GARY PADGETT'S MONTHLY GLOBAL TROPICAL CYCLONE SUMMARY AUGUST, 2000 (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.) *********************************************************************** ACKNOWLEDGEMENTS I would like up front to extend a very special thanks to two young gentlemen who helped me in a very big way with portions of the August summary. John Wallace, a student at the University of Texas in San Antonio, wrote the narratives for Northeast Pacific cyclones Fabio, Gilma, Hector, and Ileana. Eric Blake, a graduate student at Colorado State University and a member of the CSU seasonal forecasting team, wrote the summary for Hurricane Alberto while working a two-month stint at TPC/NHC. *********************************************************************** AUGUST HIGHLIGHTS --> Northeast and Northwest Pacific basins very active --> Destructive super typhoon strikes Taiwan and Chinese mainland --> Central North Pacific area remains active --> Long-lived Atlantic hurricane sets several records *********************************************************************** ***** Topic of the Month for August ***** MONTHLY STATISTICS FOR THE EASTERN NORTH PACIFIC The following table is based on information compiled by John Wallace of San Antonio, Texas. John prepared a monthly table for all cyclones reaching tropical storm intensity (NS), one for all storms reaching hurricane intensity (H), and one for intense hurricanes (IH), i.e., those hurricanes which reached Category 3 or higher on the Saffir/ Simpson Scale. I combined the NS and H tables into one in the interest of saving space, and just included the monthly totals and averages for the IHs. The only month in which four IHs formed was August, 1972, while several months produced three intense hurricanes. The first number in each column is the number of NS which initially reached tropical storm intensity (1-min avg MSW of 34 kts) during the month. The number in parentheses is the number of tropical cyclones which initially reached hurricane intensity (1-min avg MSW of 64 kts) during the month. John's counting procedure is apparently the same as I now personally prefer also (although I haven't always). Taking September and October, 1972, as an example: the only NS forming in October, Tropical Storm Kathleen, did not reach hurricane intensity. But Tropical Storm Joanne, which formed in September, reached hurricane intensity in October. Hence, Joanne is counted as a September NS but as a hurricane for October. Similarly, the only storm initially reaching tropical storm intensity in October, 1973, was Hurricane Lillian. But Tropical Storm Katherine, which formed in September, reached hurricane intensity in October; therefore, October is shown as having one NS but two hurricanes. TROPICAL STORMS and HURRICANES for the EASTERN NORTH PACIFIC (East of 140W) 1971 - 1999 -------------------------------------------- May Jun Jul Aug Sep Oct Nov Total 1971 1 (1) 1 (1) 6 (4) 5 (3) 2 (2) 2 (1) 1 18 (12) 1972 1 (1) 1 6 (6) 2 1 (1) 1 12 ( 8) 1973 3 (1) 4 (3) 1 3 (1) 1 (2) 12 ( 7) 1974 1 3 (2) 3 (2) 6 (4) 2 (2) 2 (1) 17 (11) 1975 2 (1) 4 (2) 5 (3) 3 (1) 1 (1) 1 16 ( 8) 1976 2 (2) 4 (1) 3 (2) 4 (2) 1 (1) 14 ( 8) 1977 1 1 1 (1) 1 (1) 3 (1) 1 (1) 8 ( 4) 1978 1 (1) 3 (2) 4 (3) 6 (4) 2 (2) 2 (1) 18 (13) 1979 2 (1) 2 (1) 2 (2) 1 (1) 2 (1) 1 10 ( 6) 1980 3 (2) 5 (1) 2 (3) 2 (1) 2 14 ( 7) 1981 1 1 (1) 3 (1) 4 (3) 2 (1) 4 (2) 15 ( 8) 1982 1 1 6 (4) 5 (3) 3 (3) 3 (1) 19 (11) 1983 1 (1) 1 (1) 6 (2) 3 (1) 4 (3) 4 (3) 1 21 (12)** 1984 2 (1) 3 (3) 3 (2) 4 (2) 4 (4) 1 1 18 (12) 1985 5 (2) 7 (2) 4 (2) 4 (4) 2 (2) 22 (12) 1986 1 (1) 2 (1) 3 (2) 5 (1) 5 (2) 1 (2) 17 ( 9) 1987 1 6 (2) 5 (3) 3 (3) 3 (1) 18 ( 9) 1988 2 6 (2) 3 (3) 1 (1) 1 13 ( 6) 1989 3 (2) 4 (2) 6 (2) 4 (3) 17 ( 9) 1990 1 (1) 4 (2) 4 (4) 3 (3) 5 (4) 3 (2) 20 (16) 1991 1 3 (2) 2 (2) 2 (1) 3 (2) 2 (2) 1 (1) 14 (10) 1992 3 (1) 6 (4) 5 (2) 5 (5) 5 (2) 24 (14) 1993 2 (1) 3 (3) 5 (4) 3 (2) 1 14 (10) 1994 3 (1) 4 (2) 4 (3) 5 (2) 1 (1) 17 ( 9) 1995 1 (1) 3 (2) 3 (1) 3 (3) 10 ( 7) 1996 1 2 (2) 2 (1) 4 (1) (1) 9 ( 5) 1997 3 4 (3) 3 (2) 5 (2) 1 (1) 1 (1) 17 ( 9) 1998 2 (1) 3 (2) 3 (2) 2 (1) 3 (3) 13 ( 9) 1999 1 (1) 2 (1) 3 (2) 2 (2) 1 9 ( 6) Total NS: 13 64 111 107 91 51 8 445 Avg NS: 0.45 2.21 3.83 3.69 3.14 1.76 0.28 15.35 ------------------------------------------------------------------------ Total H: 6 35 61 68 61 33 2 266 Avg H: 0.21 1.21 2.10 2.35 2.10 1.14 0.07 9.17 ------------------------------------------------------------------------ Total IH: 0 13 34 32 33 19 0 131 Avg IH: 0 0.45 1.17 1.10 1.14 0.66 0 4.51 ** - Hurricane Winnie, which formed in December, is counted in the seasonal total for NS and H. During the average Eastern North Pacific season, approximately 56% of all tropical storms reach hurricane intensity while 31% of all storms eventually become intense hurricanes. Also, about 56% of all hurricanes reach intense hurricane status. The year 1966 is most often quoted as the year in which Eastern North Pacific tropical cyclone records became reliable due to the availability of complete operational satellite coverage beginning that year. However, the Best Track file for the NEP basin does not show a single intense hurricane for the years 1966-1970, whereas in 1971 there were six. That season, as well as the following two, saw fairly extensive aerial reconnaissance of Eastern North Pacific cyclones, so, even though the initial Dvorak method had not been developed in the early 1970s, the reconnaissance data, in conjunction with some earlier methods of assessing storm intensity from satellite imagery, give the Best Track MSW values a reasonable degree of reliability beginning with 1971. The maximum number of intense hurricanes seen to develop in Eastern Pacific waters is eight, and that has occurred four times: 1983, 1985, 1992 and 1993. In 1997 seven hurricanes reached Category 3-plus status, including Hurricane Linda, which was the most intense tropical cyclone observed in the NEP basin since the advent of satellites. Several years have seen six intense hurricanes develop while the only year with no major storm was 1977, and only one was observed in 1981. *********************************************************************** ACTIVITY BY BASINS ATLANTIC (ATL) - North Atlantic Ocean, Caribbean Sea, Gulf of Mexico Activity for August: 1 tropical depression 1 possible tropical depression or storm ** 1 subtropical storm 2 tropical storms 2 hurricanes ** - no warnings were issued on this system by TPC/NHC NOTE: Much of the information presented below was obtained from the TPC/NHC discussion bulletins issued with every regular advisory. All references to sustained winds imply a 1-min averaging period unless otherwise noted. And, as noted above, the summary for Hurricane Alberto was written by Eric Blake. Atlantic Tropical Activity for August ------------------------------------- Following two tropical depressions in the month of June, the Atlantic basin was very quiet through the month of July as unfavorable upper-level westerlies spread over much of the basin at lower latitudes. However, the situation changed very quickly in August as the long-lived Hurricane Alberto formed early in the month and lasted almost three weeks. Alberto became the first hurricane and the first major hurricane (Category 3 on the Saffir/Simpson scale) of the season as winds climbed to 110 kts. Three more storms were named during the month, but only one--Debby--reached minimal hurricane intensity. In addition, advisories were issued on a tropical depression (TD-04) off the east coast of Florida. Another system attended by deep convection moved northeastward off the mid-Atlantic coast on 10-13 Aug. A reconnaissance aircraft flew into the system on the afternoon of the 10th but found no closed low-level circulation. But as the LOW moved northeastward on 11 Aug it did appear to have a circulation. SAB assigned T-numbers of 2.0 during the morning of 12 Aug and a ship reported 35-kt winds around 1200 UTC. For these reasons I am including a report on this system, which I have dubbed "Lambda", in this summary. Finally, a LOW with some distinctive subtropical characteristics developed in the western Atlantic just off the southeast U. S. coast in late August and moved inland along the North Carolina coast on the 30th. David Roth of HPC has supplied me with a track for this system, and it appears to have briefly reached subtropical storm intensity on 30 Aug when there were some ship reports of winds to 35 kts. The initial weak LOW formed on 28 Aug just east of West Palm Beach and drifted northward for about 24 hours, then moved somewhat to the northeast to a point about 200 nm south of Cape Hatteras by 30/0600 UTC. The system then turned back to the west- northwest toward the Carolina coast. At 1200 UTC it was centered approximately 125 nm southwest of Cape Hatteras, and it was at this time that David's track assigns a MSW of 35 kts. The LOW began to weaken slightly as it approached the coast, and the maximum winds were estimated at 30 kts as it made landfall near Carolina Beach (just north of Cape Fear) around 1800 UTC. The system continued moving toward the west across southeastern North Carolina and South Carolina, eventually reaching north-central Georgia by 1800 UTC on the 31st. David's notes indicate that the strongest winds were found in the northeastern quadrant except just prior to landfall when some higher winds were noted on the western side. The LOW was slowly acquiring tropical characteristics as temperatures at 500 mb rose from -12 C to -6 C in less than two days near the center, and the core was neutral to slightly warm when it made landfall. (A special thanks to David for the information he provided.) Hurricane Alberto (TC-03) 4 - 23 August -------------------------- Alberto broke a number of records for the Atlantic basin as it slowly moved across the Atlantic for about 20 days. Alberto was the longest-lived tropical storm on record in August in the Atlantic basin, maintaining tropical storm status for 19.25 days. It was also the third longest-lived tropical storm on record, behind Ginger of 1971 (21.25 days) and Carrie of 1957 (19.50 days). The storm completed the largest loop ever noted over the Atlantic, spanning approximately 5 degrees latitude by 8 degrees longitude. Alberto also became a major hurricane unusually far to the north--about 35N. It is also remarkable how low the shear was during most of the life of the storm, with the storm coming close to its Maximum Potential Intensity estimate. Alberto actually increased in strength a couple of times