 kts and it seemed only 
  a matter of time before Gaston would attain hurricane intensity.  The 
  storm had made its predicted turn towards the north in response to an 
  advancing shortwave trough embedded in the westerlies. 

     Gaston made landfall just west of McClellanville, South Carolina, 
  around 29/1400 UTC.  WSR-88D velocity data indicated that the storm 
  came ashore at just below hurricane strength.  There is every possibility
  that Gaston did move onshore as a minimal hurricane and a post-
  analysis of the system may upgrade it to a Category One hurricane 
  later.  Tracking northwards, Gaston weakened to a 35-kt tropical storm 
  at 29/2100 UTC, and after downgrading it to a tropical depression at 
  30/0300 UTC, NHC stopped issuing advisories at this point and passed 
  warning responsibility to the Hydrometeorological Prediction Center. 
  Initially, the remnants of Gaston were forecast to absorb Tropical 
  Storm Hermine, but this scenario never materialized.  Instead, Hermine 
  sped northward toward New England, leaving behind the slower moving
  Gaston.

     By 30/0300 UTC Tropical Depression Gaston had reached the South 
  Carolina/North Carolina border.  Turning northeastwards, Gaston 
  moved back over water and was located east of the Delmarva Peninsula 
  and southeast of the Chesapeake Bay at 0300 UTC 31 August.  At this 
  time NHC re-assumed warning responsibility and upgraded Gaston 
  back to tropical storm status based on observations of 43 kts (10-min 
  avg) from the Chesapeake Light C-man Station (with gusts of 49 kts) and 
  also a ship report of 36-kt sustained winds at 31/0000 UTC about 35 nm 
  southeast of the mouth of the Chesapeake Bay.  These wind reports 
  were from the south and southeastern parts of the circulation with no 
  tropical storm-force winds being recorded over land.

     Accelerating northeastwards, satellite imagery showed limited deep 
  convection in association with Gaston, and there appeared to be a 
  frontal-like band developing to the southeast of the centre.  The MSW 
  had remained at 35 kts and overall, Tropical Storm Gaston was becoming
  less tropical as it began to interact with a frontal zone.  However,
  the storm maintained enough deep convection around its centre to be 
  classified as a tropical cyclone.  Despite Gaston's passage over 16 Deg C
  SSTs, the tropical cyclone appeared to be holding its own for awhile, 
  and satellite intensity estimates at 01/0300 UTC had even risen slightly,
  although buoy reports near the centre indicated that the MSW might 
  have fallen.  However Gaston was kept at tropical storm intensity as a 
  compromise between CI estimates and surface observations.

     By dawn satellite imagery indicated that Gaston had become mostly
  extratropical, and a jet stream was creating a sharp boundary to the
  north of the cloud shield.  NHC upped the MSW to 45 kts on their final
  advisory at 01/0900 UTC to comply with satellite estimates and a 45-kt
  wind speed measurement from Buoy WRYG.  The extratropical system then
  sped east-northeastwards at around 30 kts out into the Atlantic and
  passed just southeast of Newfoundland on the 2nd of September.  Bracknell
  weather charts indicate that Gaston's remnant LOW (995 mb) continued
  northeast across the north Atlantic on the 3rd of September, losing its
  identity northwest of the British Isles the next day.


  C. Meteorological Observations
  ------------------------------

     Tropical Storm Gaston produced some impressive rainfall totals in 
  North and South Carolina.  The following were obtained from the HPC 
  storm summaries.  In South Carolina, Kingstree had a 24-hour total of 
  250 mm, while Turbeville reported 178 mm in 24-hours, both ending at
  0000 UTC 30 August.  Other 24-hour totals above 100 mm were 
  Manning (140 mm) and Charleston (103 mm) during the period 0000 
  UTC 29 August-0000 UTC 30 August.   Darlington measured a 30-hour 
  total of 133 mm ending at 0600 UTC 30 August.  The only other rainfall 
  amount exceeding 100 mm was from Florence, which recorded 119 mm 
  in the 30 hours ending at 0600 UTC 30 August.

     In North Carolina, Lauringburg-Maxton Airport received 125 mm in 
  the 30 hours ending 0600 UTC 30 August while Lumberton recorded 80 mm
  during the same period.  Camp Mackall measured a storm total of 142 mm
  and Pope AFB reported a storm total of 86 mm.


  D. Damage and Casualties
  ------------------------

     News sources indicate that Gaston unleashed torrential rains across 
  North and South Carolina. Also, central parts of Virginia received a 
  foot or more of rainfall which led to serious flooding.  Five people were
  confirmed dead in Richmond, Virginia, as a result of Gaston and 
  damages there were estimated at $20 million.  The overall death toll is 
  believed to be nine.

  (Report written by Kevin Boyle)



                       TROPICAL STORM HERMINE
                               (TC-08)
                           27 - 31 August
               --------------------------------------

     At 2100 UTC on 29 August satellite images indicated that an area of
  low pressure located roughly 280 nm southeast of Cape Hatteras, North 
  Carolina, had developed sufficiently to be classified as a tropical
  storm.  Advisories were initiated on Tropical Storm Hermine at this
  time and the initial MSW was set at 35 kts.   Hermine developed from a 
  disturbance embedded in the same frontal zone that had spawned its twin, 
  Tropical Storm Gaston.   At the time of its inception, the centre of 
  Hermine's circulation was located to the north of the associated deep 
  convection due to the effects of some vertical shear.  However, some 
  additional strengthening was expected.  Basically, Hermine was being 
  controlled by the same HIGH that had gradually pushed Tropical Storm
  Gaston into South Carolina, and the system was forecast to follow suit
  except that the forecast track kept the storm away from the U.S. East 
  Coast.  Therefore, no watches or warnings were deemed necessary.  
  Another reason for this decision was that Hermine was forecast initially
  to be swallowed up by the larger extratropical circulation of Gaston.
 
     The MSW of Hermine was nudged up a little to 40 kts at 0300 UTC on 
  30 August, but the LLCC was still exposed to the north of the deep 
  convection.  The tropical cyclone had turned towards the north-
  northwest at this time, and six hours later, it began to race northwards 
  at 16 kts.  Based on satellite intensity estimates, the intensity was
  upped to 45 kts and was kept at this intensity through 30/1500 UTC based
  on a 30/0947 UTC QuikScat pass which showed 45 to 50-kt winds southeast
  of the LLCC.   Because Hermine didn't appear to be weakening in satellite
  images, precautionary tropical storm warnings were issued at 30/1700 UTC
  for southeastern Massachusetts.  The peak intensity of 45 kts was main-
  tained for the 30/2100 UTC advisory.  Tropical Storm Hermine was now
  expected to remain a separate entity, and its rapid northerly motion
  caused it to overtake the slower-moving Gaston.

     Hermine was basically a low-level cloud swirl devoid of deep
  convection at 0300 UTC on 31 August with its MSW lowered to 35 kts, but
  it was maintained as a tropical storm for this advisory even though Buoy
  44004 reported sustained winds of only 30 kts and sea heights of
  4.6 metres.  This was to allow for a 35-kt wind maximum which could have
  possibly existed close to the tropical cyclone's centre.    Satellite
  intensity estimates also supported 35-kt winds.  Hermine then came ashore
  near New Bedford, Massachusetts, around 31/0600 UTC.  Maximum observed
  surface winds were no more than 20 to 25 kts, and as the system had no
  significant deep convection to speak of, Hermine was considered extra-
  tropical and the last advisory issued on the system at 31/0900 UTC. 

     Hermine brought some heavy rain and wind gusts to near and slightly 
  in excess of tropical storm force early on the 31st, but no sustained
  winds of tropical storm intensity were recorded in the landfall area.  
  Maximum rainfall amounts attributable to Tropical Storm Hermine were
  mostly less than 13 mm, but rainbands from Gaston were close behind,
  making it difficult to make an exact determination of rainfall totals
  between these two systems.  Highest tides reached 0.3 metre above the 
  predicted elevations as the storm passed by.

     There were no reports of damage or casualties as a result of Tropical
  Storm Hermine.

  (Report written by Kevin Boyle)

  *************************************************************************

  NORTHEAST PACIFIC (NEP) - North Pacific Ocean East of Longitude 180

  Activity for August:  1 tropical depression
                        3 tropical storms **
                        1 hurricane

  ** - one of these storms (Howard) became a hurricane and intense
       hurricane in early September


                         Sources of Information
                         ----------------------

     Most of the information presented below was obtained from the
  various tropical cyclone products issued by the Tropical Prediction
  Center/National Hurricane Center (TPC/NHC) in Miami, Florida (or the
  Central Pacific Hurricane Center (CPHC) in Honolulu, Hawaii, for
  locations west of longitude 140W):  discussions, public advisories,
  forecast/advisories, tropical weather outlooks, special tropical
  disturbance statements, etc.  Some additional information may have
  been gleaned from the monthly summaries prepared by the hurricane
  specialists and available on TPC/NHC's website.  All references to
  sustained winds imply a 1-minute averaging period unless otherwise
  noted.


               Northeast Pacific Tropical Activity for August
               ----------------------------------------------

     Tropical cyclone activity was near normal in the Eastern North Pacific
  during August.  August averages for the period 1971-2003 are about 4 NS,
  2 H, and one IH.  August, 2004, produced four NS but only one hurricane;
  however, the final storm, Howard, became a Category 4 hurricane during
  the early days of September.  Estelle formed rather far west from the
  same tropical wave which had earlier spawned Atlantic Hurricane Charley
  and moved into the Central North Pacific.  The other three storms formed
  much nearer the Mexican coast but moved generally northwestward away
  from the mainland with only minimal effects felt along the coastline.

     A special thanks to John Wallace for writing the reports on Georgette
  and Howard, and to Kevin Boyle for writing the report on Frank/TD-09E.

  NOTE: The official TPC/NHC storm reports for most of the Northeast
  Pacific basin tropical cyclones are now available online at the
  following link:

       http://www.nhc.noaa.gov/2004epac.shtml>



                         TROPICAL STORM ESTELLE
                                (TC-07E)
                             19 - 26 August
               ------------------------------------------

  A. Storm Origins
  ----------------

     According to TPC/NHC's Eastern North Pacific monthly summary for
  August, the precursor of Tropical Storm Estelle was the same African
  wave which had spawned Caribbean Hurricane Charley.   The wave was first
  mentioned in the Pacific in a STWO issued at 1700 UTC on 12 August.
  Showers and thunderstorms associated with a broad low-pressure area had
  become a little better organized about 650 nm south of Cabo San Lucas.
  However, upper-level winds were only marginally favorable for further
  development.  The system was then moving west-northwestward at about
  13 kts.  The system was dropped from the outlooks after 1100 UTC on the
  14th as it had become less organized about 700 nm southwest of Cabo San
  Lucas.

     At 0500 UTC on 18 August the system was located in the far western
  part of the Eastern North Pacific region about 1225 nm southwest of Cabo
  San Lucas.  It was moving westward at 13 kts and conditions appeared
  somewhat conducive to further intensification.  As the 18th wore on the
  disturbance gradually exhibited increased organization, and at 0900 UTC
  on 19 August, the first NHC advisory on Tropical Depression 07E was
  issued.  TD-07E was then located over 1400 nm west-southwest of Cabo
  San Lucas on the southern tip of the Baja California Peninsula.


  B. Synoptic History
  -------------------

     The depression was rather slow to intensify.  Deep convection
  decreased both in coverage and organization during the afternoon of the
  19th, and Dvorak classifications were still at T2.0 or lower.  Deep
  convection began to make a comeback during the evening, and the system
  was upgraded to Tropical Storm Estelle at 0900 UTC on 20 August when
  located about 1600 nm west-southwest of Cabo San Lucas.  Dvorak estimates
  had reached 35 kts from all three satellite agencies, and a recent Quik-
  Scat pass had shown some 35-40 kt rain-flagged vectors.  The newly-named
  Estelle was tracking west-northwestward at 12 kts south of a subtropical
  ridge to the north.  During the 20th the cyclone continued to strengthen
  with improved banding features noted in both the northern and southern
  semicircles, and by evening a cold CDO with tops to -82 C had formed.
  The MSW had risen to 55 kts by the time Estelle crossed longitude 140W
  and into Honolulu's AOR around 21/0400 UTC.

     The first CPHC advisory found Estelle located about 865 nm east-
  southeast of Hilo, Hawaii, and moving northwestward at 9 kts.  The MSW
  was upped to 60 kts based on an improved satellite signature, but
  Estelle's intensification phase was about to be halted due to increasing
  southerly shear along the eastern side of an upper-level trough running
  north/south just east of the Big Island.   Satellite CI estimates at
  21/1800 UTC were still 65 kts from SAB and the Honolulu forecast office,
  so the 2100 UTC intensity remained at 60 kts.  However, by 22/0300 UTC
  the LLCC had become exposed to the southeast of the deep convection and
  the intensity was reduced to 55 kts--and the forecaster added that this
  might be quite generous.   The MSW continued to drop rapidly and Estelle
  was reduced to tropical depression status at 23/0000 UTC when located
  about 700 nm east-southeast of Hilo.

     As Estelle weakened, it returned to a more westerly track as the LLCC,
  now decoupled from the mid and upper-level circulation, got caught up in
  the trade wind flow.   Interestingly, after Estelle had been downgraded
  to depression status, the Hawaiian trough with its associated shear began
  to move west ahead of the tropical cyclone, allowing a mid and upper-
  level ridge to build to the west and north of Estelle.  A Hurricane
  Hunter aircraft flew through the system on the morning of 23 August and
  reported a FLW of 35 kts at 1500 m about 55 nm north of the LLCC.  Based
  on this the MSW was retained at 30 kts.   During this time some of the
  models were hinting at the possibility of some modest re-intensification
  due to the slightly improved environment, but Estelle continued to slowly
  weaken as it moved west-southwestward with the trade winds.

     The MSW was reduced to 25 kts at 24/0000 UTC, but CPHC continued
  writing advisories for another 48 hours.   During the morning of the 25th
  convection increased within the outer portions of the circulation, but
  no concentration of convection near the LLCC was observed.  This renewed
  thunderstorm activity was concentrated mainly around the western to
  southwestern side some 90 nm from the circulation center with scattered
  convection seen within about 130 nm to the east of the center.  The
  convection was likely in response to instability caused by a weak trough
  extending southward from Hawaii, and also possibly to the slightly warmer
  SSTs the depression was moving over.   However, by the late afternoon of 
  the 25th the convection was waning and there appeared to be very little
  evidence of westerly winds to the south of the center.  Hence, the final
  advisory on Tropical Depression Estelle was issued at 26/0300 UTC with
  the weak 20-kt center located about 300 nm south of the southern tip of
  the Big Island of Hawaii.


  C. Damage and Casualties
  ------------------------

     No damage or casualties are known to have resulted from Tropical
  Storm Estelle.

  (Report written by Gary Padgett)



                            HURRICANE FRANK
                                (TC-08E)
                             23 - 26 August
                            ---------------

                                  AND

                          Tropical Depression
                                (TC-09E)
                             23 - 26 August
                ---------------------------------------

     Hurricane Frank originated from the remnants of Tropical Storm Earl, 
  which had degenerated into a tropical wave in the Caribbean and 
  subsequently crossed Central America before emerging into the Eastern 
  North Pacific basin around 18/19 August.  The ex-Earl wave was monitored
  on NHC/TPC's Tropical Weather Outlooks as the system tracked south of
  Mexico until finally satellite images began to show a persistent CDO on
  the morning of the 23rd.  This feature and the likelihood of further
  development prompted NHC to issue the first advisory package on Tropical
  Depression 08E at 0900 UTC on 23 August.  The depression was then centred
  approximately 375 nm south of Cabo San Lucas on the southern tip of
  the Baja California Peninsula.  Moving northwestward at 10 kts, TD-08E
  quickly organized to become Tropical Storm Frank at 23/1500 UTC.
  A 23/0927 UTC AMSU overpass indicated a tight inner core of convection
  with a long convective band that had wrapped more than halfway around
  the centre.