while the water temperatures were decreasing under its path, illustrating the dominance of vertical shear on the intensification process of tropical cyclones. The first sign of Alberto can be traced to an extremely large MCS with cloud temperatures below -88 C that fired over central Africa on 30 Jul. This system tracked to the west and was very convectively active, coming off the African coast with large pressure falls and a 50-kt jet at 700 mb over Dakar. This site in Senegal received about 25 mm of precipitation as the wave moved through, which is a rather heavy amount for this location. The system moved off the coast on 3 Aug and an unknown ship reported southeast winds of 37 kts just off the African coast. Ship MZYF3 reported 45 kts and a pressure of 1007.8 mb at 0600 UTC on the 4th and Alberto was born. It is probable that this system was already a tropical storm as it emerged from the coast, judging from the excellent satellite appearance and the ship reports. Alberto moved to the west-northwest and continued to intensify under a favorable outflow pattern. A banding-type eye was noted late on the 4th and a CDO feature was observed on the 5th, with the appearance of an eye later that afternoon. Alberto was upgraded to a hurricane about 425 nm west of the Cape Verde Islands at 05/1800 UTC. However, its intensification was limited as it began to move over cool waters of about 25-26 degrees C. The storm reached an initial peak intensity of about 80 kts on the 7th before southwesterly wind shear began to take its toll. Alberto maintained minimal hurricane status until 1800 UTC on the 8th when it took a turn to the northwest into even cooler waters and weakened into a tropical storm. A weak mid-level ridge over the Central Atlantic persisted, preventing any recurvature of the system. Two days later Alberto moved over anomalously warm water in the Mid-Atlantic. This effect, combined with weaker vertical wind shear, helped to re-strengthen the storm into a hurricane by 1800 UTC on the 9th. Alberto's track began to curve more to the north and it slowly intensified. The storm passed about 215 nm east of Bermuda around 1200 UTC on 11 Aug. Unlike most hurricanes, Alberto reached its maximum intensity after recurvature, becoming a major hurricane on the morning of the 12th north of 35N. A large, 45 nm-wide eye was noted on visible satellite images and, based upon T-numbers of 6.0 from TAFB and 5.5 from SAB, the MSW was increased to its peak intensity of 110 kts later that day, notably very close to its estimated Maximum Potential Intensity (MPI) as calculated by the Demaria/Kaplan scheme. The hurricane's center was located roughly 535 nm east-northeast of Bermuda at the time it reached its estimated peak intensity. Most model guidance took Alberto rapidly northeastward over the North Atlantic, merging with an extratropical cyclone and losing tropical characteristics, but the storm had other ideas. The GFDL was the first model to indicate that the storm would miss the trough and begin to execute a loop. The track bent from northeast to east and never crossed 40N, where extremely high wind shear and cool SSTs would have killed the storm. A large high-pressure system was building in the North Atlantic and Alberto was unexpectedly trapped underneath the ridge. Alberto slowed down and weakened some as it was moving over SSTs of about 23 C. The hurricane diminished and became a sheared tropical storm on the 14th. The track took a hard turn toward the right and Alberto began to slowly move southward, averaging a forward speed of less than 10 kts. The cloud pattern at times did not look very tropical, with limited deep convection providing some evidence of Alberto taking on subtropical characteristics on the 15th. The storm weakened further--to 40 kts--and began to move toward the southwest. However, wind shear decreased and SSTs warmed along the path of the storm, allowing for the regeneration of considerable convection on the 16th as Alberto turned toward the west-southwest. The storm began to re-intensify slowly as it proceeded to move to the west on the 17th. Alberto took a sharp turn toward the northwest as a large, slow-moving mid-latitude trough was carving out over the eastern United States. A possible eyewall was noted in AMSU images late on the 17th and Alberto became a hurricane again for the third time on the 18th. This is notable because I could not find a tropical system in the Best Track database that had re-intensified into a hurricane three times after having weakened to a tropical storm, and Alberto may very well have been the first. The hurricane continued to intensify as it moved slowly northward since wind shear was near zero. Even though water temperatures were decreasing, Alberto peaked again with an intensity of 95 kts on the afternoon of the 19th with a large, 50-nm wide ragged eye. The hurricane continued to move unusually slowly to the north as it remained contained under a large ridge. The high-pressure system gradually broke down as the large trough finally began to pick up the storm. Alberto turned to the north-northeast and accelerated while traveling over progressively cooler waters, completing its large anticyclonic loop on the morning of the 21st. The hurricane maintained an eye-like feature and hurricane status until the evening of the 22nd as it was screaming northeastward, far into the North Atlantic (nearing 50N). Alberto transitioned into an extratropical cyclone just south of Iceland, as indicated by satellite pictures. It is quite unusual for a storm to retain tropical characteristics up to 53N as Alberto did, the last storm to do so being Hurricane Frances in 1980. No reports are available on the impact of Alberto on Iceland, but it is estimated that winds in excess of tropical storm force were felt there. Some swells were reported along the U. S. East Coast a few days after the storm's recurvature. Tropical Depression (TD-04) 4 - 10 August ---------------------------- A small area of disturbed weather formed on 4 Aug approximately 525 nm east-southeast of Bermuda. The attendant weak LOW initially drifted southwestward, then westward over the next several days. By 6 Aug a few showers and thunderstorms were occurring in association with the small LOW. The system passed about 300 nm south of Bermuda around 07/0000 UTC and thereafter took a short jog to the northwest before resuming its westward motion. On the afternoon of the 8th the LOW was roughly 275 nm east of Cape Canaveral when it was visited by a U. S. Air Force Reserve Hurricane Hunter flight. The plane, flying at 450 m, reported a peak FLW of 37 kts in the north quadrant at 1851 UTC, and later reported finding winds to 47 kts in the west quadrant at 2003 UTC. The minimum pressure reported by the flight was 1010 mb. However, the system lacked any organized thunderstorm activity so TPC/NHC did not at this time classify the system as a tropical depression. During the evening some winds of 20-30 kts were reported over a limited area near the center as the system moved to the west-southwest at around 8 kts. By the morning of 9 Aug some intermittent convection was occurring over a limited area near the center, and the first advisory on TD-04 was issued at 1500 UTC with the system located approximately 175 nm east-southeast of Cape Canaveral. Satellite intensity estimates were only 25 kts, but the LLCC appeared to be better defined than the day before and the initial intensity was set at 30 kts. The depression, which really seemed to be more subtropical than tropical, was moving slowly westward at only 3 kts. A reconnaissance flight during the afternoon found maximum FLW of 39 kts in the northwest quadrant with the minimum CP still holding at 1010 mb. During the evening a buoy (WMO 41010), located about 50 nm north of the center, reported winds of 20 kts. The depression reached the westernmost point in its track around 10/0600 UTC when it was centered about 70 nm east of Cape Canaveral. During the morning Melbourne Doppler radar showed some well-organized banding features associated with the small depression, but overall the cloud pattern still looked ragged due to easterly shear. Reports from buoy 41010 suggested that the depression had lost the tight wind core it had exhibitied over the previous two days. During the 10th the system's motion became generally northeastward, and by evening the strong vertical shear had taken its toll and the depression appeared weaker. A reconnaissance aircraft did find 31-kt winds southeast of the center at 2146 UTC, but since the system was weakening and the shear was still strong, TPC/NHC issued the final advisory at 11/0300 UTC. The remnants were forecast to move to the northeast and continue to generate a few isolated thunderstorms. John Wallace passed along some interesting, but somewhat disturbing, news to the author. It seems that some meteorologists, including one with a well-known private firm, repeatedly referred to TD-04 as an unnamed tropical storm online, likely because of the stronger winds found by the reconnaissance aircraft. Also, the weather page in a major newspaper in a certain large eastern U. S. city depicted the system as Tropical Storm Beryl spinning around off the Florida coast! Possible Tropical Cyclone "Lambda" 10 - 13 August ---------------------------------- The origins of this system are a little uncertain, but appear to be associated with the northern end of a tropical wave which left the west coast of Africa around 2 Aug. The wave propagated westward across the Atlantic as tropical waves normally do, and by 1800 UTC on 7 Aug was located in the vicinity of 65W where a 1012-mb LOW had formed on the wave axis about 150 nm north-northwest of San Juan, Puerto Rico. By the 8th the wave was producing showers and thunderstorms over waters north of the eastern Bahamas for several hundred miles. All this convective activity was east of the incipient TD-04. (As I have stated earlier several times, for the early history of Atlantic/Northeast Pacific systems, I rely on a running tabular log of tropical waves and LOWs which John Wallace maintains. John's file does not indicate this (the connection was likely not clear from the Tropical Weather Discussions), but according to Todd Kimberlain, the wave which produced this system was the same one which spawned Tropical Storm Beryl and also Pacific Tropical Storm Ileana, both forming from the southern extremity of the wave.) The convection associated with the wave decreased some on the 9th, but on 10 Aug had increased and become better organized. A reconnaissance flight by the Hurricane Hunters investigated the area during the afternoon but did not find a closed low-level circulation; nonetheless, the system appeared to be quite well-organized. (Most of the following is taken from some notes I kept on 11 and 12 Aug after seeing how well-organized the system looked on the morning of the 11th. These were based largely upon animated satellite imagery available on NRL's Monterrey website.) During the evening of the 10th the convective pattern looked even more concentrated and organized. Around 11/0400 UTC a fairly tightly-wound LLCC had disengaged itself from the CDO-feature and was moving northeastward. As early as 10/2200 UTC what appeared to be curved, low-level cloud lines could be seen (in infrared imagery) emerging from the north side of the CDO- feature, and the center may have been moving out by 11/0200 UTC. The convection seemed to stay in place and did not look sheared (to me), but the LLCC continued to move away to the northeast with a quite well-organized pattern of low clouds. By the afternoon of 11 Aug convection had flared up once more near the LLCC as it moved northeastward at around 13-18 kts. This was mentioned in the 2130 UTC Tropical Weather Outlook from NHC, and the track sent to the author by David Roth begins at 1800 UTC, locating the center about 375 nm east-southeast of Charleston. The convection waned some during the night, but by 12/1000 UTC there was a nice round area of deep convection centered near 38.0 N, 65.8 W, and this appeared to be over the LLCC. Infrared satellite imagery from around 1300 to 1330 UTC on 12 Aug again depicted a circular area of deep convection, and visible imagery revealed low-level cloud lines curving into the CDO-feature. SAB assigned a Dvorak T2.0 to the system at 1145 and 1745 UTC but at 2345 UTC assigned a ST1.5 (sub- tropical). There was a ship report of 30-kt winds and a 1014-mb SLP near 32.0 N, 72.0 W at 11/1800 UTC, and more importantly, another ship report of 35-kt winds and a SLP of 1009.9 mb at 1200 UTC on the 12th. This report of gale-force winds was from a point about 150 nm south-southeast of the LOW's center when it was located about 375 nm north of Bermuda. A low-level wind analysis from CIMSS between 1215 and 1645 UTC depicted some 35-kt winds in the system's southern semicircle between 800 and 950 mb. (Thanks to John Wallace for passing this information on to me.) The system continued to scoot on off to the northeast, and by late on the 12th, the deep convection had begun to weaken and the LOW began to merge with a cold front. The exact nature of this system is a little questionable, but David Roth and Todd Kimberlain expressed the opinion (and the author concurs) that to them it looked like a tropical cyclone, and the Dvorak ratings from SAB tend to back this up. The CI numbers of 2.0 would suggest 30 kts, but the 35-kt ship report well to the south of the LLCC implies that winds could have been stronger; hence, the reason for dubbing this system as "Lambda". Hopefully, this system will at least be given a review to see if it might possibly qualify as an unnamed tropical storm. Tropical Storm Beryl (TC-05) 13 - 15 August ----------------------------- As stated in the discussion of the possible tropical cyclone which raced northeastward off the U. S. East Coast on 10-12 Aug ("Lambda"), for the pre-warning history of most Atlantic and Eastern North Pacific tropical cyclones I rely on a running tabular log of all tropical waves and LOWs compiled and maintained by John Wallace (as well as on information gleaned from the regular Tropical Weather Outlooks and Discussions issued by TPC/NHC). The first entry in the Tropical Wave file for the progenitor of Beryl depicts a wave stretching across the western Caribbean around 85W on 9 Aug. However, based on a comment in some e-mail from Todd Kimberlain, it appears that this wave was the southern portion of the same wave which had spawned "Lambda" along its northern extremity. This wave had left the coast of Africa on 2 Aug and had tracked steadily across the Atlantic. (This was also the same tropical wave responsible for the development of Tropical Storm Ileana in the Eastern Pacific.) By the early morning of 10 Aug a large area of convection had formed in association with the wave over the northwestern Caribbean. The disturbed weather spent the next couple of days crossing the Yucatan Peninsula and portions of Central America, but by 12 Aug had emerged into the Bay of Campeche and a 1011-mb LOW had formed by 1800 UTC. A reconnaissance flight by the 53rd Weather Reconnaissance Squadron of the U. S. Air Force Reserves investigated the area on the afternoon of 13 Aug and found that the LOW had acquired sufficient organization to be upgraded to a tropical depression. There was no well-defined LLCC at this point, but the plane did find FLW of 34 kts with estimated surface winds of 30 kts, so based upon these findings and the large- scale circulation evident in satellite imagery, advisories were initiated on TD-05 at 2100 UTC. The poorly-defined center of the depression was stationary about 250 nm east of Tampico, Mexico, or about 300 nm southeast of Brownsville, Texas. Beginning during the early morning hours of 14 Aug there was an enormous burst of convection which proved to be sufficient to spin the depression up to tropical storm strength. A reconnaissance flight around 1400 UTC found FLW of 53 kts in the southeast quadrant; also, the plane found that the center of circulation had reformed to the west of the previous track. The depression was upgraded to Tropical Storm Beryl at 1500 UTC with the center located about 200 nm southeast of Brownsville or 150 nm east-northeast of Tampico. The discussion bulletin noted that the system had passed over an area of unusually high heat content during the morning which may have been instrumental in the sudden strengthening of the depression. The storm was still very poorly-organized, however, and a reconnaissance crew during the afternoon had difficulty in obtaining a center fix. The position of the center indicated that Beryl was moving west-northwest at around 7 kts, which meant that landfall would likely occur too soon for the storm to reach hurricane intensity. The peak FLW winds found by the aircraft were still 53 kts so the intensity was held at 45 kts, which turned out to be the peak MSW for Beryl. Beryl's circulation remained broad and poorly-organized as it continued toward the northeast Mexican coast. Several small convective bursts were noted firing on the west side of the circulation which resulted in the spin-up of several small vortices. The center of Tropical Storm Beryl reached the coast of Mexico around 0600 UTC on 15 Aug at a point almost exactly halfway between Brownsville and Tampico--about 100 nm from each city. A reconnaissance aircraft found 55-kt winds at 850 mb to the north of Beryl's center, and a ship near Tampico reported winds of 35 kts for a few hours, so the MSW was decreased to 40 kts for the 0900 UTC advisory. The storm was down- graded to a tropical depression at 1500 UTC, and the final advisory was issued at 2100 UTC. Beryl's circulation was by then beginning to dissipate in the mountains of northeastern Mexico about 300 km north- west of La Pesca. John Wallace has sent me some observations from various locations around the western Gulf of Mexico. A CMAN near Port Aransas reported a peak 2-min avg southeast wind of 24 kts at 15/1000 UTC, the highest gust being 27 kts. Buoy 42020 (26.9N, 96.7W) reported a southwest wind of 21.3 kts (8-min avg) at 15/1400 UTC with a peak gust of 27.1 kts. The maximum wave height reported was 3.3 m at 1100 UTC. Another buoy (42002), located at 25.9N, 93.6W, recorded a maximum 8-min avg southeast wind of 25.2 kts at 0900 and 1200 UTC on 14 Aug with the peak reported gust of 31 kts occurring at 14/1200 UTC. The lowest pressure in the area was generally around 1010 mb, reported at several locales. Rainfall was rather light, at least in southern Texas. Brownsville and Corpus Christi measured 1.37 cm and 2.01 cm, respectively, in the 48-hour period from 14/1200 to 16/1200 UTC. Peak winds at these stations were in the 20-25 kt range. Rainfall may have been heavier in northeastern Mexico, but the author has not received any reports of significant damage or casualties resulting from Tropical Storm Beryl. Tropical Storm Chris (TC-06) 17 - 19 August ----------------------------- The first mention of the tropical wave which spawned Chris was in the Tropical Weather Outlook issued at 2230 UTC on 14 Aug. A large tropical wave was located midway between the African coast and the Lesser Antilles. By the afternoon of the 15th the wave was located roughly 1100 nm east of the Lesser Antilles. Thunderstorm activity had increased but was not concentrated. Temperature and moisture genesis parameters provided by NOAA/NESDIS were favorable for some development over the succeeding days. The tropical wave slowly gained in organization during the next couple of days as it moved westward. A 1010-mb LOW formed on the wave axis on the 17th, and although associated convection was somewhat meager, a closed surface circulation was apparent and advisories were begun on TD-06 at 2100 UTC. The depression was located approximately 475 nm east of Martinique and was moving west-northwestward at 10 kts. By early morning of 18 Aug a persistent cluster of convection had formed near or to the north of the presumed location of the center. Convection was beginning to wrap around the LLCC by 1200 UTC and the satellite intensity estimates from TAFB, SAB, and KGWC at that hour were all 35 kts; therefore, the depression was upgraded to Tropical Storm Chris at 1500 UTC, located about 375 nm east of Guadeloupe. However, even as Chris was named cloud tops were warming and the system was showing signs of weakening. A reconnaissance flight into the storm during the afternoon of the 18th was able to find a low-level center only with great difficulty. The maximum FLW encountered was 23 kts in the north quadrant and the minimum CP measured by dropsonde was 1011 mb. Satellite imagery indicated a better developed mid-level circulation to the east of the broad LLCC. Deep convection declined sharply, possibly due to some dry air entrainment into the storm. Chris was maintained as a tropical storm--barely--on the 2100 UTC advisory but was downgraded to a depression at 19/0300 UTC. Some convection persisted into the 19th north and northeast of the center, but an upper-level trough to the west of the depression