     Also of interest, a very small disturbance was developing about
  700 nm west of Frank and this was to become Tropical Depression 09E.
  But before the first advisory would be issued on this system, Tropical
  Storm Frank exploded into rapid strengthening mode which took NHC a
  little by surprise.  The MSW was raised to 55 kts based on the
  appearance of a banding eye feature in both visible and infrared
  satellite imagery.  A well-defined circular eye feature was also visible
  on 23/0927 UTC AQUA and 23/1459 UTC SSM/I overpasses.  Shortly after the
  23/1500 UTC advisory, NHC issued a special advisory denoting this sudden
  intensification and accordingly updated the forecast positions and 
  maximum winds.  At 1645 UTC (the issuance time of the special advisory)
  Frank was located approximately 335 nm south-southwest of Cabo San
  Lucas.  Frank reached hurricane intensity at 23/2100 UTC, the same time
  that the small disturbance was promoted to Tropical Depression 09E
  with 25-kt winds.  TD-09E was located about 850 nm west-southwest of
  the southern tip of the Baja and about 640 nm west of Hurricane Frank.

     At 2100 UTC on 23 August Tropical Depression 09E was moving slowly 
  eastward at 2 kts.  This movement was to be the only suggestion that 
  Hurricane Frank might be influencing TD-09E's track, and there was 
  no further interaction observed between the two storms during their
  lifetimes.  TD-09E was located over marginally warm water and near the
  eastern edge of dry air at mid-to upper levels--not the best ingredients
  for tropical cyclone development--and the forecast for this system was
  for no more than tropical storm intensity.  Also, the close proximity of
  Frank was expected to limit intensification.  Infrared images and micro-
  wave data showed a poorly-organized cyclone with the LLCC located to the
  south of the deep convection, and this appearance was maintained through
  the 24th.  The only event worth noting that day was that the MSW was
  raised to 30 kts after satellite estimates reached T2.0.  Also, TD-09E
  had turned onto a westerly heading which was to take it into an
  increasingly hostile environment and eventual oblivion.   Meanwhile,
  Hurricane Frank, which was on a northwesterly heading, had continued its
  strengthening phase and reached a peak intensity of 75 kts at 24/1500
  UTC when located about 300 nm southwest of Cabo San Lucas.

     Hurricane Frank crossed the 26 Deg C isotherm late on the 24th and as 
  it did so, began to weaken.  The intensity was brought down to 65 kts at 
  0300 UTC on 25 August, and as SSTs were becoming progressively colder
  ahead of the storm, both Frank, as well as its neighbour, Tropical
  Depression 09E, were forecast to be dissipating in a couple of days.
  In fact, Frank blew itself out just as quickly as it had rapidly
  strengthened into a hurricane.   By 25/0900 UTC the central convection
  had diminished considerably, leaving some weak banding features in the 
  southeastern quadrant.  Frank was downgraded to a tropical storm with 
  50-kt winds at this time, and this intensity was maintained for another 
  six hours, after which it had become a virtually convection-free vortex. 
  Frank was downgraded to a tropical depression at 26/0300 UTC when
  centred approximately 465 nm west of Cabo San Lucas.  TD-09E was faring
  no better.  The little remaining deep convection had been stripped away
  from this weak system by the moderate south-southwesterly shearing
  conditions.   Both Frank and TD-09E continued for a little while longer
  as westward-moving low-level swirls of clouds until NHC signed both their
  death warrants at 26/1500 UTC.   Frank's final position was approximately
  550 nm west of Cabo San Lucas, and TD-09E's final location was roughly
  1200 nm west-southwest of the same point.

     There were no damages or casualties associated with either of these 
  systems.

  (Report written by Kevin Boyle)



                        TROPICAL STORM GEORGETTE
                                (TC-10E)
                             26 - 30 August
              --------------------------------------------

  A. Storm Origins
  ----------------

     A cyclonic disturbance off the southwest coast of Mexico steadily
  organized late on 24 August and into the following day, quick on the
  heels of Hurricane Frank.  The disturbance tracked slowly west-
  northwestward, paralleling the coast, and satellite data, in conjunction 
  with a ship report, warranted its upgrade to Tropical Depression Ten-E
  at 1500 UTC on 16 August when located approximately 315 nm south-
  southwest of Manzanillo, Mexico.


  B. Synoptic History
  -------------------

     The depression continued on the same west-northwesterly track it had
  followed as a disturbance, and quickly intensified to become Tropical
  Storm Georgette on the second advisory, at 2100 UTC on 26 August.
  Remarkable visible GOES-10 imagery from this time showed Georgette
  accompanying no less than three tropical cyclone remnants in the North-
  east Pacific basin: ex-Hurricane Frank and ex-Tropical Depression Nine-E
  to its northwest, and the fizzling vortex that was once Tropical Storm
  Estelle southwest of the Big Island of Hawaii.

     Georgette strengthened steadily, though less than it would have 
  without moderate easterly shear.  It reached its first peak 1-min
  MSW of 50 kts at 0900 UTC on 27 August, as it made a slow turn to a
  more westerly track south of an expanding ridge.  Georgette was then
  centered about 350 nm south of Cabo San Lucas on the southern tip of
  the Baja California Peninsula.  Shear continued to erode Georgette's
  convection, causing it to weaken slightly on the 28th while it settled
  on a due westward course.  However, convection made a comeback early on
  the 29th, and the storm reached its second peak of 50 kts at 0900 UTC
  that day with the lowest CP estimated at 997 mb.

     Georgette weakened slightly again after its peak, but maintained a
  more or less steady state as it traveled parallel to the 26 C isotherm
  until the next day, when cool SSTs and stable air finally made themselves
  felt in addition to the persistent shear.  Georgette weakened rapidly to
  a depression on the 30th, and the final advisory was issued at 1500 UTC
  that day, placing the weakening center about 825 nm west-southwest of
  Cabo San Lucas.  Its large remnant vortex could be tracked in satellite
  imagery for some time; as late as September 4th it underwent a photogenic
  interaction with another LOW well to its north.     By the 6th, though, 
  Georgette's remnants had fully dissipated.


  C. Damage and Casualties
  ------------------------

     There are no known casualties or damage associated with Tropical
  Storm Georgette.

  (Report written by John Wallace)



                           HURRICANE HOWARD
                               (TC-11E)
                        30 August - 5 September
              -------------------------------------------

  A. Storm Origins
  ----------------

     The disturbance that became Hurricane Howard developed quickly off
  the southwestern Mexican coast late on 29 August and tracked west-
  northwestward in the typical fashion of NEP storms.  Its organization
  improved over the course of a day, warranting its upgrade to Tropical
  Depression Eleven-E at 2100 UTC on 30 August when located approximately
  400 nm south-southeast of Manzanillo, Mexico.  The new depression
  subsequently continued on its initial west-northwesterly heading.


  B. Synoptic History
  -------------------

     The depression steadily strengthened, and became Tropical Storm Howard
  at 0900 UTC on 31 August.    Synoptic conditions and SSTs were quite 
  favorable, and Howard intensified steadily.   Interestingly, the SHIPS
  model forecast an 80% chance of rapid intensification as early as the
  1st of September when it was only a 60-kt tropical storm.    The NHC
  followed this guidance in its forecasts, though there was little inkling
  of it from Howard itself.  As Howard reached hurricane status at 0900 UTC
  on the 1st, SHIPS backed off on the intensity forecast, as did the NHC.  

     Howard strengthened slowly as a Category 1 hurricane for almost a full
  day before dramatically exploding in intensity early on the 2nd, when it
  shot up from Category 1 to Category 4 in only 12 hours.  Howard's central
  pressure plunged an estimated 37 mb during that time frame, for an
  amazing average drop of over 3 mb per hour, firmly within the range for
  explosive deepening (1).  The rapid strengthening quickly took Howard to
  its peak MSW of 120 kts, with a CP of 940 mb, at 02/1500 UTC when the
  cyclone was located about 350 nm south-southwest of Cabo San Lucas.
  Satellite imagery at this time revealed a classic pinhole eye embedded in
  a cold, symmetric CDO.

     After 12 hours at peak intensity and a bend more to the northwest,
  Howard weakened slowly due to the influence of cooler SSTs along its
  track.  Nevertheless, Howard held on to to major hurricane status for
  roughly 36 hours as SSTs remained more favorable than anticipated.  Rapid
  weakening commenced late on the 3rd as it finally crossed into hostile
  waters and entrained stable air, and on the 4th, increasing shear
  exacerbated its decline.  Howard turned steadily more north-northwestward
  as a trough eroded the ridge to its north.    

     On the 4th Howard weakened to a tropical storm, and the next day to a
  depression.  The final advisory on Tropical Depression Howard was issued
  at 1500 UTC on 5 September when the weakening LOW was located roughly
  500 nm west-northwest of Cabo San Lucas.   Howard's low-level remnant
  vortex was unusually persistent--it lingered off the northern coast of
  Baja California for three days before turning southwestward ahead of
  Tropical Storm Isis and losing its identity late on 10 September.


  C. Damage and Casualties
  ------------------------

     No casualties or damages are known to have been caused by Hurricane
  Howard.


  D. References
  -------------

  (1.)  http://www.nhc.noaa.gov/aboutgloss.shtml>

     It's worth noting that Howard's intensification was somewhat delayed; 
  many, if not most TCs that rapidly or explosively deepen do so soon after
  reaching hurricane strength.

  (2.) http://www.nhc.noaa.gov/archive/2004/HOWARD.shtml?>

  (Report written by John Wallace)

  *************************************************************************

  NORTHWEST PACIFIC (NWP) - North Pacific Ocean West of Longitude 180

  Activity for August:  4 tropical depressions **
                        3 tropical storms ++
                        4 typhoons
                        2 super typhoons
                                   
  ** - all of these were classified as tropical depressions by JMA only

  ++ - one of these (Malou/15W) was classified as a tropical storm by
       several of the Asian TCWCs, but not by JTWC, while another (21W)
       was classified as a tropical storm by JTWC only


                        Sources of Information
                        ----------------------

     Most of the information presented below is based upon tropical
  cyclone warnings and significant tropical weather outlooks issued
  by the Joint Typhoon Warning Center of the U. S. Air Force and
  Navy (JTWC), located at Pearl Harbor, Hawaii.   In the companion
  tropical cyclone tracks file, I normally annotate track coordinates
  from some of the various Asian warning centers when their center
  positions differ from JTWC's by usually 40-50 nm or more.   All
  references to sustained winds imply a 1-minute averaging period
  unless otherwise noted.

     Michael V. Padua of Naga City in the Philippines, owner of the
  Typhoon 2000 website, normally sends me cyclone tracks based upon
  warnings issued by the Japanese Meteorological Agency (JMA) and the
  Philippines' Atmospheric, Geophysical & Astronomical Services
  Administration (PAGASA).  Also, Huang Chunliang of Fuzhou City, China,
  sends data taken from synoptic observations around the Northwest
  Pacific basin.  A very special thanks to Michael and Chunliang for
  the assistance they so reliably provide.

     In the title line for each storm I have referenced all the cyclone
  names/numbers I have available:   JTWC's depression number, the 
  JMA-assigned name (if any), JMA's tropical storm numeric designator,
  and PAGASA's name for systems forming in or passing through their
  area of warning responsibility.


               Northwest Pacific Tropical Activity for August
               ----------------------------------------------

     After a rather quiet July, which had come on the heels of a very
  active June, the tropics of the Western North Pacific exploded to
  produce a very active August.  I do not yet have monthly statistics
  available for the NWP basin, but if August, 2004, was not the most
  active August on record there, it certainly must rank in the Top Five.
  Nine systems reached tropical storm intensity, taking into account the
  classifications from all the warning agencies.  Six of these reached
  typhoon intensity and two became super typhoons per JTWC's analysis.

     All but one of the six typhoons struck populated areas in the NWP
  basin:

     Meranti - remained at sea
     Rananim - China
     Megi    - South Korea
     Chaba   - Marianas, Japan
     Aere    - Taiwan, China
     Songda  - Marianas, Okinawa, Japan

     Four systems were mentioned as tropical depressions in the High Seas
  bulletins issued by JMA.  The first of these was carried as a 30-kt
  depression on 6 and 7 August.  At 06/0600 UTC it was located roughly
  600 nm east-southeast of Iwo Jima.  The system moved generally in a
  northeasterly direction and had weakened by 1200 UTC on the 7th, being
  then located about 600 nm west-northwest of Wake Island.  JTWC did
  carry this system in their STWOs, and at one point assigned a fair
  potential for development.  A track, based on JMA's bulletins, was
  included for this system in the companion August cyclone tracks file.

     The remaining three JMA depressions were listed only in the Summary
  portion of the High Seas bulletins.  One weak circulation was located
  in the vicinity of 20N/141E on 8 August and remained quasi-stationary.
  Another depression, near 18N/133E, was mentioned only once, at 13/1200
  UTC.  The final of the three occurred on 27 and 28 August, being quasi-
  stationary just west of the International Dateline near 10N/178E.  JTWC
  did give this system a fair development potential at one point.  No
  tracks were included for any of these weaker JMA depressions in the
  August tracks file.

     A special thanks to Kevin Boyle for writing several of the NWP
  cyclone reports, and to Huang Chunliang for sending much information
  regarding meteorological observations as well as damage and casualty
  figures.



                            TYPHOON MERANTI
                           (TC-14W / TY 0412)
                              3 - 9 August
                --------------------------------------

  Meranti:  contributed by Malaysia, is a type of tree, tall and big
            which yields soft wood and is often used as a building
            material

  A. Storm Origins
  ----------------

     The first NWP tropical cyclone of a very active August had its
  beginnings with an area of convection which formed on the 2nd and was
  located approximately 80 nm southwest of Wake Island at 0600 UTC.
  Animated multi-spectral imagery revealed a weak LLCC with little deep
  convection while a 200-mb analysis indicated moderate diffluence over the
  area and weak to moderate vertical shear.   An interim STWO was issued
  at 2100 UTC relocating the disturbance 2 degrees to the south to a point
  about 225 nm south-southwest of Wake Island.    Deep convection was
  increasing and the potential for further development was upgraded to
  fair.  JTWC issued a TCFA at 03/0230 UTC, placing the center 195 nm
  south-southwest of Wake Island.  Deep convection was increasing over the
  LLCC, which was located in the divergent quadrant of a TUTT cell to the
  northwest, and an upper-level analysis indicated moderate diffluence in
  the equatorward direction and continued weak to moderate vertical shear.


  B. Synoptic History
  -------------------

     The system continued moving slowly northward and the first JTWC
  warning on Tropical Depression 14W was issued at 1800 UTC with 30-kt
  winds.  The system was then passing only about 25-35 nm west of Wake
  Island and was being steered poleward along the western periphery of
  a near-equatorial mid-level ridge located to the east of the depression.
  Also at 1800 UTC, JMA upgraded the system to a 30-kt depression.   JTWC
  upgraded TD-14W to tropical storm status on the second warning, issued
  at 04/0000 UTC, when the cyclone was centered approximately 55 nm north-
  west of Wake Island.   Initially moving slowly north-northwestward, by
  0600 UTC the tropical storm was scooting northward at 19 kts.  This
  rapid motion continued until 1800 UTC, when the depression had returned
  to a slower northwestward motion at 9 kts.  The cyclone acquired the name
  Meranti at 04/1200 UTC when JMA upgraded it to tropical storm status.
  Meranti remained a minimal tropical storm on the 4th, but by 0000 UTC on
  the 5th the system was beginning to intensify--both JMA and JTWC were
  estimating the MSW at 45 kts.  The storm was then centered approximately
  900 nm west of Midway Island, and had begun tracking north-northeastward.

     Tropical Storm Meranti continued to quickly intensify, and at 05/1200
  UTC was upgraded to typhoon status by both JMA and JTWC, based on CI
  estimates of 65 and 77 kts.  Meranti was by then tracking northeastward
  at 11 kts, and a basic easterly motion was forecast to continue as the
  near-equatorial ridge built to the south, allowing the cyclone to track
  along the northern periphery of the ridge.  Typhoon Meranti reached its
  peak intensity of 90 kts (75 kts 10-min avg per JMA) at 05/1800 UTC when
  located about 800 nm west of Midway Island.   Satellite CI intensity
  estimates were 90 kts, and a 08/1454 UTC TRMM image depicted a well-
  organized system with a small, ragged eye.  Meranti's intensification
  had been aided by an increase in the poleward outflow channel.  The
  minimum CP estimated by JMA was 960 mb, and at its peak, gales covered
  an area approximately 170 nm in diameter.