continued to induce southwesterly shearing over the system. Even though the advisory package issued at 0900 UTC forecast that Chris would regain tropical storm intensity, a reconnaissance flight during the morning was unable to find a well-defined closed circulation, so apparently the system had degenerated into an open wave about 325 nm east of San Juan. The weak remnant LOW continued moving to the west-northwest and was absorbed into a frontal boundary on 22 Aug. It is perhaps open to question as to whether Chris ever really was a tropical storm. However, three satellite intensity estimates of 35 kts argues in favor of the system having been at tropical storm strength early on 18 Aug. Many tropical depressions have been upgraded to tropical storms on the basis of only two or sometimes even only one Dvorak rating of T2.5. Eric Blake, who was employed at NHC at the time, has informed me that Chris will likely be considered a tropical storm for a six hour period. Hurricane Debby (TC-07) 19 - 24 August ------------------------ A tropical wave emerged from the west coast of Africa on 15 Aug and continued moving westward across the tropical Atlantic. A 1010-mb LOW had formed by the 16th but the disturbance at that time showed no imminent signs of developing. Early on 17 Aug a TWO issued by TPC/NHC indicated that the system, located a few hundred miles west- southwest of the Cape Verdes, was beginning to exhibit increased organization and had the potential to develop into a depression over the next couple of days. The tropical wave and associated LOW continued to track westward with little change through the 17th and 18th, but during the night of 18 Aug convection increased somewhat and this trend continued through 19 Aug. By afternoon the system had developed enough circulation and organized convection for advisories to be initiated on TD-07, located about 875 nm east of the island of Barbados and estimated to be moving westward at 15 kts. As the 20th progressed banding features began to develop around the depression's center, and with good upper-level outflow, low vertical shear, and progressively warmer waters, the system continued to slowly strengthen and was upgraded to Tropical Storm Debby at 20/1500 UTC. Debby was centered roughly 700 nm east of Martinique and the MSW was estimated at 40 kts. During the afternoon Debby's convective banding became better organized, but the intensity was held at 40 kts due to a lack of vertical alignment between the low- and mid-level circulation centers. A 20/1506 UTC TRMM overpass showed a tight mid-level circulation or even eye-like feature in the 85 GHz (ice) data, but the 37 GHz data and visible imagery suggested that the LLCC was located about 25 nm to the southwest of the mid-level center. This lack of alignment in the vertical persisted over the next few days and helped to prevent Debby from intensifying as rapidly as most of the model guidance suggested. During the night of 20-21 Aug a large, symmetric, cold CDO feature developed, but due to uncertainty in the location of the center, the intensity was held at 45 kts. The storm also encountered some light southwesterly shear which inhibited the intensification process somewhat. A reconnaissance aircraft investigated Debby during the morning of the 21st and had a difficult time closing off a well-defined low-level center; nonetheless, the plane found winds of 88 kts at an altitude of 300 m in some convection north of the center. Based on this and some other wind reports exceeding 80 kts, the MSW was increased to 60 kts in the 1500 UTC advisory. The storm was moving just north of due west at a fairly fast clip of 19 kts. This rather rapid translational speed was the primary factor causing the shear, resulting in the low-level circulation moving faster than the mid- and upper-level circulations. Convective bursts were breaking out in the northeast quadrant, and with each new burst, the LLCC would redevelop near the burst and then move out to the west. A reconnaissance flight during the evening of 21 Aug reported that the central pressure had fallen 8 mb to 996 mb in a couple of hours. Also, the Guadeloupe radar showed a partial eyewall, thereby giving an indication that Debby might become a hurricane. The tropical storm was upgraded to a hurricane around 0600 UTC on 22 Aug with the center located near the island of Barbuda. The discussion bulletin noted that the earlier sharp drop in pressure was likely caused by a supercell-type thunderstorm which had developed near the center. This feature soon dissipated and the Guadeloupe radar began to show the formation of a classic banding pattern. An eye-like feature had been spinning up and spinning down in about three-hourly increments over the past 24 hours, and Debby appeared to be developing an eye again during the morning of the 22nd. The hurricane passed over or very near Anguilla and was located just west of the island at 1200 UTC. Wind gusts to 66 kts and 52 kts were reported on St. Barthelemy and St. Martin, respectively, during the early morning hours of 22 Aug. A ragged 25-nm diameter eye was noted by the San Juan radar during the afternoon, and several convective bursts helped to spin up small eyewall mesovortices which tended to distort the overall shape of the eye. Debby's center passed about 35 nm north of San Juan around 2100 UTC, moving west-northwest at about 18 kts. After passing to the north of Puerto Rico, Debby's track began to curve more to the west due to a 500-mb ridge to the north of the hurricane. During an early morning flight into the storm a Hurricane Hunter aircraft found a minimum pressure of 995 mb and FLW of 74 kts northeast of the center, so Debby remained a hurricane for the 23/0900 UTC advisory. The San Juan WSR-88D radar indicated that a series of eyewall mesovortices had formed and then dissipated during the night and that the eyewall convection had had trouble wrapping more than halfway around the center. Dropsonde data indicated a 10-20 kt southerly flow at 200- 300 mb undercutting the outflow of Debby, and a reconnaissance flight during the morning of the 23rd found a tight mid-level circulation with most of the convection displaced to the northeast. Based on several GPS dropsondes indicating 55-60 kt winds north of the center, the MSW was decreased to 60 kts at 1500 UTC and Debby was downgraded to a tropical storm with the center approximately 25 nm east-northeast of Puerto Plata in the Dominican Republic. During the afternoon of 23 Aug the LLCC became separated from the mid-level circulation and went scooting westward along the northern coast of Hispaniola. Strong southwesterly shear and the high mountains of Hispaniola took their toll on the storm, and by 24/0000 UTC Debby was a weakening storm with 40-kt winds located over the northwest peninsula of Haiti. A reconnaissance flight during the early morning hours found a very poorly-defined 1011-mb center south of eastern Cuba with 35-40 kt winds well to the northeast. Since some sporadic convection was still firing, the MSW was maintained at 35 kts for the 0900 UTC advisory. However, another reconnaissance plane during the mid-morning was unable to find a closed low-level center, so Debby was downgraded to an open tropical wave at 24/1500 UTC between eastern Cuba and Jamaica. The discussion bulletin noted that there was still an area of deep convection with heavy rain and winds to near tropical storm force near Cuba and Jamaica which would be spreading westward to the Cayman Islands later in the day. Debby's remnant tropical wave continued moving westward for another week or more, crossing the Caribbean and Bay of Campeche and eventually emerging into the Eastern Pacific off the west coast of Mexico. The wave was last mentioned in the Tropical Weather Discussions near 131W on 4 Sep. Storm effects of Hurricane Debby were minimal in the Leeward and Virgin Islands and also on Puerto Rico. There was some disruption to shipping and airline schedules and of course to the tourist industry. A 78-year old man in San Juan was killed as he fell off a roof while trying to dismantle a television antenna. In Haiti and the Dominican Republic heavy tropical rains were more of a worry than winds. However, it appears that while some heavy downpours fell on these areas and in eastern Cuba, there were no fatalities nor major damage resulting from the rains. A boat was sunk in the harbor at the coastal village of Carenage in Haiti. One of the biggest stories connected with Hurricane Debby was the failure of most of the numerical model guidance to correctly predict the dissipation of the storm. Based for the most part on the output of the various models, from early on 22 Aug the 72-hour forecast position from the official TPC/NHC forecasts indicated a strong Category 1 or Category 2 hurricane on the Saffir/Simpson scale near or approaching the southern tip of the Florida peninsula. This continued through 1500 UTC on the 23rd, which was the advisory package downgrading Debby to a tropical storm. Not until the 23/2100 UTC advisory, issued after Debby had obviously de-coupled during the afternoon, was this threat to Florida no longer forecast. The GFDL model in particular depicted a severe threat to southern Florida. As late as 23/1200 UTC the GFDL was tracking a 932-mb hurricane near Key West. Large numbers of tourists left the Keys and many residents boarded up or otherwise prepared for a hurricane. It is certainly to be hoped that the blown forecast in the instance of Hurricane Debby will not lead to complacency or apathy in connection with a future storm. NOTE: Operationally, as noted above, Debby was upgraded to a hurricane at 0600 UTC on 22 Aug, and the highest MSW given for any regular or intermediate advisory was 65 kts. However, the monthly summary article for August on TPC/NHC's website states that Debby became a hurricane early on 21 Aug and reached its peak estimated intensity of 75 kts later that day. Apparently a re-evaluation of the reconnaissance report from the 21 Aug morning flight, along with other data and perhaps taking into account the rapid translational speed, has led the hurricane specialists to this conclusion which likely will be reflected in the official Best Track file for Debby. *********************************************************************** NORTHEAST PACIFIC (NEP) - North Pacific Ocean East of Longitude 180 Activity for August: 5 tropical storms ** 2 hurricanes ** - one of these was a "warningless" system which Mark Lander felt reached minimal tropical storm intensity NOTE: Much of the