     Typhoon Meranti maintained its peak intensity for 18 hours.  By 0600
  UTC on 6 August satellite imagery revealed that the eye was no longer
  distinct, and microwave and water vapor imagery indicated that dry air
  was being entrained into the system from the west.  The MSW was reduced
  to 80 kts at 1200 UTC, based on CI estimates ranging from 55 to 90 kts,
  and at 1800 UTC the MSW was drastically reduced to 50 kts with satellite
  CI estimates then ranging from 45 to 65 kts.   The cyclone had shown a
  rapid decrease in deep convection over the past few hours with the LLCC
  becoming exposed to the southwest of the deepest convection.  (JMA also
  downgraded Meranti to a tropical storm at 1800 UTC, estimating the 10-min
  avg MSW at 60 kts.)  The cyclone was then located roughly 670 nm west-
  northwest of Midway Island, moving northeastward at 14 kts.

     Meranti continued tracking northeastward along the periphery of a mid-
  level steering ridge as it retreated eastward.  The storm underwent a
  slight re-intensification to 60 kts at 07/0000 UTC due to enhanced out-
  flow into an upper-level LOW located to the southeast.   But after that
  it was downhill all the way.  The MSW dropped to 55 kts at 07/0600 UTC
  and to 45 kts at 1800 UTC.   The system was experiencing shear with a
  partially-exposed LLCC southwest of the deep convection, and there was
  also evidence that extratropical transition was beginning.   At 1800 UTC
  the center of Meranti was located approximately 700 nm northwest of
  Midway Island, and had turned to the north due to the approach of a major
  shortwave trough.  JTWC issued the final warning on Meranti at 0600 UTC
  on 8 August, placing the center about 800 nm northwest of Midway Island.
  The MSW was estimated at 40 kts, and the system was deemed to have
  completed extratropical transition.  Ever since the downgrade of Meranti
  from typhoon status on the 6th, JMA's equivalent 1-min avg MSW had been
  running higher than JTWC's, and JMA maintained Meranti as a tropical
  cyclone for another 24 hours following JTWC's final warning, declaring
  the system extratropical at 0600 UTC on 9 August.  However, Meranti's
  remnants quickly weakened--by 1800 UTC consisting of only a 25-kt LOW
  about 930 nm northwest of Midway Island, or several hundred miles south
  of the southwestern Aleutian Islands.


  C. Damage and Casualties
  ------------------------

     No damage or casualties are known to have resulted from Typhoon
  Meranti.

  (Report written by Gary Padgett)



                          TROPICAL STORM MALOU
                           (TC-15W / TS 0411)
                              4 - 6 August
                ----------------------------------------

  Malou:  contributed by Macau, China, is the name for 'agate'--a very
          hard stone with bands of colour, and is often used for 
          ornaments and souvenirs

  A. Storm History
  ----------------

     At 0600 UTC on 31 July, JTWC released a STWO on a limited area of 
  convection which had persisted for 12 hours approximately 190 nm north
  of Guam.  The potential for development was assessed as poor.  Multi-
  spectral imagery revealed a large, well-defined LLCC in association with
  the convection.   A day passed with little to write about, but at 02/0600
  UTC JTWC upgraded the development potential to fair based on improved
  organization.  However, multi-spectral imagery depicted a partially-
  exposed system with cycling deep convection to the east of the LLCC.
  The 03/0300 UTC STWO demoted the suspect area back to poor status as a
  result of the convection decreasing markedly over the poorly-defined
  LLCC.  Upper-level analysis of UW/CIMSS satellite-derived winds showed
  that the LLCC had entered a region of confluent flow from the southeast
  and increased vertical shear.  The shear later eased and convection
  increased over the centre, prompting JTWC to upgrade the development
  potential back up to fair.  At 03/2330 UTC a TCFA was issued and the
  first warning on Tropical Depression 15W followed at 04/0000 UTC.

     The first warning placed the newly-formed depression approximately 
  420 nm southeast of Iwakuni, Japan, and moving northwestward at 21 kts.
  This rapid movement, southwest of the low-to mid-level ridge, was
  expected to take the system inland over Japan within 24 hours of 
  inception.  As a result Tropical Depression 15W was not expected to 
  intensify further.  Also, the environment was becoming less supportive 
  of tropical cyclone maintenance.  By 04/1200 UTC Tropical Depression 
  15w was only 70 nm south-southwest of Kyoto, Japan.  The system had 
  been named by JMA, who considered TD-15W a tropical storm and 
  accordingly assigned the international codename Malou.  The peak 10-min
  MSW reached 40 kts per JMA warnings at 04/0600 UTC and this intensity
  was carried through the next bulletin at 04/1200 UTC.  The agency then
  downgraded Malou to depression status at 05/0000 UTC.

     Tropical Storm Malou made landfall over Japan between 1800 UTC on 
  4 August and 05/0000 UTC.  By 0000 UTC 5 August it was located over the
  Sea of Japan and, having completed recurvature, was moving north-
  northeastward at 21 kts.  The system had lost most of its organization 
  during its transit across Japan, and JTWC called it a day at 05/0600 UTC 
  on the now extratropical Malou which was then centred 275 nm northwest
  of Tokyo, Japan.   The remnants of the former tropical cyclone then
  tracked across northern Honshu on 6 August and into the North Pacific
  before losing its identity late on the 7th.

  Editor's Note:  The recording of sustained winds at or exceeding gale
  force by two stations at sea level strongly suggests that Malou was
  indeed a tropical storm at landfall in Japan.  See Section B below.
 
     No reports of fatalities or significant damages were received.


  B. Meteorological Observations
  ------------------------------

     The following observations from Japan were compiled and sent by
  Huang Chunliang.   A special thanks to Chunliang for the information.


  (1) Landfalls (based on the JMA warnings)
  =========================================

  1. Typhoon 0411 (MALOU) made first landfall near Anan City, Tokushima 
  Prefecture around 04/1330 UTC with a MSW of 40 kts and a CP of 994 hPa.

  2. Typhoon 0411 (MALOU) made second landfall near Aioi City, Hyogo 
  Prefecture around 04/1600 UTC with a MSW of 40 kts and a CP of 996 hPa.


  (2) Rains
  =========

  1. The most torrential hourly rain during the storm was recorded in 
  Miyagawa, Mie, which reported a record-breaking amount of 101 mm during
  the 1-hr period ending at 05/0110 UTC.

  2. The most significant storm totals [03/1500-05/0700 UTC] included: 
  Kamikitayama, Nara (731 mm); Miyagawa, Mie (589 mm); Mt. Hidegatake,
  Nara (571 mm); etc.


  (3) Winds
  =========

  1. Five stations reported sustained winds of gale force or higher during
  the storm (values represent 10-min means):

    Tomogashima, Wakayama (JMA65036, Alt 43m)----40.8 kts
    Murotomisaki, Kochi (WMO47899, Alt 185m)----39.5 kts
    Akashi, Hyogo (JMA63496, Alt 3m)----36.9 kts
    Himezi, Hyogo (WMO47769, Alt 38m)----35.6 kts
    Tsu, Mie (WMO47651, Alt 3m)----33.8 kts

  2. The highest peak gust occurred in Himezi, Hyogo (WMO47769, Alt 38m),
  which reported 58.9 kts.


  (4) References (Japanese versions only)
  =======================================

     http://www.data.kishou.go.jp>
     http://www.osaka-jma.go.jp>

  (Report written by Kevin Boyle with significant contributions by
  Huang Chunliang)



                           TYPHOON RANANIM
                      (TC-16W / TY 0413 / KAREN)
                            6 - 13 August
            ----------------------------------------------

  Rananim:  contributed by the Federated States of Micronesia, is the
            Chuukese word for 'hello'

  A. Introduction
  ---------------

     Typhoon Rananim was the first of two tropical cyclones to strike
  China during the month of August and became the worst storm to affect
  that nation since Typhoon Winnie (1997).


  B. Storm Origins
  ----------------

     At 0600 UTC on 5 August JTWC issued a STWO for a persistent area of 
  deep convection located approximately 380 nm west-northwest of Guam. 
  Multi-spectral imagery and a QuikScat pass revealed a possible weak LLCC
  with isolated, disorganized convection.  Upper-level analysis indicated
  that the suspect area was located in a region of weak to moderate wind
  shear and moderate diffluence.  The potential for development of a
  significant tropical cyclone within the next 24 hours was poor.  The LLCC
  consolidated on the 6th and the deep convection became more persistent
  and better organized.  Based on this, JTWC upgraded the potential to
  fair.  This was superceded by a TCFA at 06/2100 UTC.  At this time the
  disturbance was relocated to a position 560 nm east of Manila in the
  Philippines.  A second relocation was required six hours later, placing
  the centre nearly a degree south near 14.6N/130.7E.  Warnings began on
  Tropical Depression 16W at 07/1800 UTC with the system moving northward
  at 10 kts approximately 590 nm south of Okinawa, Japan. 


  C. Synoptic History
  ------------------- 

     At 0000 UTC 8 August, the poorly-defined centre of Tropical Depression
  16W was continuing northward along the northwestern periphery of the 
  near-equatorial ridge.  A TUTT was located between that HIGH and the 
  subtropical ridge, and the prognostic reasoning issued at the time
  indicated that after 24 hours the TUTT would move east, allowing the two
  ridges to combine.  This synoptic situation would result in a north to
  northeast heading for the depression.  QuikScat imagery indicated only
  weak winds associated with the LLCC with the strongest winds to the
  southwest.  At 08/0600 UTC multi-spectral imagery indicated cycling
  convection over the broad centre and also several weak low-level vortices
  rotating around the main area of convection.    The name Rananim was
  assigned when JMA upgraded the system to a tropical storm at 08/1200 UTC.
  JTWC also promoted TD-16W to tropical storm intensity at this time,
  setting the MSW at 35 kts.  Turning northeastward at 08/1800 UTC Rananim
  strengthened a little to 40 kts with a recent TRMM pass showing a sheared
  LLCC on the eastern edge of the deep convection.

     Tropical Storm Rananim reverted back to its northward motion at 0000
  UTC on 9 August, moving to a position 440 nm south-southeast of Okinawa, 
  Japan.  At this time both QuikScat and microwave imagery showed a well-
  defined LLCC which still lay on the eastern side of the deep convection.
  The MSW increased during the day from 45 kts at 09/0000 UTC to 55 kts at
  09/1800 UTC.  The forward motion slowed for a time as Rananim turned 
  west-northwestward before picking up again to around 7-8 kts.  As the
  10th dawned, Rananim was close to typhoon strength.  The storm had turned
  towards the northwest and was located 290 nm south of Okinawa.   Deep 
  convection had increased by this time in the northern semicircle while
  the poleward outflow pattern had also improved. The MSW reached 65 kts
  at 10/1200 UTC as the newly-upgraded typhoon continued northwestward 
  along the western periphery of the mid-level steering ridge to the
  northeast.

     By 0000 UTC on 11 August Typhoon Rananim had closed to within 195 nm 
  south-southwest of Okinawa, its strengthening phase bringing the MSW up 
  to 75 kts.  Intensification continued, and satellite imagery showed the 
  development of a small, ragged eye at 11/0600 UTC.  After the intensity
  had reached 85 kts Rananim began to weaken slightly as the outflow became
  more restricted, but intensification soon resumed again and the storm 
  reached its peak intensity of 90 kts at 12/0000 UTC as it began to
  approach the Chinese coast.

     At 0000 UTC on 12 August Rananim was moving north-northwestward 
  approximately 150 nm northeast of Taipei, Taiwan, at its peak intensity
  of 90 kts.  This was maintained as the storm passed north of the island,
  its track altering to a west-northwesterly heading.     This track
  ultimately carried the cyclone inland at approximately 12/1300 UTC near
  Wenling, Zhejiang Province, China.     As Typhoon Rananim crossed the
  Chinese coast it began to weaken.   By the time the 1800 UTC warning was
  issued the MSW had fallen to 70 kts.   Further weakening occurred as the
  storm progressed farther west into China.   JTWC downgraded Rananim to a
  60-kt tropical storm in their final warning at 13/0000 UTC, and JMA
  issued their final warning on the system as a 30-kt tropical depression
  at 13/0600 UTC.  (See Section E below for more information on the inland
  progress of the weakening depression.)

  Editor's Note:   JMA's peak 10-min avg MSW for Rananim was 80 kts, and 
  this was also the peak intensity estimated by the Central Weather Bureau
  of Taiwan.   However, HKO assigned a peak MSW of 85 kts, and NMCC 
  estimated Rananim's peak 10-min avg MSW at 90 kts--equivalent to a 1-min
  avg MSW of 105 kts.  Thus, NMCC and HKO considered Rananim a more intense
  typhoon than did JTWC.   PAGASA named this system Karen, and estimated
  the peak intensity of Typhoon Karen at 75 kts during the time it was
  within that agency's AOR.    The cyclone had exited PAGASA's AOR by
  11/1800 UTC, shortly before reaching its overall peak intensity.  The 
  minimum CP estimated by JMA was 950 mb.  At Rananim's peak intensity, 
  gales covered a zone 220 nm in diameter.   (Rather nostalgic having a 
  typhoon named Karen in the NWP again, even if only an unofficial name.
  Typhoon Karen of November, 1962, was one of the fiercest and most famous
  typhoons to strike the island of Guam in modern times.)


  D. Meteorological Observations from Japan
  -----------------------------------------

     The observations in this section were compiled and sent by Huang
  Chunliang from data obtained at the following URL:

     http://www.okinawa-jma.go.jp>


  (1) Wind and Pressure Observations (all dates 11 September)
  ===========================================================

  Station         Min SLP (hPa)     Peak Sust Wind (kts)    Peak Gust (kts)
  -------------------------------------------------------------------------
  Miyakojima      952.9 (1236 UTC)    50.9 (1050 UTC)       94.9 (1110 UTC)
  Ishigakijima    975.3 (1410 UTC)    51.9 (2030 UTC)       82.0 (0721 UTC)
  Iriomotejima    978.7 (1656 UTC)    46.1 (2030 UTC)       70.2 (2022 UTC)
  Yonagunijima    983.5 (1618 UTC)    39.8 (1630 UTC)       68.6 (1618 UTC)


  (2) Rainfall Measurements
  =========================

  Station        WMO Number  Alt (m)  Storm Total (mm)     Period (UTC)
  ------------------------------------------------------------------------
  Miyakojima       47927       40        229.0 (*)       09/1500 - 12/1500
  Ishigakijima     47918        6        160.0 (#)       09/1500 - 12/1500
  Iriomotejima     47917        9        146.5           09/1500 - 12/1500
  Yonagunijima     47912       30        165.5           09/1500 - 12/1500

  Notes:

  (*) - 188.0 mm out of the total fell during the 24-hour period ending at
        12/0000 UTC

  (#) - 138.5 mm out of the total fell during the 24-hour period ending at
        12/0000 UTC


  E. Meteorological Observations from China
  -----------------------------------------

  (1) Landfall Observations
  =========================

     According to the NMCC warnings, Typhoon 0414 (Rananim) made landfall
  in Shitang Town, Wenling City (a sub-city of Taizhou City), Zhejiang
  Province around 12/1200 UTC with a MSW of 87 kts and a CP of 950 hPa.
  Interestingly, the town of Shitang was exactly the place which saw the
  first sunbeam of the new century in Mainland China on January 1, 2000.

     After rampaging through southern Taizhou City, northern Wenzhou City,
  northern Lishui City and Quzhou City, Rananim eventually departed from
  Zhejiang Province and entered Jiangxi Province as a tropical storm around
  13/0300 UTC.  The storm didn't finish its 22-hour journey in northern
  Jiangxi Province until 14/0100 UTC, by which time it had weakened to a
  depression. 

     Tropical Depression Rananim then moved through southeastern Hubei
  Province and northeastern Hunan Province, where it finally dissipated as
  a significant tropical cyclone.


  (2) Rainfall Observations
  =========================

  (a) Zhejiang Province
  +++++++++++++++++++++

     During the 72-hour period ending at 14/0000 UTC, rains > 200 mm were
  recorded at 35 weather stations, among which 14 stations reported rains
  > 300 mm with Shabu (located in Huangyan District, Taizhou City), Zhuxi
  (located in Xianju County, Taizhou City) and Wugen (located in Wenling
  City--a sub-city of Taizhou City) ranking the top three and reporting
  454 mm, 453 mm and 436 mm, respectively.   However, it should be noted
  that it was a hydrological station that reported the most torrential rain
  during the typhoon--Futou, Yueqing City (a sub-city of Wenzhou City)
  recorded a 24-hour accumulation of 703.5 mm, which turned out to be a
  new record for Zhejiang Province's daily rainfall amounts. 


  (b) Jiangxi Province
  ++++++++++++++++++++

     During the 60-hour period ending at 15/0000 UTC, rains > 100 mm were
  recorded in 39 counties/cities, among which Lushan, Yongxiu, Xinjian,
  Jinxian, Dongxiang, Fengcheng and Yujiang reported rains > 200 mm with
  Lushan reporting the highest amount of 268 mm (200 mm out of this total
  fell during the 24-hour period ending at 14/0000 UTC).


  (c) Hunan Province
  ++++++++++++++++++

     Rains > 100 mm were recorded at 9 stations during the 24-hour period
  ending at 15/0000 UTC with Qiyang County reporting the highest amount of
  120.6 mm.


  (d) Taiwan Region
  +++++++++++++++++

     Neuchusan, Hsinchu County, reported the highest storm total of 345 mm
  during the 35-hour period ending at 12/0300 UTC.


  (e) Other Provinces
  +++++++++++++++++++

     Some of the stations located in Fujian, Hubei, Henan, Anhui and
  Jiangsu also reported torrential rains during the typhoon.


  (3) Wind Observations
  =====================

     Typhoon-force winds were reported in the coastal region of Zhejiang
  Province with gusts exceeding 78 kts being recorded by 11 stations, among
  which Dachen Dao (WMO 58666, 28.45E/121.88, Alt 84 m) reported the
  highest gusts, topping 114.1 kts--a new record for the station--at
  12/0721 UTC.  Other significant gust observations include:  Shamen, 
  Wenling City--104.0 kts, and Sanjiaotang, Sanmen County--90.2 kts.

     Most areas of Shanghai Municipality reported peak gusts of Beaufort
  Force 7 to 9 during the storm.


  F. Damage and Casualties
  ------------------------

     Over 18,000,000 residents in Mainland China were affected by the
  typhoon.  Official statistics indicated that Typhoon Rananim had killed
  168 people and caused 20.1 billion yuan of direct economic losses in the
  mainland.

     In Zhejiang Province Typhoon Rananim killed at least 164 people
  (falling houses killed 109; landslides claimed 28 lives; another 27 died
  in flooding or other accidents caused by the storm) with 24 still
  missing.  Also, more than 1800 people were injured in the typhoon. 
  Direct economic losses in the province were estimated at 18.1 billion
  yuan (2.2 billion US dollars).  Official statistics indicated that the
  typhoon affected 765 towns of 75 counties/cities/districts in the
  province, where 64,300 houses were toppled, 55,000 head of livestock were
  killed and the infrastructure destroyed.  The typhoon also destroyed
  1,163 kilometers of roads, damaged many embankments and water conservancy
  facilities and cut off power supplies and communication.  Some 467,900
  people were evacuated and more than 9,900 ships were called back before
  the catastrophe.

     Additional articles on the effects of Typhoon Rananim in China may be
  found at the following URL:

     http://www.reliefweb.int/w/rwb.nsf/vLND>


  G. Historical Note
  ------------------

    The China Meteorological Administration (CMA) has evaluated Typhoon
  Rananim  as the strongest landfalling typhoon for the Chinese
  Mainland since Typhoon Sally  made landfall in Wuchuan,
  Guangdong Province on September 9, 1996, and the most intense for
  Zhejiang Province since Typhoon Wanda (no number was assigned by NMC,
  which did not begin to number tropical storms until 1959) made landfall
  in Xiangshan, Zhejiang Province on August 1, 1956.

  (Report written by Kevin Boyle with significant contributions by
  Huang Chunliang)



                         TROPICAL STORM MALAKAS
                           (TC-17W / TS 0414)
                             10 - 13 August
               ------------------------------------------

  Malakas:  contributed by the Philippines, means 'strong' or 'powerful'

  A. Storm Origins
  ----------------

     During the second week of August a reverse-oriented monsoon trough
  extended from the Philippine Sea northeastward for hundreds of miles.
  Several weak circulations formed and died along the trough.  One
  disturbed area with an exposed LLCC can be seen in satellite imagery
  around 22N/150E at 0600 UTC on 8 August.  JMA began mentioning this
  system as a weak tropical depression at 09/1200 UTC, locating it near
  23N/152E and moving slowly eastward.  JTWC cashed in on the system at
  0600 UTC on the 10th, locating the weak LLCC about 730 nm west-northwest
  of Wake Island.  Convection associated with the partially-exposed LLCC
  was primarily located to the west of the center, and a 200-mb analysis
  indicated moderate diffluence aloft and weak vertical shear.  The
  development potential was initially assessed as poor.  Also at 0600
  UTC, JMA upped the MSW (10-min avg) to 30 kts.

     JTWC upgraded the potential for development to fair at 1700 UTC.
  The LLCC was then located about 670 nm west-northwest of Wake Island,
  moving northeastward at 11 kts.  Deep convection was in a cycling
  mode, and animated water vapor imagery, enhanced infrared and micro-
  wave imagery indicated that the system was basically subtropical in
  nature.  Nonetheless, JTWC initiated warnings on Tropical Depression
  17W at 1800 UTC, and at the same time JMA upgraded the depression to
  Tropical Storm Malakas.  Satellite CI estimates at the time ranged
  from 25 to 35 kts.  Malakas was located approximately 670 nm west-
  northwest of Wake Island, moving northeastward at 11 kts.


  B. Synoptic History
  -------------------

     Tropical Storm Malakas tracked rapidly northeastward along the
  northern periphery of the subtropical ridge throughout its existence.
  JTWC upgraded the system to tropical storm status at 0000 UTC on the
  11th when it was located about 1300 nm west of Midway Island and moving
  northeastward at 22 kts.  JTWC never estimated the MSW any higher than
  35 kts, although the remarks in the JTWC warnings noted that some CI
  estimates were reaching 45 kts.  At 11/1800 UTC water vapor imagery
  indicated that Malakas was becoming extratropical.   JTWC issued their
  final warning on Tropical Storm Malakas at 0600 UTC on 12 August, placing
  the center approximately 960 nm west-northwest of Midway Island.  Current
  intensity estimates ranged from 25 to 45 kts, but the system appeared
  very ragged and was declared extratropical.

     Both JMA's and NMCC's intensity estimates for Malakas were higher than
  JTWC's.    JMA maintained the MSW (10-min avg) at 40 kts for more than
  48 hours with the peak intensity of 45 kts reached at 11/1800 UTC for
  twelve hours.   NMCC's peak estimated MSW (10-min avg) was 40 kts.
  Whereas JTWC declared Malakas extratropical at 12/0600 UTC, JMA kept the
  system alive as a tropical cyclone through 1200 UTC on 13 August as it
  continued to move generally northeastward across the North Pacific.  JMA
  finally declared the system extratropical at 13/1800 UTC, placing the
  weak 25-kt LOW approximately 575 nm north-northwest of Midway Island.


  C. Damage and Casualties
  ------------------------

     No damage or casualties are known to have resulted from this short-
  lived marine cyclone.

  (Report written by Gary Padgett)



                              TYPHOON MEGI
                       (TC-18W / TY 0415 / LAWIN)
                             14 - 22 August
             ----------------------------------------------

  Megi:  contributed by South Korea, is the catfish--a large fish found
         mainly in rivers, lakes, etc and which has long whiskers around
         its mouth

  A. Introduction
  ---------------

     Typhoon Megi was the fourth of eight significant tropical cyclones to 
  form during August.  After Megi formed in the Northwest Pacific, JTWC
  issued warnings on tropical cyclones without a break through the rest of
  the month.  Megi formed well to the east of the Philippines, moved north-
  westward through the Ryukyu island chain before recurving northeastward
  towards South Korea and Japan.  Despite peaking at only minimal typhoon
  intensity,  Megi had quite a significant impact on both these nations. 


  B. Storm Origins
  ----------------

     At 2200 UTC on 11 August an area of convection persisted 
  approximately 260 nm west of Guam and was included in JTWC's STWO with
  the development potential being assessed as poor.  Animated enhanced
  infrared satellite imagery indicated a possible LLCC in connection with
  this convection.  An upper-level analysis showed moderate diffluence
  aloft and moderate wind shear over the area.  The potential for develop-
  ment remained poor through the 12th and much of the 13th.   At 13/0600
  UTC the system was relocated to a position approximately 65 nm south of
  Guam, and then repositioned again at 13/2300 UTC to a point 60 nm to the
  north-northwest of Guam.  A recent QuikScat pass indicated that the LLCC
  had consolidated significantly over the previous 12 hours with stronger
  winds within the deep convection.  On this basis, a TCFA was issued at
  this time.  The disturbance was upgraded to Tropical Depression 18W at
  14/0000 UTC.


  C. Synoptic History
  -------------------

     Tropical Depression 18W formed approximately 150 nm northwest of 
  Guam and initially tracked west-northwestward at 8 kts under the
  influence of a mid-level steering ridge to its northeast.  This heading
  persisted through the 14th while the forward speed accelerated.  There
  was little change in intensity and deep convection had become less
  organized by 14/1800 UTC.  At this time animated enhanced infrared
  satellite imagery suggested multiple LLCCs, meaning that the system
  resembled a monsoon depression.  The system appeared to have become a
  little more consolidated by 15/0000 UTC, as depicted in satellite
  imagery, but remained at depression status through the 15th.

    At 0000 UTC on 16 August Tropical Depression 18W was centred 490 nm
  south-southeast of Okinawa and continuing on its westward journey at a
  slower pace of 7 kts.  The storm still had not become any better
  consolidated at this time.  However, both JTWC and JMA upgraded the 
  system to a tropical storm and it was named Megi.  From there, Megi 
  strengthened slowly, reaching 45 kts at 16/1800 UTC after turning 
  northwestward six hours earlier.  This new heading was caused by Megi's 
  reaching the end of the subtropical ridge at the same time an upper-
  level trough was moving eastward over eastern China. 
  
     Tropical Storm Megi was accelerating north-northwestward at 0000 UTC
  on 17 August approximately 160 nm south of Okinawa.  The upper-level
  trough began to enhance Megi's outflow and the storm responded by
  strengthening to 60 kts at 17/1800 UTC.  Warnings issued by JTWC
  indicated that Megi passed 75 nm west of Okinawa at 17/1200 UTC with the
  island lying well within the radius of gale-force winds.  Although the
  system was upgraded to typhoon intensity at 18/0000 UTC (by both JTWC and
  JMA) satellite images showed a distorted circulation as it became more
  involved with the upper-level trough.  At this time Megi had reached its
  maximum intensity of 65 kts and this was to be maintained for the
  following 24 hours.  The typhoon completed recurvature at 18/1200 UTC
  approximately 210 nm west-southwest of Sasebo, Japan.

     At 0000 UTC on 19 August Megi was downgraded to tropical storm 
  status as it moved north-northeastward at a quickening pace 
  approximately 620 nm west-southwest of Misawa, Japan.  Extratropical 
  transition was well underway as its circulation crossed the southern 
  part of the Korean peninsula, northern Kyushu and then entered the 
  Sea of Japan.  From there Megi sped across northern Honshu before 
  completing extratropical transition off the east coast of Hokkaido.
  Its rapid translational speed likely limited heavy rainfall to a degree
  over South Korea and Japan but allowed little time for the storm to
  significantly weaken.   JTWC issued its final bulletin at 19/1200 UTC,
  but JMA continued to follow the storm through their bulletins.  In fact,
  that agency retained typhoon intensity until 19/1800 UTC, at which time
  Megi was demoted to severe tropical storm status.  The last mention of
  the system as a tropical cyclone was at 20/0600 UTC when it was located
  southeast of Hokkaido and moving eastward at 33 kts.   The resulting
  extratropical storm continued moving rapidly eastward, reaching a point
  near 42N/174E by 22/0600 UTC when it was last referenced in JMA's High
  Seas Bulletins.

     The lowest CP estimated by JMA was 970 mb.  This coincided with that 
  agency's peak estimated intensity of 65-kts (10-min avg).  During the
  time that Lawin (Filipino name for the system) was within PAGASA's AOR,
  the highest MSW estimated by that agency was 40 kts.   The cyclone 
  remained a tropical storm during the period it was within PAGASA's 
  boundary lines, and thus was never upgraded to typhoon status.  PAGASA
  began issuing warnings at 15/0900 UTC and ended warning coverage at
  17/0600 UTC after Lawin had exited their AOR.   NMCC's peak 10-min avg
  MSW for Megi was also 65 kts.


  D. Meteorological Observations from Japan
  -----------------------------------------

     The data in this section, and in the two following, was compiled
  and sent by Huang Chunliang.  A special thanks to Chunliang for
  sending the information.

  NOTE:  I have left all the wind observations in metres per second (mps).
  To convert to knots, divide the mps value by 0.51444.  For a quickly
  obtained approximation, just double the mps value.  An asterisk (*)
  preceding an entry denotes a record-breaking value for the relevant
  station.

  {Part I}. Landfall
  ==================

     According to the JMA warnings, Typhoon 0415 (MEGI) made landfall in
  Tsugaru Peninsula, Aomori Prefecture, around 19/2100 UTC with a MSW of
  30 m/s and a CP of 975 hPa.


  {Part II}. Top-5 Storm Totals [16/1500-20/1500Z]
  ================================================

  Ranking    Prefecture        Station        Rainfall (mm)
  ------------------------------------------------------------
  01         Ehime             Tomisato          610
  02         Kochi             Hongawa           602
  03         Miyazaki          Mikado            487
  04         Kochi             Funato            445
  05         Kochi             Ikegawa           424


  {Part III}. Top-5 Daily Rainfall Observations
  =============================================

  Ranking    Prefecture        Station               Rainfall (mm)
  ---------------------------------------------------------------------
  01         Ehime             Tomisato         *398 [16/1500-17/1500Z]
  02         Miyazaki          Mikado            338 [16/1500-17/1500Z]
  03         Kochi             Ikegawa           297 [17/1500-18/1500Z]
  04         Kochi             Hongawa           289 [17/1500-18/1500Z]
  05         Kochi             Funato            273 [17/1500-18/1500Z]


  {Part IV}. Top-5 Hourly Rainfall Observations
  =============================================

  Ranking    Prefecture        Station               Rainfall (mm)
  ---------------------------------------------------------------------
  01         Miyazaki          Mikado            123 [17/1210-17/1310Z]
  02         Kanagawa          Hakone             96 [17/0840-17/0940Z]
  03         Hyogo             Sumoto             82 [17/0810-17/0910Z]
  04         Ehime             Tomisato           70 [17/0610-17/0710Z]
  05         Kochi             Hongawa            64 [17/0540-18/0640Z]


  {Part V}. Top-5 Peak Sustained Wind (10-min avg) Observations
  =============================================================

  Ranking    Station                                   Peak wind (mps/dir)
  ------------------------------------------------------------------------
  01      Tobishima, Yamagata (JMA35002, Alt 58m)      34  /WSW [19/2000Z]
  02      Erimomisaki, Hokkaido (JMA22391, Alt 63m)    28  /NE  [20/0130Z]
  03      Izuhara, Nagasaki (WMO47800, Alt 4m)        *27.1/SSE [18/2050Z]
  04      Nomozaki, Nagasaki (JMA84596, Alt 190m)      27  /SE  [18/1600Z]
  05      Ryotsu, Niigata (JMA54166, Alt 2m)          *26  /SW  [19/1740Z]


  {Part VI}. Top-5 Peak Gust Observations
  =======================================

  Ranking    Station                                   Peak wind (mps/dir)
  ------------------------------------------------------------------------
  01      Izuhara, Nagasaki (WMO47800, Alt 4m)        *48.7/SSE [18/2036Z]
  02      Fukue, Nagasaki (WMO47843, Alt 25m)          41.2/S   [18/1723Z]
  03      Akita, Akita (WMO47582, Alt 6m)              41.1/SW  [19/1839Z]
  04      Sakata, Yamagata (WMO47587, Alt 3m)          39.9/SSW [19/1853Z]
  05      Hachinohe, Aomori (WMO47581, Alt 27m)        39.2/SW  [20/0006Z]


  {Part VII}. Top-5 SLP Observations
  ==================================

  Ranking    Station                              Min SLP (hPa)
  ---------------------------------------------------------------
  01         Izuhara, Nagasaki (WMO47800)        974.1 [18/2112Z]
  02         Fukaura, Aomori (WMOWMO47574)       978.7 [19/1939Z]
  03         Kumejima, Okinawa (WMO47929)        980.7 [17/0937Z]
  04         Aomori, Aomori (WMO47575)           981.3 [19/2104Z]
  05         Hachinohe, Aomori (WMO47581)        982.5 [19/2328Z]