information presented below was obtained from the TPC/NHC discussion bulletins issued with every regular advisory (CPHC for locations west of 140W.) All references to sustained winds imply a 1-min averaging period unless otherwise noted. As mentioned above, the write-ups for Tropical Storms Fabio and Ileana and Hurricanes Gilma and Hector were prepared by John Wallace. Northeast Pacific Tropical Activity for August ---------------------------------------------- In contrast to July when tropical cyclone activity was somewhat below average, the month of August was rather active in the eastern half of the North Pacific Ocean. As the month opened, former Hurricane Daniel was a weakening tropical storm north of Hawaii. (See the July summary for the full report on Hurricane Daniel.) Five tropical storms formed east of 140W and were assigned names by TPC/NHC. Of these, two--Gilma and Hector--became hurricanes. Another depression in mid-month formed in subtropical latitudes just west of the Dateline and moved northeastward into the Central North Pacific, strengthened, and was named Wene by CPHC. Additionally, another system late in the month southeast and south of Hawaii was not covered in any official warnings, but Dr. Mark Lander of the University of Guam sent me a track for this system which he feels briefly reached tropical storm intensity. I have designated this system with the Greek letter "Mu". Finally, a new tropical depression (TD-13E) formed on the last day of August and became Tropical Storm Kristy on 1 September. This brief cyclone will be covered in the September summary. Tropical Storm Fabio (TC-08E) 3 - 6 August ------------------------------ The origin of Fabio can be traced back to a tropical wave of apparent African origin that tracked off the coast on 21 Jul. The wave tracked uneventfully across the Atlantic and entered the Eastern Pacific on the 30th. The wave spawned a tropical LOW on 2 Aug, after which the convection and organization increased enough to warrant the LOW's upgrade to Tropical Depression Eight-E on 3 Aug at 2100 UTC about 650 nm west-southwest of Manzanillo. The depression tracked westward under the influence of a deep-layer ridge to its north. The tropical cyclone had trouble from the start. Though Eight-E generated strong convection, the cloud pattern indicated stiff easterly shear from a very strong upper-level anticyclone north of the system; the low-level bands were exposed in its eastern semicircle, and the ill-defined LLCC was displaced from the deepest convection. The unfavorable shear was not strong enough to prevent intensification, however, and SSTs were adequate; a slight improvement in the cloud pattern combined with QuikScat data warranted Eight-E's upgrade to Tropical Storm Fabio on 4 Aug at 0300 UTC about 700 nm west of Manzanillo. Fabio's deep convection remained intermittent and asymmetric, though a large convective burst over the LLCC allowed it to strengthen slightly. The storm intensified to a MSW of 45 kts and a CP of 1000 mb at 1500 UTC on the 4th when located roughly 825 nm west of Manzanillo. This was to be Fabio's peak strength, and its organization decayed slowly thereafter. The 2100 UTC advisory on the 4th indicated that the storm might have been weaker than it was earlier that same day, but Dvorak constraints and a persistent bursting pattern allowed for only slow weakening, which was the case through the 4th and well into the 5th. Fabio held its own surprisingly well, and since shear was forecast to weaken and the storm was expected to make a southward turn into warmer waters, re-intensification was consistently forecast. However, the intermittent, sheared deep convection weakened beginning at 1500 UTC on the 5th, and the storm embarked on a more rapid decaying trend as it made an expected, but somewhat erratic, west-southwestward turn. Fabio was downgraded to a depression at 0300 UTC on 6 Aug about 1125 nm west of Manzanillo as convection weakened and became more displaced from the center. Soon, it was little more than a low-level vortex with brief bursts of convection southwest of the LLCC. Compounded with the shear was the entrainment of dry air at the middle to upper levels, as indicated by water vapor imagery. Even so, possible re-intensification was forecast right up until the last advisory, which was issued at 2100 UTC on 6 Aug when the depression was 1250 nm west of Manzanillo. At this time there were signs of interaction with Tropical Storm Gilma to the east, and the circulation became large and amorphous. Satellite data indicated winds of less than 25 kts. The remnants of Fabio generated intermittent convection for some time after the final advisory and were not absorbed by Gilma. In fact, there may have been some interaction between the two cyclones, particularly during Gilma's weakening stage, but there is no proof of this. There were no known deaths or damage associated with Fabio. Hurricane Gilma (TC-09E) 5 - 11 August ------------------------- The origin of Hurricane Gilma can be traced back to a tropical wave that was first noted in the mid-Atlantic on 27 Jul. The wave tracked uneventfully across the Atlantic, entering the NEP basin late on 2 Aug. The wave generated strong convection, and a tropical LOW developed late on the 4th. The organization of the LOW increased steadily, and QuikScat data along with impressive banding warranted its upgrade to Tropical Depression Nine-E at 0300 UTC on 5 Aug about 325 nm south- southwest of Manzanillo. The depression tracked initially west- northwestward along the periphery of a mid-level HIGH to its north. Nine-E was impressive, with plenty of strong convection and a good cyclonic signature. Outflow was well-developed, and overall upper- level conditions were favorable. With this in mind, it is strange that the tropical cyclone intensified at only a modest rate. It was upgraded to Tropical Storm Gilma on the second advisory, at 0900 UTC on 5 Aug, when it was centered some 300 nm southwest of Manzanillo. Gilma briefly maintained this strength, but a north-south elongation of the convection, a poorly-defined circulation, and multiple LLCCs resulted in its downgrade to a tropical depression at 2100 UTC on the 5th. Easterly shear impinging on Gilma also hindered further development. However, outflow remained good in all but the eastern quadrant. Even the NHC was puzzled by its lack of intensification. Convection remained strong, but amorphous and displaced north and west of the center. Occasional flare-ups of intense convection, as well as banding away from the center, made location of the broad center difficult. A discussion at 0900 UTC on the 6th remarked that the easterly shear apparently had no effect on Gilma's convection. By the next advisory, though, Gilma had finally pulled its structure together enough to be re-upgraded to tropical storm status at 1500 UTC on 6 Aug. Easterly shear remained a moderate negative factor, however, and the cyclone's forecast time over warm waters was limited. Gilma intensified steadily from its upgrade through 0900 UTC on the 7th when the trend leveled off somewhat. By the end of this period, Gilma was just below hurricane strength and had a strong CDO with a hint of a warm spot. The SSTs were very favorable, but less-than-ideal upper-level conditions prevented any rapid intensification. Gilma didn't attain hurricane status for a full day after strengthening to 60 kts--not until 0900 UTC on 8 Aug, when it became Hurricane Gilma roughly 525 nm west-southwest of Cabo San Lucas. The 65-kt initial intensity was slightly lower than the satellite estimates, based on a ragged 30-nm eye apparent in SSMI and TRMM imagery. The upgrade occurred as upper-level conditions and outflow improved, allowing Gilma to become more symmetrical. The eye quickly disappeared, but an improving satellite signature warranted an increase in intensity to a MSW of 70-kts and a CP of 984 mb at 1500 UTC on 8 Aug when Gilma was centered some 575 nm west-southwest of Cabo San Lucas. This was to be Gilma's peak, and was maintained through the next advisory, though Gilma was already decaying by then as it entered cooler waters and entrained more stable air. Gilma weakened fairly rapidly, weakening to tropical storm strength by 0300 UTC on the 9th and to a depression by 0300 UTC on the 10th; meanwhile, its track bent slightly more westward as a ridge built to its north. The last advisory on Tropical Depression Gilma was issued at 0300 UTC on 11 Aug when it was located 825 nm west of Cabo San Lucas. The remnant circulation tracked westward in tandem with the remnants of Fabio. No deaths or other damage are known from Gilma; indeed, no public advisories were even issued for this hurricane. Hurricane Hector (TC-10E) 10 - 16 August -------------------------- The "pre-history" of Hurricane Hector deserves as much mention as the hurricane itself; it is an interesting case. Hurricane Hector originated from an African tropical wave that apparently exited the African coast on 30 Jul. It tracked uneventfully across most of the central Atlantic. However, it generated strong convection upon entering the western Caribbean; indeed, a tropical LOW developed east of the wave on 6 Aug. It seemed that formation of a tropical depression was imminent. This was not the case, however--the tropical LOW apparently dissipated by the 8th, and the bulk of the disturbance remained over the Isthmus of Tehuantepec. The disturbance became re-invigorated as it tracked off the Mexican coast as a massive cyclonic disturbance. A tropical LOW formed and was upgraded to Tropical Depression Ten-E the same day, at 2100 UTC on 10 Aug, some 200 nm west-southwest of Manzanillo, Mexico. The upgrade was based on satellite estimates and a report from ship ELXB9. The depression tracked steadily westward to the south of a ridge, its convection fluctuating in a diurnal pattern. The tropical cyclone boasted a large but poorly-organized envelope, and fixes were initially difficult. Intensification was slow to occur as well, though overall conditions were favorable. Information from Socorro Island in conjunction with satellite data indicated that Ten-E was not vertically aligned as of 2100 UTC on the 11th. However, its organization improved enough to warrant its upgrade to Tropical Storm Hector in the next advisory, issued at 0300 UTC on 12 Aug, while located roughly 300 nm southwest of Cabo San Lucas. A report from ship C6RHB of 31-kt surface winds 60 nm northwest of the center was considered in the upgrade. Hector tracked westward, intensifying steadily through the 12th and into the 13th, but slightly hindered by northerly shear. Two surprising developments occurred late on the 13th. Hector reached minimal hurricane intensity at 2100 UTC on 13 Aug, with an estimated MSW of 65 kts and a CP of 987 mb, while located roughly 575 nm west- southwest of Cabo San Lucas. This was to be Hector's peak intensity. Almost simultaneously, the storm veered abruptly northwest, off its previous westward track. Neither event was forecast, though a gradual west-northwest turn had been expected. The turn may have been in response to a weakness in the ridge north of the system. Hector tracked slowly northwestward, and sported a 10 to 15 nm diameter eye from 14/0900 UTC through 15/0300 UTC. The system maintained hurricane intensity until 1500 UTC on 15 Aug, at which time it weakened to tropical storm strength. A combination of southerly shear, stable air entrainment, and cooler SSTs began to take their toll on the tropical cyclone, and its deep central convection quickly collapsed. It's interesting to note that as Hector's upgrade to hurricane status coincided with its right turn, so did its downgrade to a tropical storm coincide with a left turn--also an abrupt change. The primary steering became the low-level easterlies as convection weakened rapidly; the storm was a low-level vortex with no convective activity by 0300 UTC on the 16th. Hector weakened to a depression, and the final advisory was issued the same day, at 2100 UTC on 16 Aug, with the dissipating center about 1100 nm west of Cabo San Lucas. Ship reports at this time suggested that no closed surface circulation was present. The remnants of Hector tracked westward across the Pacific and generated squally weather in the Hawaiian Islands on 20-21 Aug as they interacted with a strong upper-level trough. No casualties or damage were reported from either Hawaii or Mexico in association with Hector. Tropical Storm Ileana (TC-11E) 13 - 17 August ------------------------------- NOTE: I have left John's summary pretty much as he wrote it. As mentioned in the write-up for Tropical Storm Beryl, the wave which spawned Beryl and Ileana was likely traceable back to one which left the coast of Africa around 2 Aug. Tropical Storm Ileana seems to have been spawned by the same wave that spawned Tropical Storm Beryl in the Gulf of Mexico. A tropical wave was first noted in the western Caribbean on 9 August. It propagated steadily westward, generating strong convection as it crossed the Yucatan peninsula into the Bay of Campeche. The wave crossed into the Pacific by the 13th, whereupon a tropical LOW quickly formed, nearly straddling the Mexican coast. Visible satellite and microwave data indicated that the organization of the LOW was sufficient to warrant its upgrade to Tropical Depression Eleven-E at 2100 UTC on 13 August about 100 nm south of Manzanillo. The depression tracked northwest, paralleling the Mexican coast south of a ridge. The Mexican government simultaneously issued a tropical storm warning for the coast after the upgrade, extending from Lazaro Cardenas to Cabo Corrientes. The rainbands from Eleven-E impinged on the coast even as the warning was issued. Based on satellite data, the depression was upgraded to Tropical Storm Ileana at 0300 UTC on 14 August while located 75 nm south- southwest of Manzanillo. All factors were favorable for steady intensification: outflow was good, shear was relatively light, and SSTs were warm. The storm maintained a northwestward track toward Baja California. A tropical storm warning and hurricane watch were issued for Baja California at 1500 UTC on the 14th from La Paz around the peninsula to Todos Los Santos since hurricane intensity was forecast. The corresponding warnings for southwest Mexico were dropped as Ileana left its vicinity. The storm maintained a more northerly component than originally forecast, perhaps due to a weakness in the subtropical ridge; a hurricane warning replaced the tropical storm warning and hurricane watch in Baja California at 0300 UTC on the 15th while tropical storm warnings were issued for the rest of the peninsula south of 25N. Though overall conditions and the storm's organization were somewhat mercurial, Ileana intensified steadily, reaching its estimated peak MSW of 60 kts and lowest CP of 990 mb at 0900 UTC on 15 August about 90 nm southeast of Cabo San Lucas. However, the 60-kt peak MSW was attained slightly before the lowest pressure. Ileana finally began to make an expected westward turn as it interacted with a ridge to its north; a fortunate development, as it steered the storm away from Baja California. It made its closest approach to Cabo San Lucas soon afterward, near 1500 UTC on the 15th when it was located roughly 45 nm to the south. This distance theoretically placed Cabo San Lucas within the storm-force wind radii. Even so, all warnings for Baja California were dropped at 2100 UTC on the 15th. Ileana maintained an estimated 60-kt MSW until 0900 UTC on the 16th when persistent northeasterly shear initiated a weakening trend, decoupling the LLCC from the convection. Intensification to hurricane strength had been forecast up until 0300 UTC the same day. Ileana weakened very rapidly; its estimated MSW dropped from 60 to 30 kts in only 18 hours while its central pressure rose 13 mb. It was devoid of deep convection by 2100 UTC on the 16th, and was declared dissipated by 0300 UTC on 17 August roughly 340 nm west-northwest of Cabo San Lucas. Ileana was a compact system; no casualties or significant damage are known to the author. An Associated Press report states that Cabo San Lucas did get some rain from Ileana, but there was no significant damage aside from power grid disruptions. Unfortunately, no useful surface observations from the area are known or available to the author. The effects of Ileana in southwest Mexico are an unknown as well. It is assumed that effects, if any, were slight. Tropical Storm Wene (TC-16W) 15 - 17 August ----------------------------- Wene: a Hawaiian name, is the transliteration of Wayne Tropical Storm Wene was the second tropical cyclone in less than a month to be named by the Central Pacific Hurricane Center in Honolulu. Prior to Upana in July, the last storm to have a Hawaiian name bestowed upon it was Paka in December, 1997. Very atypically for a Central Pacific tropical cyclone, Wene formed just west of the International Dateline and intensified just after entering the Central North Pacific moving northeastward. Wene's origin lay within an area of disturbed weather in the subtropics which had persisted for several days in mid-August west of the Dateline. An area of convection about 325 nm northwest of Midway Island began to develop rapidly on 15 Aug. The first warning on TD-16W was issued by JTWC at 0600 UTC. Animated satellite imagery depicted a well-defined LLCC with convection developing over the center. The depression was tracking slowly north-northeastward and was initially not forecast to intensify. By 1800 UTC the system's center was just east of the Dateline and the MSW was increased slightly to 30 kts. Satellite imagery depicted a partially-exposed LLCC with a band of convection extending from north to east of the center. However, an amended warning was released shortly thereafter upgrading the depression to a tropical storm about 345 nm north-northwest of Midway. Multi-spectral satellite imagery indicated that the system was more intense than previously thought with a CDO located directly over the LLCC. The CPHC assumed warning responsibility for the storm, issuing their first advisory at 16/0300 UTC and naming the system Tropical Storm Wene. Dvorak intensity estimates were 45 kts, so the MSW was increased to that value. At 0600 UTC Wene was located about 430 nm north of Midway Island and moving northeastward at 13 kts. Animated satellite imagery revealed persistent deep central convection with good organization. By 1200 UTC, however, Wene had begun to experience increased shearing due to an approaching trough in the westerlies and most of the convection lay in the northern half of the system. Dvorak estimates were ranging from 35 to 55 kts, but the MSW was decreased to 40 kts based on scatterometer data. As Wene continued moving to the north-northeast over the next 12 to 18 hours it maintained a well-defined circulation with some deep convection in spite of the shear and cooler waters. However, by 0600 UTC on 17 Aug the tropical storm was rapidly transitioning into a cold-cored extratropical LOW as it moved over even colder waters and began to be absorbed into an upper-level trough. Deep convection had disappeared and the wind field, as depicted by scatterometer data, was expanding outward. The final advisory on Wene placed the storm about 725 nm north of Midway or about 720 nm south-southeast of Adak, Alaska. Tropical Storm John (TC-12E) 28 August - 1 September ----------------------------- A tropical wave which was likely the precursor of Tropical Storm John left the African coast on 9 Aug. The wave continued moving west- ward across the tropical Atlantic over the succeeding days with an associated tropical LOW being mentioned for several days in the Tropical Weather Discussions issued by TPC/NHC. The system reached the eastern Caribbean Sea on 15 Aug and continued tracking to the west, reaching the Eastern Pacific by 19 Aug. This information was taken from John Wallace's tropical wave log, which indicates that a 1009-mb LOW had formed along the wave near 15N, 137W, by 1800 UTC on 27 Aug. As best I can determine, the first mention of the pre-John disturbance by TPC/NHC in a Tropical Weather Outlook was early on the 27th, indicating that a large area of disturbed weather was located about 1300 nm east-southeast of Hilo, Hawaii. However, convection along the ITCZ in the Central and Eastern Pacific became quite active during late August, and Mark Lander suggested that the spate of tropical cyclones during this time, including John, the system described below ("Mu"), and Tropical Storm Kristy were not of African wave origin but rather had their roots in the Pacific. This increase in tropical activity in the Northeast Pacific basin was possibly due to an active phase of the Madden-Julian Oscillation in the area. Mark Lander is of the opinion that the system in question was likely of or near tropical storm intensity by late on 27 Aug, but no advisories were issued until the first visible images on the 28th revealed that deep convection had become involved with the LLCC and the overall cloud pattern had become more impressive than