  {Part VIII} References (Japanese versions only)
  ===============================================

  http://www.data.kishou.go.jp
  http://www.osaka-jma.go.jp


  E. Meteorological Observations from Coastal Zhejiang, China
  -----------------------------------------------------------

  1. Significant gust observations from the western periphery of Typhoon
  Megi [Aug 17-18, locally]
  ======================================================================

     Shulang Lake----30.6 m/s
     Langgang----27.7 m/s
     Haijiao----27.3 m/s
     Dongtou----26.8 m/s

  2. Significant rainfall observations from the western periphery of
  Typhoon Megi [18/0000-19/0000Z]
  ==================================================================

     Dongtou----86.4 mm
     Yuhuan----62.9 mm
     Jinhua----59.2 mm


  F. Rainfall Observations from the Republic of Korea
  ---------------------------------------------------

  17/1200-18/1200Z (Only amounts >= 100 mm listed)
  ================================================
 
  WANDO (34.40N 126.70E 35m)              332.5 mm
  GWANGJU (35.17N 126.90E 74m)            305.5 mm
  ANDONG (36.57N 128.72E 141m)            210.0 mm
  JINDO RADAR (34.47N 126.32E 477m)       191.5 mm
  MOKPO (34.82N 126.38E 39m)              177.0 mm
  JEONJU (35.82N 127.15E 55m)             171.5 mm
  JEJU (33.52N 126.53E 23m)               139.5 mm
  JEJU UPPER/RADAR (33.28N 126.17E 73m)   139.0 mm
  SANGJU (36.40N 128.15E 100m)            116.0 mm
  CHUPUNGNYEONG (36.22N 128.00E 245m)     114.0 mm


  18/0000-19/0000Z (Only amounts >= 100 mm listed)
  ================================================

  GWANGJU (35.17N 126.90E 74m)            319.5 mm
  WANDO (34.40N 126.70E 35m)              254.5 mm
  ANDONG (36.57N 128.72E 141m)            237.0 mm
  DAEGWALLYEONG (37.68N 128.77E 844m)     220.5 mm
  JINDO RADAR (34.47N 126.32E 477m)       214.0 mm
  SOKCHO (38.25N 128.57E 19m)             199.0 mm
  GANGNEUNG (37.75N 128.90E 26m)          197.5 mm
  MOKPO (34.82N 126.38E 39m)              195.5 mm
  DONGHAE RADAR (37.50N 129.13E 37m)      187.5 mm
  ULJIN (36.98N 129.42E 51m)              183.5 mm
  DAEGU (35.88N 128.62E 59m)              179.5 mm
  JEONJU (35.82N 127.15E 55m)             176.5 mm
  POHANG (36.03N 129.38E 4m)              167.5 mm
  SANGJU (36.40N 128.15E 100m)            154.0 mm
  CHUPUNGNYEONG (36.22N 128.00E 245m)     144.0 mm
  YEONGWOL (37.18N 128.47E 237m)          113.5 mm
  DAEJEON (36.37N 127.37E 72m)            104.5 mm


  18/1200-19/1200Z (Only amounts >= 100 mm listed)
  ================================================

  DAEGWALLYEONG (37.68N 128.77E 844m)     219.0 mm
  DONGHAE RADAR (37.50N 129.13E 37m)      217.5 mm
  GANGNEUNG (37.75N 128.90E 26m)          209.5 mm
  SOKCHO (38.25N 128.57E 19m)             169.0 mm
  ULJIN (36.98N 129.42E 51m)              165.0 mm
  POHANG (36.03N 129.38E 4m)              143.5 mm
  DAEGU (35.88N 128.62E 59m)              108.0 mm


  G. Damage and Casualties
  ------------------------

    News sources indicated that five people were reported dead or missing 
  after Typhoon Megi lashed South Korea with heavy rains and strong winds.
  The number left homeless by the storm rose to more than 2400.  Dozens of
  domestic flights were cancelled.  Typhoon Megi left at least ten dead in
  Japan, where the previous month's floods had already caused 15 deaths.
  Most of the casualties were due to floods and landslides while two 
  persons were lost at sea.  Also, a man was killed after being struck by
  wind-borne advertising boarding.  Shikoku and the nearby Tsushima Islands
  were particularly hard hit--205 mm of rain had fallen on some areas of
  Shikoku by 0000 GMT 20 August.    Megi's landfall on northern Japan
  resulted in large blackouts as electricity to 130,000 homes was cut.
  Some 700 people were evacuated from their homes due to the heavy rains
  and 24 airline flights were cancelled.  A group of around 165 primary
  school students stranded by a landslide in western Japan were success-
  fully rescued by helicopter.

  (Report written by Kevin Boyle with significant contributions by
  Huang Chunliang)



                           SUPER TYPHOON CHABA
                            (TC-19W / TY 0416)
                         18 August - 3 September
               -------------------------------------------

  Chaba: contributed by Thailand, is a tropical flower--the shoeflower
         (genus Hibiscus)

  A. Storm Origins
  ----------------

     The origins of the long-lived Chaba lay in an area of convection
  which developed and persisted on 18 August about 135 nm north-northwest
  of Kwajalein.  Animated water vapor imagery indicated cycling convection
  while a 200-mb analysis depicted weak to moderate vertical shear and good
  diffluence associated with an upper-level LOW to the northeast.  The
  development potential was assessed as fair.  Later on the 18th, at 1230
  UTC, JTWC issued a TCFA for the system which was then located about
  410 nm northeast of Pohnpei.  Convection was increasing and the system's
  organization was improving.  An 18/0952 UTC SSM/I pass indicated the
  presence of a LLCC associated with the deep convection.  JTWC issued the
  first warning on developing Tropical Depression 19W at 18/1800 UTC,
  locating the center approximately 1080 nm east of Guam, moving west-
  northwestward at 8 kts with an initial intensity of 30 kts.  JMA also
  classified the system as a 30-kt tropical depression at 1800 UTC.


  B. Synoptic History
  -------------------

     Tropical Depression 19W was located in a favorable environment for
  intensification with good poleward outflow and an outflow linkage with
  an upper-level LOW to the northeast.  JTWC upgraded the system to
  tropical storm status at 0600 UTC on 19 August, and JMA followed suit
  six hours later, assigning the name Chaba.  At 1200 UTC Tropical Storm
  Chaba was located approximately 900 nm east of Guam with 45-kt winds,
  moving northwestward at 12 kts.  The cyclone was forecast to continue in
  a general westerly to west-northwesterly direction under the steering
  influence of a ridge to the north.   Chaba continued to slowly intensify
  as it moved westward on the 19th and 20th.  Winds had reached 60 kts
  by 20/0000 UTC, and at 1200 UTC JTWC upgraded the cyclone to typhoon
  intensity, based on CI estimates of 45 and 65 kts.   The storm was then
  centered approximately 620 nm east of Guam and moving westward at 14 kts.
  (JMA did not upgrade Chaba to typhoon status for another 30 hours.)
  JTWC's upgrade to typhoon status was possibly premature--a detailed
  analysis around 1800 UTC indicated that the upper-level circulation
  might be decoupled from the surface circulation.

     However, on the 21st intensification began to accelerate.  A 21/0401
  UTC AMSU-B pass revealed a large, irregular, cloud-filled eye.  The
  MSW was upped to 75 kts at 1200 UTC and to 90 kts at 1800 UTC, based
  on CI estimates of 77 and 90 kts.  Also at 1800 UTC, JTWC upgraded
  Chaba to a 75-kt typhoon.  The cyclone was located approximately 185 nm
  east-southeast of Saipan and tracking westward at 15 kts.  The 22nd of
  August was a stormy day in the Mariana Islands.  The intensifying typhoon
  approached, passing between Tinian and Rota shortly after 1200 UTC.  The
  MSW had increased to 125 kts by this time, and by 1800 UTC had jumped
  up to 155 kts, making Chaba the 4th super typhoon of the year.  At 1200
  UTC the center of Typhoon Chaba was located about 20 nm northeast of Rota
  and about 40 nm south of Tinian. 

     After passing the Marianas Super Typhoon Chaba began to track toward
  the west-northwest and later northwest as it moved around the south-
  western periphery of the steering ridge.  Satellite CI estimates remained
  at 140 and 155 kts throughout the 23rd; consequently JTWC's MSW estimate
  remained at 155 kts until 1800 UTC, when it was reduced slightly to
  150 kts.  JMA increased their 10-min avg MSW estimate to 105 kts at
  22/1800 UTC where it remained pegged for 36 hours.  At its peak Chaba
  was an average-sized typhoon with gales covering an area roughly 400 nm
  in diameter.  At 23/1800 UTC the typhoon was centered approximately
  465 nm south of Iwo Jima, moving northwestward at 11 kts.  Chaba under-
  went a slight weakening on the 24th with the intensity dropping to
  140 kts at 1200 UTC, but by 1800 UTC had re-intensified slightly to
  150 kts.  JMA's 10-min avg MSW reached a peak of 110 kts at 24/0600
  UTC and remained at that level until 26/0000 UTC.  The minimum central
  pressure estimated by JMA was 910 hPa.  Super Typhoon Chaba continued
  tracking northwestward on the 24th and at 1800 UTC was located about
  325 nm south-southwest of Iwo Jima.

     Chaba's track became increasingly north-northwesterly on 25 August
  as it tracked around the western periphery of the ridge.  The storm
  maintained intensity with JMA's 10-min avg MSW remaining at 110 kts
  and JTWC's 1-min avg MSW dropping to only 145 kts.   At 1800 UTC on
  the 25th the cyclone was centered about 530 nm east-southeast of Okinawa
  and moving north-northwestward at 8 kts.  Some weakening occurred on the
  26th with a substantial weakening of the convection in the northern
  semicircle due to dry air entrainment.     By 1800 UTC the MSW had
  dropped to 110 kts (95 kts per JMA) and deep convection was confined
  mainly to the southern and eastern sides of the system.  Typhoon Chaba
  was located at 1800 UTC about 470 nm southeast of Sasebo, Japan, and
  still moving slowly in a north-northwesterly direction.  Over the next
  24 hours the cyclone's forward motion gradually slowed down such that
  by 27/1800 UTC it had almost come to a standstill about 390 nm south-
  east of Sasebo.  Convection along the northern periphery made a slight
  comeback during the 27th, and a 27/0932 UTC SSM/I pass revealed that
  Chaba had developed concentric eyewalls.  After weakening to 100 kts
  at 27/0600 UTC, the MSW was upped back to 115 kts at 1800 UTC.

     Chaba began to track more to the west-northwest on 28 August due to
  weak ridging northeast of the system.  The intensity remained at 115 kts
  until 28/1200 UTC, when it was lowered to 110 kts.  Satellite imagery
  around this time suggested that Typhoon Chaba was undergoing an eyewall
  replacement cycle.  At 1800 UTC the cyclone was centered approximately
  330 nm south-southeast of Sasebo, tracking west-northwestward at 7 kts.
  Slow weakening ensued on the 29th due to dry air intrusion and land
  interaction with Chaba's intensity dropping to 90 kts by 1800 UTC.  In
  response to a longwave trough entering the Sea of Japan, the storm's
  motion became increasingly poleward--at 1800 UTC Chaba was located
  165 nm south of Sasebo and moving north-northwestward at 8 kts.  By
  0000 UTC on 30 August the typhoon was accelerating to the north-
  northeast as it made landfall over extreme southeastern Kyushu.  The 
  MSW estimates from JTWC and JMA at this time were 85 kts and 80 kts,
  respectively.

     At 30/0600 UTC Chaba was located over northern Kyushu about 80 km
  southwest of Iwakuni, and six hours later was crossing southwestern
  Honshu as it continued to accelerate.  By 1800 UTC the weakening cyclone
  had emerged over the Sea of Japan and was downgraded to tropical storm
  status by both JTWC and JMA.  It was then centered approximately 315 nm
  west-southwest of Misawa, Japan, and was racing northeastward at 31 kts.
  Peak winds were estimated at 60 kts by both agencies as Chaba slowly
  began the transition into an extratropical cyclone.  Chaba subsequently
  crossed the extreme northern tip of Honshu and eastern Hokkaido.  JTWC
  issued their final warning at 31/0600 UTC when the by-now extratropical
  gale was located about 220 nm north-northeast of Misawa and racing north-
  eastward at 43 kts.  JMA followed suit and declared Chaba extratropical
  at 1200 UTC.  The extratropical LOW moved into the Sea of Okhotsk and
  became quasi-stationary just east of Sakhalin Island for several days
  as it slowly wound down.  By 0000 UTC on 3 September the former super
  typhoon had weakened into a 25-kt low-pressure area.


  C. Meteorological Observations in the Marianas
  ----------------------------------------------

     As Typhoon Chaba passed through the Mariana Islands, Saipan reported
  a maximum 1-min avg sustained wind of 49 kts with a peak gust of 61 kts
  at 22/1209 UTC.   A report of 130-kt winds gusting to 150 kts at 22/1454
  UTC which was noted in one of the JTWC warnings appears to be spurious.
  The NWS office at Agana, Guam, measured a 24-hour rainfall total of 
  144 mm between 21/1200 and 22/1200 UTC.


  D. Japanese Meteorological Observations
  ---------------------------------------

     The information in this section was supplied by Huang Chunliang
  of Fuzhou City, China.  A special thanks to Chunliang for sending
  the data.    (To convert metres/second (m/s) to knots, divide by
  0.51444, or for an approximate conversion, just double the m/s value.)

  NOTE: "*" = record-breaking values for relevant stations.


  {Part I}. Landfalls (based on the JMA warnings)
  ===============================================

  1. Severe Typhoon 0416 (CHABA) made landfall near Kushikino City,
  Kagoshima Prefecture around 30/0030 UTC with a MSW of 40 m/s and a
  CP of 950 hPa, making it the first typhoon (named tropical cyclone)
  that made landfall over Kyushu in nearly 5 years (since Bart, which
  made landfall on the 23rd of Sep, 1999).

  2. Severe Typhoon 0416 (CHABA) made landfall near Hofu City, Yamaguchi 
  Prefecture around 30/0800 UTC with a MSW of 35 m/s and a CP of 965 hPa.

  3. Typhoon 0416 (CHABA) made landfall near Tomakomai City, Hokkaido
  around 31/0500 UTC with a MSW of 30 m/s and a CP of 975 hPa.


  {Part II}. Top-5 storm totals [26/1500-31/1500 UTC]
  ===================================================

  Ranking    Prefecture        Station           Rainfall (mm)
  ------------------------------------------------------------
  01         Miyazaki          Ebino             821
  02         Nara              Mt.Hidegatake     797
  03         Miyazaki          Mikado            756
  04         Miyazaki          Kamishiiba        560
  05         Miyazaki          Miyakonojo        544


  {Part III}. Top-5 daily rainfall obs
  ====================================

  Ranking    Prefecture        Station           Rainfall (mm)
  ---------------------------------------------------------------------
  01         Miyazaki          Mikado           *584 [29/1500-30/1500Z]
  02         Miyazaki          Ebino             531 [29/1500-30/1500Z]
  03         Ehime             Jojushya          485 [29/1500-30/1500Z]
  04         Kochi             Hongawa           483 [29/1500-30/1500Z]
  05         Kagoshima         Koniya            417 [28/1500-29/1500Z]


  {Part IV}. Hourly rainfall obs
  ==============================

  No value >= 100 mm.