during the previous evening. The first advisory on TD-12E was issued by NHC at 1700 UTC, locating the depression's center approximately 1650 nm west- southwest of Cabo San Lucas or about 1025 nm east-southeast of Hilo. There was some southwesterly shear over the area and strengthening was forecast to be slow. However, at 28/2100 UTC the depression was upgraded to a 45-kt tropical storm and named John. Visible satellite imagery showed a well-organized tropical cyclone with banding features over the eastern semicircle, and Dvorak T-numbers from TAFB had reached 3.0. Tropical Storm John was moving west-northwestward at 4 kts and was forecast to intensify to 60 kts within 36 hours. The MSW was increased to 55 kts six hours later based on estimates from TAFB and SAB. Upper-level outflow was good to the east and improving to the west, due in part to the influence of the disturbance about 500 nm to the west ("Mu") which was helping to deflect strong 200-mb shearing winds to the north of John. A small CDO feature with -80 C overshooting tops had developed near the center, indicating that the storm was strengthening. However, John's intensity leveled off and remained at 55 kts, likely due to some westerly flow approaching the storm and undercutting the outflow. Convection weakened some on the 29th but made a comeback late in the day. The storm's motion was being impeded by the southwesterly shear it was running into, but a low- and mid-level ridge to the north was strong enough to keep the cyclone moving on a slow west-northwesterly track. By 0000 UTC on 30 Aug John's center was located about 875 nm east-southeast of Hilo and was just west of 140W, so TPC/NHC issued its final advisory and warning responsibility was transferred to the CPHC in Honolulu. Tropical Storm John continued to move slowly westward after entering the Central Pacific. The North Pacific HIGH at 45N was centered far to the north of its mean position for the summer months, and this made for weak steering flow at 17N. The storm maintained its intensity in spite of the shear, and at 30/1800 UTC the center had become hidden under a cold overcast and Dvorak estimates from SAB and Honolulu edged up to T4.0--65 kts. The MSW was bumped up to 60 kts, which was the peak intensity for the storm's history. (The minimum estimated CP of 994 mb had occurred on the 29th.) However, by 31/0000 UTC the convection had dramatically declined due to increased shear and the center was difficult to locate, there being uncertainty as to which side of 140W the center was located on. The MSW was decreased back to 55 kts, and six hours later the intensity was further dropped to 45 kts. Almost all the deep convection was gone due to the persistent southwesterly shear. At this point John had been essentially stationary for the previous 24 hours just west of 140W. After 0600 UTC the weakening cyclone did drift slowly to the west for about 12 hours until it became stationary again at 1800 UTC. By this time John was a minimal tropical storm with winds down to 35 kts. The storm was downgraded to a tropical depression at 0000 UTC on 1 Sep. The center was still stationary near the 31/1800 UTC position, and the CPHC discussion indicated that John was undergoing a weak Fujiwhara- type rotation with TD-13E (later TS Kristy) which was located about 575 nm to the east-southeast. The dissipating depression, consisting of a low-level cloud swirl with no associated deep convection, remained quasi-stationary on 1 Sep with scatterometer data indicating winds of only about 25 kts, and the final advisory on John was issued at 2100 UTC with the center located approximately 775 nm east-southeast of Hilo. Tropical Cyclone "Mu" 26 August - 2 September ----------------------- As mentioned above in the discussion of Tropical Storm John, the ITCZ became active in the Central and Eastern North Pacific area in late August, possibly due to the active phase of a Madden-Julian Oscillation moving through the region. Also mentioned in the discussion of John was another disturbance to the west of the tropical storm which, at one point, likely helped to deflect some strong 200-mb westerlies to the north of the storm, thereby enabling it to become stronger than it otherwise might have. No advisories were issued on this disturbance by either NHC or CPHC, but the author received a track for this system from Dr. Mark Lander of the University of Guam who estimates that the disturbance was a minimal tropical storm for a 36-48 hour period on 26-28 Aug. A Tropical Weather Outlook (TWO) issued by NHC on 17 Aug mentioned that an area of cloudiness and showers had developed about 475 nm south of the southern tip of Baja California. This seems to be the parent disturbance from which the subject cyclone developed. The system moved steadily westward for the next few days with a weak low- level circulation occasionally referred to in the TWOs. Mark's track begins at 0000 UTC on 26 Aug with a 25-kt depression located roughly 825 nm east-southeast of Hilo, Hawaii. A TWO issued by CPHC at 1400 UTC on the 26th indicated that convection associated with the system had intensified. Mark's track indicates that the system had reached tropical storm intensity by 0000 UTC on 27 Aug when it was centered about 650 nm east-southeast of Hilo. A TWO issued at 27/1400 UTC noted that peripheral convection associated with the system had weakened since the previous evening but that new thunderstorms had recently fired near the center of circulation. The system continued to move rather slowly generally to the west. Mark Lander's track, which consists of only 12-hourly positions, main- tains the cyclone as a tropical storm through 28/0000 UTC, then down- grades it to a depression. The depression passed about 300 nm south of the Big Island around 31/0000 UTC, and the final position in the track places the system about 450 nm east-southeast of Johnston Atoll at 0000 UTC on 2 Sep. In a message received along with the track, Mark writes that the track may be a bit crude, but his primary objective was to point out that scatterometer data and a strict adherence to Dvorak analysis of satellite imagery indicate that the system was likely a 35-40 kt tropical storm from around 26/1800 through 28/1800 UTC. In another e-mail Mark cited a 27/2000 UTC ERS-2 pass with 30-kt wind vectors depicted near the center of the system. These represent 8-min avg wind speeds, averaged over a pixel, thus implying that the 1-min avg MSW was likely around 35 kts. *********************************************************************** NORTHWEST PACIFIC (NWP) - North Pacific Ocean West of Longitude 180 Activity for August: 2 tropical depressions ** 3 tropical storms ++ 3 typhoons 1 super typhoon ** - one of these was a short-lived system carried only by JMA ++ - one of these was treated as a tropical depression by JTWC and JMA but Mark Lander felt that it reached tropical storm intensity NOTE: Most of the information on each cyclone's history presented in the narrative will be based upon JTWC's advisories, and references to winds should be understood as a 1-min avg MSW unless otherwise noted. However, in the accompanying tracking document I have made comparisons of coordinates with JMA (Japan) and the Philippines (PAGASA) when their positions differed from JTWC's by usually 40-50 nm or more. A special thanks to Michael V. Padua, owner of the Typhoon 2000 website, for sending me the PAGASA and JMA tracks. In the title line for each storm I plan to reference 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 responsibility. Northwest Pacific Tropical Activity for August ---------------------------------------------- August turned out to be a very active month in the Northwest Pacific basin. Six tropical storms were named by JMA with four of these reaching typhoon intensity. Bilis became an intense super typhoon and Jelawat nearly did do (according to Mark Lander it did briefly reach super typhoon intensity). All of the named cyclones eventually made landfall on the Asian continent except for Typhoon Ewiniar. JTWC issued warnings on two additional depressions (14W and 17W), and JMA issued two bulletins on another tropical depression. However, Mark Lander sent me alternate tracks for some of the systems, and his track for TD-14W assigns a peak MSW of 50 kts. Another depression, TD-16W, formed just west of the Dateline and moved into the Northeast Pacific basin where it developed into Tropical Storm Wene. Finally, Mark Lander sent a satellite picture on 30 Aug of a very tiny vortex in the subtropics accompanied by some deep convection. It isn't exactly clear just what the nature of this "micro-midget" system was. In addition to the named cyclones Bilis and Kaemi, there were two weak tropical depressions in the NWP basin. JTWC issued six warnings on a system designated as TD-17W which formed on 17 Aug about 525 nm west-northwest of Midway Island in an area of disturbed weather in the subtropics which had persisted after Tropical Storm Wene had formed in the same general area a couple of days earlier. The depression moved erratically at first, then began to move off to the northeast. The last warning, issued at 19/0000 UTC, placed the center about 475 nm north-northwest of Midway. Maximum winds for the system were 25 kts. On 20 Aug JMA mentioned a tropical depression in a couple of their High Seas Warning and Summary bulletins. This system tracked eastward from a position about 225 nm south of Tokyo on 19 Aug. At 20/0000 UTC it was located about 350 nm southeast of Tokyo, and it was at this time that it was first referred to as a tropical depression. By 1200 UTC it was continuing to move rather quickly to the east and was described as a LOW. The final reference to the system was at 1800 UTC when it was located about 550 nm east-southeast of Tokyo. At my request Mark Lander reviewed some imagery of this system and concurred that it was classifiable as a 30-kt tropical depression. Given the latitude of formation and the fact that JTWC never mentioned the system in their STWOs, it seems likely that the LOW was perhaps more of a hybrid system than purely tropical. Regarding the "micro-midget" alluded to above--the first visible image Mark sent was taken at 30/0530 UTC and shows a very small vortex centered near 30.5N, 164.5E. The center appears to be located near the southern edge of a small convective mass no more than 60 nm in diameter. A second visible image taken at 31/0500 UTC depicts a tiny vortex near 30.9N, 165.4E, and Mark indicates he checked several infrared pictures to make sure it was