  {Part V}. Top-5 peak sustained wind (10-min avg) obs
  ====================================================

  Ranking    Station                                      Peak wind (mps)
  -----------------------------------------------------------------------
  01         Murotomisaki, Kochi (WMO47899, Alt 185m)     46.8 [30/1030Z]
  02         Seto, Ehime (JMA73341, Alt 143m)             39   [30/0610Z]
  03         Aburatsu, Miyazaki (WMO47835, Alt 3m)       *37.0 [29/2320Z]
  04         Tobishima, Yamagata (JMA35002, Alt 58m)     *36   [31/0040Z]
  05         Tomogashima, Wakayama (JMA65036, Alt 43m)   *33   [30/1410Z]


  {Part VI}. Top-5 peak gust obs
  ==============================

  Ranking    Station                                      Peak wind (mps)
  -----------------------------------------------------------------------
  01         Murotomisaki, Kochi (WMO47899, Alt 185m)     58.3 [30/1028Z]
  02         Makurazaki, Kagoshima (WMO47831, Alt 30m)    58.1 [29/1737Z]
  03         Aburatsu, Miyazaki (WMO47835, Alt 3m)       *55.8 [30/0147Z]
  04         Tokushima, Tokushima (WMO47895, Alt 2m)      54.1 [30/1043Z]
  05         Yakushima, Kagoshima (WMO47836, Alt 36m)     50.6 [29/1356Z]
  

  {Part VII}. Top-5 SLP obs
  =========================

  Ranking    Station                             Min SLP (hPa)
  ---------------------------------------------------------------
  01         Makurazaki, Kagoshima (WMO47831)    953.7 [29/2255Z]
  02         Hitoyoshi, Kumamoto (WMO47824)      954.1 [30/0233Z]
  03         Kagoshima, Kagoshima (WMO47827)     955.5 [30/0041Z]
  04         Naze, Kagoshima (WMO47909)          961.3 [29/0750Z]
  05         Yakushima, Kagoshima (WMO47836)     961.4 [29/1940Z]


  {Part VIII} References (Japanese versions only)
  ===============================================

  http://www.data.kishou.go.jp>
  http://www.fukuoka-jma.go.jp>


  D. Damage and Casualties
  ------------------------

     The casualty toll in Japan due to Typhoon Chaba was thirteen dead
  with four reported missing.  Electrical power was lost to 341,700
  households, and 13,000 homes were flooded.

     Additional articles on the aftermath of Typhoon Chaba can be found
  at the following link:

     http://www.reliefweb.int/w/rwb.nsf/vLND>

  (Report written by Gary Padgett with significant contributions by
  Huang Chunliang)



                             TYPHOON AERE
                      (TC-20W / TY-0417 / MARCE)
                            19 - 31 August
            ----------------------------------------------

  Aere: contributed by the United States, is the Marshallese word
        for 'storm'


  A. Storm Origins
  ----------------

     An area of convection developed approximately 250 nm east of Pohnpei
  and was included as a suspect area with poor development potential in
  JTWC's STWO at 0600 UTC on 13 August.  Animated multi-spectral satellite
  imagery revealed a weak LLCC situated in an environment of weak vertical
  shear and favourable divergence aloft.  Deep convection increased in
  association with this LLCC on the 14th, and as the system began to
  consolidate JTWC issued the first in a series of TCFAs at 15/2200 UTC.
  This statement relocated the centre to a position 205 nm east of Fananu.
  The next day at 2200 UTC the disturbance was passing 40 nm north of
  Chuuk.  A second TCFA was required at 17/0300 UTC to cover for a
  relocation and placed the centre 220 nm south-southeast of Guam.  While
  microwave imagery revealed a well-defined mid-level circulation, an
  upper-level analysis indicated the area was under moderate shear at this
  time.

     Another TCFA was issued at 18/0300 UTC as the suspect area passed
  260 nm south of Guam.  Signs of a weak LLCC were noted south of the deep
  convection in animated multi-spectral imagery.   A STWO issued at 18/0600
  UTC mentioned that while the development potential remained good, the
  system had become less organized over the previous six hours.  At 18/1230
  UTC the potential was dropped to poor after an 18/0330 UTC AMSR-E micro-
  wave image failed to show a distinct LLCC.  However, deep convection
  began to consolidate once again over the LLCC and the potential was
  raised to fair at 2200 UTC.  The final TCFA was issued at 19/0100 UTC
  when the system was passing 100 nm north of Yap.  A 200-mb analysis
  indicated a more favourable environment with weak shear and good
  diffluence aloft.  JMA first mentioned the disturbance as a tropical
  depression at 19/0600 UTC, and this was followed six hours later by 
  JTWC's first warning.


  B. Synoptic History
  -------------------

     At the time of the first warning, issued at 1200 UTC on 19 August,
  Tropical Depression 20W was located 500 nm west of Guam and heading in a
  northwesterly direction at 10 kts along the southwestern periphery of a
  mid-level steering ridge.  The system didn't appear particularly well-
  organized at this time, but it exhibited moderate convection and radial
  outflow.  Based on low shear in the immediate vicinity of the storm,
  further development was expected and the system reached tropical storm
  status at 0000 UTC on 20 August.  JMA also at this time upgraded the MSW
  to 35 kts (10-min avg) and assigned the name Aere.   Tropical Storm Aere
  was relocated at 20/0600 UTC after animated multi-spectral imagery
  revealed that the LLCC had consolidated approximately 100 nm to the
  northeast of the 0000 UTC position.  An upper-level LOW had been
  inhibiting development up to this point.  However, the LOW shifted to
  the northwest of the tropical cyclone and freed the outflow pattern to
  the north.  Overall, Tropical Storm Aere was looking healthier and by
  20/1800 UTC the MSW had increased to 55 kts.

     At 20/0000 UTC Typhoon Aere crossed into PAGASA's AOR and that agency
  named it Marce from their internal names list.  At 21/0000 UTC Tropical
  Storm Aere/Marce was still tracking northwestward some 630 nm southeast
  of Okinawa, Japan.   It was upgraded to typhoon intensity at 21/0600 UTC
  when the MSW had reached 65 kts.  Multi-spectral satellite imagery showed
  a decrease in convective coverage but cloud tops had cooled sufficiently
  enough to support continued intensification.  Typhoon Aere's intensity
  remained at 65-kts through the 21st and there was little change during
  the 22nd.  At 22/0000 UTC enhanced water vapor imagery showed that dry
  air was being advected into the eastern and northern parts of its
  circulation while microwave data showed a lack of symmetric convection
  around the LLCC.  The system was moving on its continuing northwesterly
  track and was located approximately 380 nm south of Naha, Okinawa.  Aere
  edged a little closer to the Japanese island during the course of the
  day.  JMA raised their 10-min MSW estimate to typhoon intensity at
  22/1200 UTC.

     At 23/0000 UTC Typhoon Aere was downgraded to a tropical storm briefly
  as shear increased due to a passing shortwave trough.  The cyclone was
  located 200 nm south of Naha, Okinawa, at this time.  Once the trough
  had passed by and the shear relaxed, JTWC raised the MSW back up to
  65 kts and upgraded Aere back to typhoon strength.   This intensity was
  maintained for the rest of the 23rd as the storm swayed from west-
  northwest to north-northwest and slowed its forward speed.  Typhoon Aere
  began to intensify and had reached 75 kts by 23/1800 UTC when its 50-nm
  eye was located 250 nm south of Naha, Okinawa, and moving away from the
  island.  At 24/0000 UTC deep convection had decreased to the north of the
  eye but the MSW continued to climb and reached a peak intensity of 85 kts
  at 24/1200 UTC.   Aere's heading had been wobbling from northwest to
  west-northwest, but a definite westerly heading was finally established.
  As the storm crossed the northern tip of Taiwan it started to feel the
  effects of land interaction and subsequently began to weaken.

     Typhoon Aere then turned west-southwestward at 25/0000 UTC and made
  its closest approach to Taipei, Taiwan, passing approximately 30 nm to
  the north.  The MSW started to fall as the storm crossed northern Taiwan
  and headed into the Taiwan Strait.  Aere had weakened to a 65-kt typhoon
  by the time it reached the Chinese coastline near Pingtan at 25/1200 UTC.
  It then turned southwestward, and this heading persisted into the next
  day.  This peculiar track carried the storm past Xiamen around 25/1800
  UTC and close to Shantou at 26/0000 UTC, seemingly en route to Hong Kong.
  At this time Aere was downgraded to tropical storm status and had lost
  much of its deep convection, leaving the LLCC completely exposed.
  Surface observations from Shantou reported wind speeds of around
  10-15 kts.  Aere lingered off Guangdong for awhile before turning west,
  and at the time of the final warning at 26/1200 UTC, was moving further
  inland as a 30-kt tropical depression approximately 115 nm northeast of
  Hong Kong.  The remnants of Typhoon Aere remained identifiable in
  satellite images until around 28 August, and JMA maintained the left-
  overs as a tropical depression until 0000 UTC 31 August.  JMA, NMCC, and
  HKO all estimated peak intensities of 80 kts (10-min avg).  CWB
  considered Aere as a moderate typhoon with the MSW estimated at 75 kts.
  During the time that Aere/Marce was within PAGASA's AOR, the MSW set by
  that agency was 65 kts.  The typhoon reached its maximum intensity after
  crossing PAGASA's western boundary at 23/1800 UTC.    The lowest CP
  estimated by JMA was 955 mb.


  C. Meteorological Observations from Japan
  -----------------------------------------

     The data in this section was compiled and sent by Huang Chunliang
  of Fuzhou City, China.  A special thanks to Chunliang for sending the
  information.  (To convert metres/second (m/s) to knots, divide by
  0.51444, or for an approximate conversion, double the m/s value.)

  
  {Part I}. Ryukyu obs
  ====================

  Station        Min SLP (hPa)      Peak SW (m/s)     Peak Gust (m/s) 
  ------------------------------------------------------------------- 
  Miyakojima     971.0 [23/1546Z]   27.2 [23/1600Z]   51.3 [23/1546Z] 
  Ishigakijima   961.7 [23/2305Z]   34.3 [24/0110Z]   56.0 [24/0036Z] 
  Iriomotejima   961.1 [24/0037Z]   30.6 [24/0220Z]   45.2 [24/0204Z] 
  Yonagunijima   976.6 [23/2241Z]   27.3 [23/2320Z]   43.5 [23/2316Z] 


  Station           Storm total (mm)
  -------------------------------------------
  Miyakojima        280.5  [22/0600-24/1600Z]
  Ishigakijima      314.5  [22/2100-25/1400Z]
  Iriomotejima      265.0  [22/1400-25/1100Z]
  Yonagunijima      202.0  [22/2000-25/0200Z]

  Note 1: The 24-hr [23/1500-24-1500Z] accumulations reported by the four 
  stations reached 164.0 mm, 204.0 mm, 192.5 mm and 147.5 mm, respectively.

  Note 2: Miyakojima------WMO47927, Alt 40 m
          Ishigakijima----WMO47918, Alt  6 m
          Iriomotejima----WMO47917, Alt  9 m
          Yonagunijima----WMO47912, Alt 30 m


  {Part II}. Focus on Ishigakijima, Okinawa (WMO 47918, ROIG,
             24.34 N 124.16E, Alt 6 m)
  ===========================================================

  1. Introduction
  ---------------

     The Island of Ishigakijima spent as long as 8 hrs (approximately)
  within Aere's eye, which was about 110 km in diameter, during the storm.
  See the next section for the "eye obs".

  2. Hourly sustained wind/rain/pressure obs
  ------------------------------------------

  DD/HH (UTC)    Wind (mps/dir)    Rain in past 1 hr    Pressure in hPa 
                                                        (Land/SLP)
  ---------------------------------------------------------------------
  23/08          17.9/NNE           6.5 mm              983.2/984.5
  23/09          18.4/NNE          10.0 mm              982.3/983.6
  23/10          16.0/NNE           4.5 mm              980.8/982.0
  23/11          19.1/NNE           8.0 mm              979.6/980.8
  23/12          19.5/NNE           7.0 mm              978.9/980.1
  23/13          18.6/NNE           7.5 mm              976.9/978.1
  23/14          21.8/N             8.0 mm              972.7/973.9
  23/15          24.6/N            23.0 mm              967.7/968.9
  23/16          11.7/N             6.0 mm              963.8/965.0
  23/17           6.6/NW            0.0 mm              961.8/963.0
  23/18           6.6/NW            0.0 mm              960.9/962.1
  23/19           8.2/W             0.0 mm              960.9/962.1
  23/20          11.0/WSW           0.5 mm              961.6/962.8
  23/21          15.8/SW            0.0 mm              961.8/963.0
  23/22          17.7/SW            0.5 mm              961.9/963.1
  23/23          23.8/SW            0.0 mm              961.1/962.3
  24/00          26.9/SW            2.5 mm              963.1/964.3
  24/01          30.6/SW           20.0 mm              965.0/966.2
  24/02          32.2/SSW          35.0 mm              971.1/972.3
  24/03          28.3/SSW          18.0 mm              974.0/975.2
  24/04          27.7/SW            8.5 mm              976.4/977.6
  24/05          26.4/SSW          15.0 mm              977.7/978.9
  24/06          25.3/S            21.5 mm              980.2/981.5
  24/07          22.4/SSW          26.5 mm              980.9/982.2
  24/08          22.2/SSW           8.5 mm              982.1/983.4
  24/09          22.5/S             9.0 mm              983.1/984.4
  24/10          20.8/S             7.0 mm              983.8/985.0
  24/11          16.4/S             9.5 mm              986.3/987.6
  24/12          15.5/S             5.0 mm              987.2/988.5
  24/13          15.8/S             5.0 mm              988.0/989.3
  24/14          17.4/S             5.5 mm              987.9/989.2

  Note 1: None of the Aere-related hourly sustained winds recorded outside
  the period [23/08-24/14Z] reached gale force.

  Note 2: The hourly values may not represent the true extrema.  Please
  refer to Part I for the extrema of the Ishigakijima obs.


  {Part III} References (Japanese versions only)
  ==============================================

  http://www.data.kishou.go.jp>
  http://www.okinawa-jma.go.jp>


  D. Meteorological Observations from China
  -----------------------------------------

     All the information in this section was compiled and sent by Huang
  Chunliang.  A special thanks to Chunliang for his efforts.  To convert
  wind speed in meters/second (m/s) to knots (kts), divide m/s by 0.51444.
  For an approximate conversion, simply double the m/s value.

  {Part I} Landfalls
  ==================

     According to the NMC bulletins, Typhoon 0418 (Aere) made four land-
  falls in the mainland of Fujian Province, which possesses the most
  flexuous coastline of China:  Typhoon 0418 (Aere) made landfall in
  Gaoshan Town, Fuqing City (a sub-city of Fuzhou City), Fujian Province,
  around 25/0830 UTC with a MSW of 36 m/s and a CP of 970 hPa.  As a 
  result, Aere turned out to be the first tropical cyclone to make land-
  fall in Fuzhou with typhoon intensity since Typhoon 0102 (Chebi).
  Interestingly, the town of Gaoshan was exactly where deadly Typhoon 
  0102 (Chebi) made landfall on June 23, 2001.

     Typhoon 0418 (Aere) next made landfall in Shishi City (a sub-city of
  Quanzhou City), Fujian Province, around 25/1330 UTC with a MSW of 33 m/s
  and a CP of 970 hPa.  Severe Tropical Storm 0418 (Aere) also made land-
  fall in Gangwei Town, Longhai City (a sub-city of Zhangzhou City), Fujian
  Province, around 25/1830 UTC with a MSW of 30 m/s and a CP of 975 hPa.

     Finally, Tropical Storm 0418 (Aere) made landfall in Dongshan County,
  Zhangzhou City, Fujian Province, around 26/0230 UTC with a MSW of 20 m/s
  and a CP of 985 hPa.

     This typhoon also passed over at least two of the Fujian islands, 
  including Pingtan Dao (Fuzhou City) and Nanri Dao (Putian City), the 
  former island being the fifth biggest one in China:

  (1) Typhoon Aere made landfall in Pingtan County, Fuzhou City, Fujian
  Province, around 25/0750 UTC. 

  (2) Typhoon Aere made landfall in Nanri Town, Xiuyu District, Putian
  City, Fujian Province, just shortly after moving into the sea from
  Fuqing's Longgao Peninsula, where the typhoon made its first landfall
  on the mainland.


  {Part II} Fujian Obs
  ====================

  (1) Rain
  ========

     During the 72-hr period ending at 27/0000Z, rains >100 mm were 
  recorded in 26 cities/counties, 7 of which reported rains >200 mm with 
  Fuding (located in Ningde City) reporting the highest amount of 663 mm.
  (Qinglan Reservoir located in Zherong County, Ningde City, reported the
  highest 24-hr accumulation of 504 mm.)

  (2) Wind
  ========

     13 WMO stations of coastal Fujian reported gusts of gale force or
  higher:

         Station                    Peak Gust
  ------------------------------------------------
  Zherong, Ningde City             32 m/s (ESE)
  Fu'an, Ningde City               20 m/s (ENE)
  Fuding, Ningde City              20 m/s (ENE)
  Pingtan, Fuzhou City             27 m/s (NNE)
  Fuqing, Fuzhou City              20 m/s (NE)
  Xianyou, Putian City             20 m/s (NNE)
  Tong'an, Xiamen City             22 m/s (NW)
  Xiamen, Xiamen City              33 m/s (NNW)
  Chongwu, Quanzhou City           25 m/s (NW)
  Zhangpu, Zhangzhou City          21 m/s (N)
  Dongshan, Zhangzhou City         19 m/s (W)
  Longhai, Zhangzhou City          21 m/s (WNW)

     All the insular automatic stations of northern and middle Fujian
  reported gusts of typhoon force or higher: 

  Station                    Peak Gust           DD/HH
  --------------------------------------------------------
  Taishan                  37.1 m/s (N)         24/1600Z
  Xiyang                   43.8 m/s (ENE)       25/0300Z
  Pingtan                  35.8 m/s (NNE)       25/0800Z
  Nanri                    43.8 m/s (NW)        25/0800Z
  Weitou                   34.4 m/s (S)         25/1600Z

  {Part III} Taiwan Obs
  =====================

  (1) Rain
  ========

     A few stations recorded storm totals exceeding 1000 mm:

        Station                        Rainfall (mm)
  --------------------------------------------------------
  Matala, Miaoli County          1546 [22/1600-25/1200Z]
  Pai Lan, Hsinchu County        1335 [22/1600-25/1200Z]
  Hsuen-lin, Taichung County     1243 [22/1600-25/1500Z]

  (2) Wind
  ========

     Only those stations that reported peak sustained winds of gale force
  or peak gusts of typhoon force are given:

  Station                 Peak Sustained Wind            Peak Gust
                          (mps/dir/Local Date)       (mps/dir/Local Date)
  ------------------------------------------------------------------------
  An Bu (WMO46691)          25.9/350/24th              42.2/10 /24th
  Taipei (WMO46692/58968)   13.3/320/24th              33.1/350/24th
  Chu-tzu-hu (WMO46693)     10.2/180/25th              33.1/20 /25th
  Keelung (WMO46694)        17.9/210/25th              34.4/240/25th
  Ilan (WMO46708)           20.9/330/24th              34.1/330/24th
  Lanyu (WMO46762/59567)    30.4/250/24th              44.1/250/24th
  Kinmen (WMO46736/59135)   24.1/270/25th              31.9/280/25th

  {Part IV} Zhejiang Obs
  ======================

     During the 31-hr period ending at 25/0700Z, rains >100 mm were
  recorded at 14 stations with Haishan (162.7 mm), Pingyang (157.4 mm)
  and Wencheng (152.6 mm) ranking the top three.  Significant gust reports
  included:  Zhaoshandu, Rui'an City (coastal station)--34.9 m/s; Beiji
  (insular station)--31.4 m/s; Nanji (insular station)--30.4 m/s; Dachen
  (insular station)--30.1 m/s.

     Around 24/1755Z, 6 villages located in Gaoqiao Town, Yinzhou District,
  Ningbo City, were struck by a tornado, which was triggered by Typhoon
  Aere.  The tornado did cause some economic losses, but no casualties were
  reported.

  {Part V} Guangdong Obs
  ======================

     Torrential rains lasted for nearly five days (Sep 26--30) in Guangdong
  during the period when the remnant depression, formerly Typhoon Aere,
  traversed the province from the neighboring Fujian. 

     Zhuhai City was hammered by Aere's downpours when the remnant arrived
  in the mouth of Pearl River on the 29th.  Five stations recorded 6-hr
  [0000-0600Z] rainfall amounts that ranged from 100 mm to 160 mm.  The
  most torrential rain--38 mm/minute--was reported by the urban area of
  the city.

  {Part VI} Other Obs (from Hainan Province, Hong Kong & Macao Special
            Administrative Regions
  ====================================================================

  Station                         Coordinates            Rainfall (mm)
  ------------------------------------------------------------------------
  Haikou, Hainan                 20.03N 110.35E      154.0 [28/00-29/00Z]
  Dongfang, Hainan               19.10N 108.62E      312.8 [27/00-30/00Z]
  Danxian, Hainan                19.52N 109.58E      252.9 [27/00-30/00Z]
  Hong Kong Int. AP, Hong Kong   22.32N 113.92E      184.1 [28/00-30/00Z]
  Taipa Grande, Macao            22.17N 113.57E      215.0 [28/00-30/00Z]

     The HKO report on TY Aere can be found at the following link:

     http://www.weather.gov.hk/informtc/aere/aere.htm>

  {Part VII} Damage and Casualties
  ================================

  (1) Fujian
  ==========

     Preliminary statistics on August 26 indicated that the typhoon had
  caused 2.485 billion yuan of direct economic losses and was responsible
  for two deaths in the province.  Aere also affected 3,479,900 residents
  in 421 towns of 48 counties of 6 cities in Fujian, where three cities
  were flooded, 10,100 houses were toppled, 236 embankments and thousands
  of water conservancy facilities were damaged.  Some 937,000 people were
  evacuated and 10,676 vessels were called back before the typhoon's
  arrival.

  (2) Taiwan
  ==========

     Typhoon Aere pounded northern Taiwan with torrential rains and strong
  winds before hugging the coast of Fujian, causing widespread disruption
  to air and sea transport.  It caused 24 deaths and left nine people
  missing in the region.  Water supply to 910,000 households was cut off,
  and power supply to 360,000 households was disrupted.  The economic
  losses were estimated to have been at least NT$ 400 million.

  {Part VIII} First "Black Typhoon" for Fuzhou
  ============================================

     The local government of Fuzhou, the provincial capital of Fujian,
  ordered work to stop at all construction sites and cancelled after-class
  activities at schools and universities when the city was under their
  first Black Typhoon Signal in history.  (Starting from 2003, a system of
  color-coded typhoon warning signals labeled white, green, yellow, red
  and black in an ascending order was employed in Fujian Province to give
  information to the residents on the existence and the potential threat
  of a tropical cyclone.  A Black Typhoon Signal, the most severe of the
  five grades, indicates that a tropical cyclone is affecting the district
  or is to affect the district within the next 12 hours with sustained
  wind of Beaufort Force 12 or higher.)

     Typhoon Aere played havoc with the traffic of the city.  Flights
  associated with the Changle Int. AP, which happened to be the destination
  of my flight which departed from Shanghai (please refer to Part IX for
  HCl's encounter with Typhoon Aere), were either cancelled or delayed and
  the whole airdrome had to be shut down for several hours during the
  typhoon.  Meanwhile, the traffic of the urban area, as well as several
  thruways starting from the city, were also under extraordinary control.
  Trains and buses travelling on local short-distance routes, however,
  were immune from the storm.

  {Part IX} HCl's "Reconnaissance Mission" of Aere
  ================================================

     Below is my experience of coming up against Typhoon Aere in the air
  10,000 meters above the sea level.  (Time in BJT, i.e., GMT + 8 hours.)

     I booked an airline ticket of MF8548 (Shanghai--Fuzhou, 05:05 p.m. on
  the 25th of August) 9 days ahead of schedule.   Of course, I didn't know
  beforehand that it would become my first "reconnaissance mission" of a
  typhoon, which shared the same destination with me exactly on the same
  afternoon.

     I had just arrived at the Shanghai Hongqiao Int. AP that afternoon 
  when I saw the notice, saying that MF8548 had been cancelled due to the 
  severe weather condition at the port of destination.  Nevertheless, a few
  hours later we were informed that all the scheming passengers of MF8548 
  had been incorporated into another delayed flight, MF8542, which should 
  have been the first one heading for Fuzhou that afternoon.

     My plane eventually took off from Shanghai Hongqiao Int. AP in good
  weather condition at 09:35 p.m.  So the take-off and climb were smooth
  enough for me to enjoy the beautiful nocturnal view of Shanghai in a
  merry mood.  As soon as we reached cruising altitude, the public address
  announced to the passengers that the plane was scheduled to arrive at
  the port of destination on time at 10:35 p.m.

     Being on a night flight for the first time (though not out of my
  original intention), I was fairly busy looking out of the window just
  next to my seat--12F--during the in-flight services.  At first I
  managed to see nothing but the faint horizon.  However, as we flew more
  and more southward, the grey clouds emerged and then increased gradually.
  I also noticed that all the video screens overhead were kept off all
  the way.  Besides, everything remained well-regulated with a few weak
  turbulences until the "landing time" (of Plan A) drew near.

     A powerful turbulence burst occurred around 10:20 p.m. and lasted 
  for nearly two minutes.    The plane was bounced so violently that one 
  of the passengers became sick.  I realized afterwards that the "Fasten 
  Seat Belts" sign lit before the turbulence, was not cancelled until the 
  plane pulled in.)

     It was 10:35 p.m., the exact scheduled time for landing, but the
  plane refused to even drop in altitude!  As a result, the majority
  of the passengers (including me), whether acquaintances or not, began to
  whisper to each other.  Before long, the public address sounded again in
  due course, notifying that the landing time had been delayed to 11:05
  p.m. (so they were going to go with Plan B), which would become the
  bona fide one.

     During most of the "overtime", we were lacking in everything except
  one-by-one turbulences, among which, another powerful one made another
  passenger ill.  This one also lasted for nearly two minutes when the
  plane was struggling through the typhoon as if were riding on an
  obstinate and unruly bronco.
 
     Surprisingly, the eventual landing turned out to be a relatively
  smooth one. "Thank goodness!  At last we're safe now!"...some of the
  passengers remarked when the plane was sliding on the damp airstrip of
  Fuzhou Changle Int. AP around 11:05 p.m.
  
     Looking back on the satellite pics, the hazy eye of Typhoon Aere,
  which had weakened into an intensity of 65 knots (per NMC), was located
  near 24.6 N/118.6 E when my plane was landing near 26.0 N/119.5 E.


  E. Damage and Casualties
  ------------------------

     News sources to date indicate that Taiwan took the brunt of Typhoon
  Aere.  Thirty-four people were killed as a result of the storm, and
  fifteen died as a mudslide buried a remote mountain village in the north
  of the island.  Agricultural losses were estimated at 7.7 million New
  Taiwan dollars ($US 313,000).

     No casualties were reported from China, thanks to the evacuation of
  930,000 people from low-lying and coastal areas.    More than 40,000
  fishing boats were returned to port and flights in the region were
  cancelled.

     Forty-three deaths in the Philippines were caused by heavy rains
  induced by the typhoon.  Nearly 16,000 people were evacuated from homes
  engulfed in floodwaters.  A swollen river near the northern province of
  Nueva Ecija blocked traffic on a main road and stranded hundreds of
  commuters overnight.  Eight provinces in northern and central Luzon
  were most severely affected with 70% of the provinces under water at
  one point.

     Additional articles on the aftermath of Typhoon Aere may be found
  at the following link:

     http://www.reliefweb.int/w/rwb.nsf/vLND>

  (Report written by Kevin Boyle with significant contributions by
  Huang Chunliang)



                            TROPICAL STORM
                               (TC-21W)
                            26 - 31 August
                  ----------------------------------

     A STWO issued by JTWC on 24 August noted that an area of convection
  associated with a possible weak LLCC had developed approximately 700 nm
  east-southeast of Guam.  A QuikScat pass revealed a broad, weak circu-
  lation center with increasing, though unorganized, deep convection.  An
  upper-level analysis indicated weak to moderate vertical shear with
  weak diffluence aloft.  The potential for development was assessed as
  poor.   Twenty-four hours later the disturbance had moved westward to
  a point approximately 450 nm east-southeast of Guam with little change
  in organization.  Around 1500 UTC a QuikScat pass indicated consolidation
  of the LLCC; hence, the development potential was upgraded to fair.  A
  TCFA was issued at 25/2130 UTC as convection continued to consolidate
  around the LLCC in the face of weak vertical shear and favorable
  divergence.

     JMA classified the system as a 30-kt tropical depression at 0000 UTC
  on 26 August, and JTWC issued their first warning on Tropical Depression
  21W at 0600 UTC.  The depression was located about 435 nm east of Guam
  and moving slowly west-northwestward at 4 kts.  Animated multi-spectral
  imagery revealed well-defined low-level cloud lines converging in on
  the LLCC with deep convection consolidating over the center.   The
  depression gradually intensified and was upgraded to Tropical Storm 21W
  by JTWC at 26/1800 UTC when located about 365 nm east of Guam.  Satellite
  CI estimates were 30 and 35 kts, and convection had strengthened some
  around the LLCC.   By 27/0600 UTC the center had become fully-exposed
  with the deep convection being displaced westward over Guam.  Unfavorable
  vertical shear was forecast to continue, and the storm was downgraded
  back to depression status at 1200 UTC when centered approximately 245 nm
  east-northeast of Guam.

     The slowly weakening tropical cyclone continued to move initially
  west-northwestward, then gradually turned back to the west.  The final
  warning from JTWC at 0600 UTC on 28 August placed the center 60 nm
  north of Saipan.   JMA continued to follow the residual depression in
  their High Seas Bulletins for a few more days, the final reference being
  at 0000 UTC on 31 August when the weak LOW was located approximately
  750 nm west of Saipan.   JTWC was the only warning agency which classi-
  fied this system as a tropical storm.

     No damage or casualties are known to have resulted from short-lived
  Tropical Storm 21W.

  (Report written by Gary Padgett)



                          SUPER TYPHOON SONGDA
                        (TC-22W / TY 0418 / NINA)
                         27 August - 11 September
              ---------------------------------------------

  Songda:  contributed by Vietnam, is a branch of the Red River (the
           largest in northern Vietnam) which rises in China and is
           characterized by waterfalls with high hydroelectric potential

  A. Introduction
  ---------------

     Songda was the second super typhoon to affect the Marianas and Japan
  in a week, forming a one-two with Super Typhoon Chaba.  Songda formed in
  a similar location to Chaba and followed an almost identical track across
  the Pacific, through the northern Marianas before recurving and making
  landfall over Japan, the third typhoon to strike that nation so far this
  year.


  B. Storm Origins
  ----------------

     At 1100 UTC 26 August JTWC issued a STWO including a new area of 
  convection which had developed and persisted approximately 210 nm north-
  east of Kwajalein.  This is roughly the same area that spawned Super
  Typhoon Chaba.  Animated multi-spectral satellite imagery revealed that
  the deep convection was becoming more organized over an already
  established LLCC.  As upper-level analysis indicated weak vertical shear
  and favourable divergence over the area, the potential for development
  was assessed as fair.  A TCFA followed at 27/1130 UTC, by which time the
  disturbance was passing north of Kwajalein.  Deep convection continued
  to consolidate over the centre and multi-spectral imagery noted a weak
  spiral banding feature.  The first warning on Tropical Depression 22W
  was released by JTWC at 27/1200 UTC, locating the centre 270 nm east of
  Eniwetak, moving westward at 6 kts.


  C. Synoptic History
  -------------------

     From the time of the first warning at 1200 UTC 27 August it was all 
  systems go.  Because of the ideal environmental conditions it was located
  in, Tropical Depression 22W underwent rapid strengthening and was soon
  upgraded to tropical storm intensity at 27/1800 UTC, but the system had
  to wait a further six hours to be assigned the name Songda.  JMA upgraded
  the MSW to 35 kts (10-min avg) at 28/0000 UTC.  Rapid intensification
  continued, bringing Songda to the verge of typhoon intensity by 28/1200
  UTC.  The early stages of eyewall development were noted on a 29/1444 UTC
  AMSR-E microwave pass.  Intensification then ceased for awhile.  Mean-
  while, Songda had been tracking steadily west-northwestwards and was
  located approximately 175 nm northwest of Eniwetak at 29/0000 UTC.  The
  storm passed north of that island between 28/1400-1500 UTC, bringing
  sustained winds of tropical storm force and gusts to typhoon force.  The
  tropical cyclone was following the periphery of the low to mid-level
  ridge located to the north and was expected to continue to do so over
  the next few days.

     Continuing on its west-northwesterly heading, Songda reached typhoon
  intensity at 0600 UTC 29 August approximately 760 nm east of Saipan.  At
  this time the MSW was raised to 70 kts based on CI estimates of 55 and 65
  kts.  There was no further strengthening during the 29th, and in fact
  there was very little else to report through the day, other than a
  typhoon watch being issued for the island of Agrihan at 29/1558 UTC, and
  also a brief westerly turn at 29/1800 UTC.  Typhoon Songda was still
  located some 580 nm east of Saipan, but the other island communities of
  the northern Marianas were, by this time, more than aware of the storm's
  presence, especially after Super Typhoon Chaba's rampage through there
  only a week before.  Songda strengthened a little more to 75 kts at
  30/0000 UTC and the system began to expand in areal coverage.  The
  tropical cyclone took a brief westerly jog as it resumed its intensifi-
  cation phase at 30/1200 UTC.  By 1800 UTC Songda had become a rather
  strong typhoon with a MSW of 95 kts.

     At 0000 UTC 31 August Typhoon Songda was moving west-northwest at
  10 kts across the Pacific with an increased MSW of 105 kts.  At this time
  it was located 250 nm east-northeast of Saipan.  During the 31st Typhoon
  Songda intensified significantly to 120 kts at 0600 UTC, to 125 kts at
  1200 UTC, and to 130 kts six hours later.  Thus, Songda became the fifth
  super typhoon of 2004.  The wind radii was representative of an average
  to large-sized typhoon with gales extending up to 180 nm southwest of the
  centre and 64-kt winds up to 50 nm in all quadrants.  For the northern
  Marianas there was good news and bad news.  The good news was that Songda
  had made a brief northwest turn at 31/1200 UTC, ensuring that Saipan and
  Tinian would escape the worst of the winds.  Unfortunately, this left
  Pagan and Agrihan to bear the brunt of the storm, and typhoon-force wind
  gusts were being observed on Pagan and Agrihan as the eyewall of Songda
  approached.  At 31/2100 UTC the eye was located about 29 nm east of Pagan
  and 40 nm northeast of Alamagan.
 
     Songda was not a super typhoon for long. The MSW were lowered to
  125 kts at 0000 UTC 1 September, but this intensity was maintained for
  the rest of the day.  At this time the typhoon had slowed to around
  7 kts and was moving towards the northwest.  The centre of Songda passed
  about 17 nm north-northeast of Agrihan at 01/0300 UTC.  The tropical
  cyclone subsequently accelerated and turned back towards the west-
  northwest at 01/1200 UTC.  Continuing west-northwestward, Typhoon Songda
  began to slowly weaken on the 2nd with the MSW falling 5 kts per warning,
  bringing the intensity down to 105 kts at 02/1200 UTC.  However, the
  storm began to pick up once again and the MSW rose back up to 110 kts at
  02/1800 UTC.  Animated satellite imagery indicated an increase in pole-
  ward outflow which had resulted in improved convection on the northern
  side of Songda's circulation.  The cyclone had also turned westwards and
  was now heading for its next port of call--Okinawa. 

     At 0000 UTC 3 September Typhoon Songda was moving westwards at 10 kts
  approximately 525 nm east-southeast of Okinawa.  The intensity had held 
  steady during the past six hours, but once again increased to 115 kts at 
  03/0600 UTC and to 120 kts at 03/1200 UTC.  At 03/0600 UTC Typhoon Songda
  had earned another name--Nina--after entering PAGASA's area of warning 
  responsibility.   The storm began to slow as it turned west-northwestward
  at 03/1800 UTC.  By 04/0000 UTC Songda had moved to a position 300 nm
  southeast of Okinawa and turned northwestward toward the island at
  04/0600 UTC.  At this time, a combination of AMSR-E, TRMM, and AMSU
  microwave data revealed nearly symmetric convection surrounding the eye
  with the most intense convection located in the eastern eyewall.   As
  Songda began to approach Okinawa it strengthened a little more, reaching
  a secondary peak of 125 kts at 04/1200 UTC.   This was maintained for
  the rest of the day.

     Typhoon Songda's strength began to wane as it neared Okinawa.  The
  MSW dropped to 120 kts at 0000 UTC 5 September when it was located 80 nm
  southeast of Okinawa.  Typhoon Songda passed a short distance north of
  the island at 05/1000 UTC with the lowest SLP of 924 mb recorded at 0928
  UTC.  Weakening continued as the storm tracked to the northwest.  The
  intensity fell to 110 kts at 05/1200 UTC and remained at this strength
  for another six hours.  Songda turned north-northwestward at 06/0000 UTC,
  and then northwards as it pushed through the ridge axis on its way
  towards Japan.  The 06/0000 UTC placed the centre 285 nm south-southwest
  of Sasebo, Japan.  The MSW fell below 100 kts at this time and down to
  90 kts at 06/1200 UTC as Songda recurved north-northeastwards and began
  to accelerate.  Water vapor imagery at 06/1800 UTC showed dry air being
  sucked into the southwestern quadrant. 

     At 0000 UTC 7 September Songda was about to make landfall on the 
  northwestern coast of Kyushu and at this time was centred 50 nm southwest
  of Sasebo, Japan.  The intensity had held at 90 kts since 1200 UTC of
  the 6th, but began to slowly drop off as the storm tracked further
  inland.  Songda came ashore near the city of Nagasaki with the MSW
  (10-min avg) at 80 kts and a CP of 945 hPa.  The tropical cyclone
  accelerated northeastwards at 34 kts as it moved across southwestern
  Japan.  Weakening continued as the typhoon moved into the Sea of Japan.
  The MSW dropped below typhoon strength and Songda was downgraded to a
  tropical storm at 07/1800 UTC.  By this time its forward speed had
  increased to 50 kts.  JTWC issued the final warning at 1800 UTC, placing
  the centre approximately 100 nm north of Misawa, Japan.  Satellite
  imagery indicated that Songda was embedded in the subtropical jet and had
  completed its transformation into an extratropical LOW.  JMA released
  their final bulletin at 08/0600 UTC.  The extratropical gale continued
  eastward as it slowly weakened.  By late on the 10th it had crossed the
  Dateline in the Bering Sea and was last referenced in JMA's bulletins
  at 11/0000 UTC.

     JMA regarded Songda as a Very Severe Typhoon with a peak intensity of
  90 kts, and the lowest CP estimated by that agency was 935 hPa.  NMCC and
  CWBB estimated the MSW at 120 kts and 100 kts, respectively.  During the
  time that Songda/Nina was in PAGASA's AOR, the typhoon's maximum
  intensity was estimated at 85 kts.  HKO did not issue warnings on this
  system since it remained outside their AOR.


  D. Meteorological Observations
  ------------------------------

    Following are some observations from Japan and Korea compiled and sent
  by Huang Chunliang.  A special thanks to Chunliang for sending the
  information.  An asterisk (*) beside any number indicates a new record
  value for the relevant station.  To convert wind speed in metres/second
  (m/s) to knots (kts), divide m/s by 0.51444, or to approximate, just
  double the m/s value.


  {Part I}. Landfall (based on the JMA warnings)
  ==============================================

  1. Very Severe Typhoon 0418 (SONGDA) made landfall over northern Okinawa
  Island around 05/1000 UTC with a MSW of 45 m/s and a CP of 925 hPa.

  2. Severe Typhoon 0418 (SONGDA) made landfall near Nagasaki City around
  07/0030 UTC with a MSW of 40 m/s and a CP of 945 hPa.


  {Part II}. Top-5 Storm Totals [03/1500-08/1500Z]
  ================================================

  Ranking    Prefecture        Station           Rainfall (mm)
  ------------------------------------------------------------- 
  01         Miyazaki          Morotsuka         905
  02         Miyazaki          Mikado            573
  03         Miyazaki          Nishimera         549
  04         Ehime             Jojushya          545
  05         Tokushima         Kitou             542


  {Part III}. Top-5 Daily Rainfall Obs
  ====================================

  Ranking    Prefecture        Station           Rainfall (mm)
  ---------------------------------------------------------------------- 
  01         Miyazaki          Morotsuka         358 [05/1500-06/1500Z]
  02         Ehime             Jojushya          342 [06/1500-07/1500Z]
  03         Miyazaki          Nishimera         304 [05/1500-06/1500Z]
  04         Miyazaki          Ebino             301 [06/1500-07/1500Z]
  05         Ehime             Tomisato          282 [06/1500-07/1500Z]


  {Part IV}. Top-5 Hourly Rainfall Obs
  ====================================

  Ranking    Prefecture        Station           Rainfall (mm)
  ---------------------------------------------------------------------- 
  01         Shizuoka          Shimizu          *102 [04/1320-04/1420Z]
  02         Nagano            Nagiso            *89 [04/0700-04/0800Z]
  02         Mie               Kiraramine        *89 [05/0940-05/1040Z]
  04         Shizuoka          Inatori           *88 [04/1120-04/1220Z]
  05         Mie               Kiinagashima      *85 [05/1130-05/1230Z]


  {Part V}. Top-5 Peak Sustained Wind (10-min avg) Obs
  ====================================================

  Ranking    Station                                 Peak wind (mps/dir)
  ----------------------------------------------------------------------
  01     Seto, Ehime (JMA73341, Alt 143m)            *42   [07/0230Z]
  02     Nomozaki, Nagasaki (JMA84596, Alt 190m)      38   [06/2330Z]
  03     Okinoerabu, Kagoshima (WMO47942, Alt 27m)    36.7 [05/1330Z]
  04     Hiroshima, Hiroshima (WMO47765, Alt 4m)      33.3 [07/0540Z]
  05     Ube, Yamaguchi (JMA81436, Alt 5m)           *32   [07/0110Z]


  {Part VI}. Top-5 Peak Gust Obs
  ==============================

  Ranking    Station                                 Peak wind (mps/dir)
  ----------------------------------------------------------------------
  01     Hiroshima, Hiroshima (WMO47765, Alt 4m)     *60.2 [07/0520Z]
  02     Asosan, Kumamoto (WMO47821, Alt 1142m)       57.1 [07/0419Z]
  03     Saigou, Shimane (WMO47740, Alt 27m)         *55.8 [07/0809Z]
  04     Okinoerabu, Kagoshima (WMO47942, Alt 27m)    53.6 [05/1314Z]
  05     Unzendake, Nagasaki (WMO47818, Alt 678m)    *53.2 [07/0122Z]


  {Part VII}. Top-5 SLP Obs
  =========================

  Ranking    Station                             Min SLP (hPa)
  ----------------------------------------------------------------
  01         Nago, Okinawa (WMO47940)           *924.4 [05/0928Z]
  02         Saga, Saga (WMO47813)               944.3 [07/0140Z]
  03         Nagasaki, Nagasaki (WMO47817)       948.0 [07/0044Z]
  04         Iizuka, Fukuoka (WMO47809)          948.7 [07/0228Z]
  05         Naha, Okinawa (WMO47936)            950.0 [05/0723Z]


  {Part VIII} References (Japanese versions only)
  ===============================================
 
     http://www.data.kishou.go.jp>
     http://www.osaka-jma.go.jp/gyomusyokai/kikocho/saigai/h16/ty200418.pdf>


  {Part IX} Rainfall Obs from the REPUBLIC OF KOREA
  =================================================

  Only amounts >= 100 mm listed:

  ULLEUNGDO (37.48N 130.90E 220m)         112.0 mm [06/00-07/00Z]
  ULLEUNGDO (37.48N 130.90E 220m)         101.5 mm [07/00-08/00Z]
  MUNSAN (37.88N 126.75E 31m)             103.4 mm [06/12-07/12Z]
  POHANG (36.03N 129.38E 4m)              110.5 mm [06/12-07/12Z]


  E. Damage and Casualties
  ------------------------

     News reports indicate that Typhoon Songda killed 20 people and injured
  700 others in Japan.  In addition, 15 crew members of a vessel were
  reported missing.  Songda arrived shortly after three earthquakes had
  struck the country a few days prior.

  (Report written by Kevin Boyle with contributions by Huang Chunliang)

  *************************************************************************

  NORTH INDIAN OCEAN (NIO) - Bay of Bengal and Arabian Sea

  Activity for August:  No tropical cyclones

  *************************************************************************

  SOUTHWEST INDIAN OCEAN (SWI) - South Indian Ocean West of Longitude 90E

  Activity for August:  1 tropical depression


             Southwest Indian Ocean Tropical Activity for August
             ---------------------------------------------------

     The 2004-2005 Southern Hemisphere season got off to an early start
  with the formation of a tropical depression (designated as Tropical
  Depression 01) by Meteo France La Reunion.  This system formed just
  west of 90E and subsequently moved southeastward into Perth's AOR where
  it became Tropical Cyclone Phoebe on 2 September (TC-01S per JTWC).
  Since Phoebe became a named cyclone in September, it will be covered
  in next month's summary.

  *************************************************************************

  NORTHWEST AUSTRALIA/SOUTHEAST INDIAN OCEAN (AUW) - From 90E to 135E

  Activity for August:  No tropical cyclones

  *************************************************************************

  NORTHEAST AUSTRALIA/CORAL SEA (AUE) - From 135E to 160E

  Activity for August:  No tropical cyclones

  *************************************************************************

  SOUTH PACIFIC (SPA) - South Pacific Ocean East of Longitude 160E

  Activity for August:  No tropical cyclones

  *************************************************************************

                             EXTRA FEATURE

     In order to shorten the amount of typing in preparing the narrative
  material, I have been in the habit of freely using abbreviations and
  acronyms.   I have tried to define most of these with the first usage
  in a given summary, but I may have missed one now and then.  Most of
  these are probably understood by a majority of readers but perhaps a
  few aren't clear to some.  To remedy this I developed a Glossary of
  Abbreviations and Acronyms which I first included in the August, 1998
  summary.  I don't normally include the Glossary in most months in
  order to help keep them from being too long.  If anyone would like to
  receive a copy of the Glossary, please e-mail me and I'll be happy
  to send them a copy.

  *************************************************************************

  AUTHOR'S NOTE:  This summary should be considered a very preliminary 
  overview of the tropical cyclones that occur in each month. The cyclone
  tracks (provided separately) will generally be based upon operational
  warnings issued by the various tropical cyclone warning centers.  The
  information contained therein may differ somewhat from the tracking and
  intensity information obtained from a "best-track" file which is based
  on a detailed post-seasonal analysis of all available data. Information
  on where to find official "best-track" files from the various warning
  centers will be passed along from time to time.

    The track files are not being sent via e-mail.  They can be retrieved
  from the archive sites listed below.  (Note: I do have a limited e-mail
  distribution list for the track files.    If anyone wishes to receive
  these via e-mail, please send me a message.)

    Both the summaries and the track files are standard text files
  created in DOS editor.  Download to disk and use a viewer such as
  Notepad or DOS editor to view the files.

     The first summary in this series covered the month of October,
  1997.   Back issues can be obtained from the following websites
  (courtesy of Michael Bath, Michael V. Padua, Michael Pitt, and
  Chris Landsea):

    http://australiasevereweather.com/cyclones/>
    http://www.typhoon2000.ph>
    http://mpittweather.com>
    ftp://ftp.aoml.noaa.gov/hrd/pub/landsea/padgett/>


     Another website where much information about tropical cyclones may
  be found is the website for the UK Meteorological Office.  Their site
  contains a lot of statistical information about tropical cyclones
  globally on a monthly basis.  The URL is:

    http://www.met-office.gov.uk/sec2/sec2cyclone/sec2cyclone.html>


                    TROPICAL CYCLONE REPORTS AVAILABLE

     JTWC now has available on its website the complete Annual Tropical 
  Cyclone Report (ATCR) for 2003 (2002-2003 season for the Southern 
  Hemisphere).  ATCRs for earlier years are available also.

     The URL is:  http://199.10.200.33/jtwc.html>

     Also, TPC/NHC has available on its webpage nice "technicolor"
  tracking charts for the 2003 Atlantic and Eastern North Pacific
  tropical cyclones; also, storm reports for all the 2003 Atlantic
  and Eastern North Pacific cyclones are now available, as well as
  track charts and reports on storms from earlier years.

     The URL is:  http://www.nhc.noaa.gov>


     A special thanks to Michael Bath of McLeans Ridges, New South Wales,
  Australia, for assisting me with proofreading the summaries.


  PREPARED BY

  Gary Padgett
  E-mail:  garyp@alaweb.com
  Phone:  334-222-5327

  Kevin Boyle  (Eastern Atlantic, Western Northwest Pacific, South
                China Sea)
  E-mail:  newchapelobservatory@btinternet.com

  John Wallace (Assistance with Eastern North Pacific)
  E-mail:  dosidicus@aol.com

  Huang Chunliang  (Assistance with Western Northwest Pacific, South
                    China Sea)
  E-mail:  huangchunliang@hotmail.com

  Simon Clarke  (Northeast Australia/Coral Sea, South Pacific)
  E-mail:  saclarke@iprimus.com.au

  *************************************************************************
  *************************************************************************

Document: summ0408.htm
Updated: 17th May, 2005

[Australian Severe Weather index] [Copyright Notice] [Email Contacts] [Search This Site]