the same vortex. The 31 Aug image reveals a somewhat more ragged-looking system with the convection having decreased in areal extent and the LLCC located near the western edge of the convection. Mark estimates that the maximum winds were likely in the 20-25 kt range. No track is given for this system in the cyclone tracks file. Typhoon Jelawat (TC-13W / TY 0008) 30 July - 11 August ----------------------------------- Jelawat: submitted by Malaysia, is the name of a freshwater carp also known as Sultan fish. This tasty fish normally inhabits large rivers and is much sought after by gourmets. Typhoon Jelawat was a far-travelled typhoon which formed in the subtropical latitudes of the Western North Pacific a few hundred miles west of Wake Island and followed a long trajectory which eventually took it inland into China south of Shanghai by way of Okinawa. The storm peaked at 125 kts--just under super typhoon intensity. According to an alternate track for the early portion of Jelawat's history sent to me by Dr. Mark Lander of the University of Guam, the typhoon did briefly reach super typhoon status at 0000 UTC on 3 Aug. Another interesting aspect of Jelawat was the huge variation in eye diameter. Early in the storm's life, during its first rapid intensification spurt, the typhoon displayed a tiny, pinhole eye as small as 3 nm in diameter at one point. By the time Jelawat was menacing Okinawa, it had developed a huge eye approximately 90 nm in diameter. The earliest mention by JTWC of the pre-Jelawat disturbance was in a STWO issued at 2200 UTC on 31 Jul. An area of convection was rapidly developing about 720 nm east-southeast of Iwo Jima. A TRMM pass at 31/1908 UTC depicted tightly-curved convective banding. The area was given a Fair development potential. However, according to Mark Lander the initial development of a circulation had occurred on 30 Jul and several hundred miles to the east. Mark's alternate track locates a weak 25-kt system about 300 nm west-northwest of Wake Island at 0000 UTC on the 30th. The disturbance moved rather steadily to the west, reaching tropical storm intensity (per Mark's track) at 31/0000 UTC when located about 875 nm east-southeast of Iwo Jima. The first JTWC warning at 0000 UTC on 1 Aug placed the center about 675 nm east-southeast of Iwo Jima with 25-kt winds whereas Mark Lander's track estimates the MSW to be 55 kts at that time. JTWC upgraded the depression to a tropical storm with 45-kt winds on the second warning (at 0600 UTC)--in Mark's opinion typhoon intensity was reached about this time. JMA officially named the cyclone Jelawat at 1200 UTC, estimating the maximum 10-min avg wind to be 35 kts. JTWC's MSW estimate was still 45 kts at 1200 UTC but was 70 kts six hours later. Mark Lander's 1200 and 1800 UTC estimates were 85 kts and 105 kts. JMA did bring Jelawat to typhoon intensity by 02/0000 UTC. This wide divergence in intensity estimates among various analysts seems to be par for the course with small, midget tropical cyclones, especially those developing in subtropical latitudes. Last August there was one instance where simultaneous satellite current intensity estimates for Typhoon Tanya ranged from 25 kts to 65 kts. There was quite a bit of e-mail discussion about this subject which the author had access to, and one likely reason for these oft-seen discrepancies (suggested by Lori Chappel at Darwin) is that the Dvorak rules were developed for normal-sized tropical cyclones and that the minimum width criteria might not necessarily be valid for midgets. Also, these midgets are capable of intensifying and weakening at very rapid rates which break the normal Dvorak constraints for changes in cyclone intensity. Last January with Tropical Cyclone Iris in the South Pacific (another midget which intensified and then weakened very rapidly), there was a great divergence between Dvorak estimates based upon various methods. Current intensity numbers based upon visible imagery analysis reached 4.0 while simultaneously the IR-derived number was 6.0 (this from a JTWC Satellite Bulletin), and objective T-numbers averaged above 6.5 for a few hours, briefly reaching 6.9 at one point. Considering the extreme danger that such small, intense systems can present to any ship or island in their paths, further research and reconnaissance of midget tropical cyclones is something which is badly needed. To return to Typhoon Jelawat--guided by ridging to its north, the developing typhoon continued to move to the west-northwest. Jelawat continued to intensify, reaching its estimated peak intensity of 125 kts (per JTWC warnings) at 0600 UTC on 3 Aug when it was located about 160 nm east of Iwo Jima. (According to Mark Lander's track Jelawat's MSW peaked at 130 kts at 03/0000 UTC.) Deep convection surrounded a tiny eye only 3-nm in diameter. The typhoon was located beneath a diffluent region generated by a HIGH to the northwest and a TUTT to the southwest. At its peak gales covered an area slightly over 200 nm in diameter while the radius of 100-kt winds was estimated to be no greater than 20 nm. (The peak 10-min avg wind value assigned by JMA was 85 kts from 0000 through 1800 UTC on 3 Aug.) After reaching its peak intensity Typhoon Jelawat weakened somewhat with the MSW dropping to 100 kts by 03/1800 UTC. (Mark's estimate was down to 95 kts at this point.) The still-intense typhoon was centered only about 25 nm northeast of Iwo Jima at this time. Jelawat had been moving on a west-northwesterly track for a couple of days, but after passing Iwo Jima the storm began to move straight toward the west just north of the 26th parallel. After passing Iwo Jima Jelawat's intensity held fairly constant in the 90-95 kt range for about the next three days as it trekked toward Okinawa. (Mark's MSW estimate was slightly higher; JMA's 10-min avg estimate held at 80 kts.) A GMS-5 visible image taken at 04/0231 UTC (sent by Roger Edson) shows concentric eyewalls in Jelawat. Roger's comments indicate that it is very unusual to see this phenonmenon in small typhoons. As the typhoon continued westward the eye diameter began to steadily increase, jumping from 29 nm at 06/0000 UTC to 56 nm six hours later. Although the diameter of the eye fluctuated some, decreasing to around 40 nm by 1200 UTC, it had increased again to 64 nm by 07/0600 UTC and to a whopping 90 nm by 1200 UTC. Gales at this time extended outward 170 nm to the northeast of the center and 120 nm elsewhere while 50-kt winds covered an area about 110 nm in diameter. Jelawat's MSW had increased to 100 kts on the 6th as the eye contracted some (Mark Lander's track estimates 110 kts), but had dropped slightly to 90 kts by 07/1200 UTC. Between 0600 and 1200 UTC on 6 Aug Typhoon Jelawat passed over the Japanese island of Minamidaito-jima, located about 180 nm east of Okinawa. A report from Masashi Nagata, a forecaster at JMA's National Typhoon Center, states that the island reported a minimum pressure of 958.9 mb at 0720 UTC. The maximum 10-min avg wind of 75.4 kts was measured at 0920 UTC, and a peak gust of 119.5 kts was recorded at 0913 UTC. The report states that sea level pressure series at the island indicated very steep pressure gradients in a narrow ring just outside the large eye with almost uniform pressure within the eye. As Jelawat approached Okinawa on 7 Aug the storm turned to a north- westerly course which it followed for a couple of days. As early as 07/0000 UTC gale-force winds were being felt on Okinawa. Convective tops warmed some on 7 Aug and the MSW had dropped to 80 kts by the time Jelawat had moved over the northeastern end of the island around 1800 UTC. At 07/1200 UTC a ship approximately 100 nm to the southeast of the center was reporting 10-min mean south-southwest winds of 39 kts. By 1800 UTC the huge eye was positioned over the northeastern end of Okinawa and was tracking northwestward at 6 kts. Jelawat by this time had moved north of the upper-level ridge and into an area of increased vertical shear. The 50-kt wind radii had been adjusted outward to over 100 nm due to the enormous eye, and gales now covered an area exceeding 300 nm in diameter. Mark Lander passed along a report from the small island of Okinoerabu north of Okinawa. At 08/0100 UTC winds were from the southeast at 71 kts (presumably a 10-min avg), gusting to 92 kts with a pressure of 980 mb. Mark indicated that none of the wind reports from Okinawa had been all that high--Kadena had had a peak gust of 68 kts. A press report received by the author mentioned winds to 80 kts on Okinawa, but it is unknown whether this was measured or estimated nor whether it represents a sustained wind or a peak gust. Typhoon Jelawat continued to move slowly northwestward for a couple of days after passing Okinawa; then, as a mid-level ridge over China strengthened, the storm turned back to a westerly course which it followed until it made landfall in China. The intensity held steady at 80 kts (per JTWC's warnings) through 8 Aug, then increased to 90 kts for about 12 hours on the 9th as deep convection increased in the northeastern quadrant. (JMA's 10-min avg wind estimate had dropped to 70 kts on 8 Aug and did not reflect this short-term strengthening.) The eye diameter remained large but not quite as large as when the typhoon passed over Okinawa. As Jelawat approached the Chinese coast it encountered increasing vertical shear south of the subtropical ridge axis and slowly began to weaken. Landfall occurred near Shipu, about 130 nm south-southeast of Shanghai, around 1200 UTC on 10 Aug. The storm was a minimal typhoon with 65-kt winds per JTWC's warnings; JMA estimated the maximum 10-min mean winds near 50 kts at landfall. The weakening storm turned to the north-northwest after landfall and the final JTWC warning on Jelawat at 11/0600 UTC placed the dissipating depression's center about 150 km west of Shanghai. The effects of Typhoon Jelawat appeared to be minimal on Okinawa and in China. On Okinawa two children were injured, and 19,000 homes were without electrical power, mainly in the northern city of Nago. The author has located no reports of injuries or fatalities in China as a result of the typhoon. The primary impact seems to have been disruption of traffic and closure of airports and shipping ports. Jelawat's large eye was very unusual but not unprecedented. In August, 1996, Typhoon Kirk passed directly over Okinawa with an eye diameter in excess of 70 nm. The eye required 12 hours to pass over the island. Also, 1996 typhoons Orson and Violet had at some point in their evolution eye diameters on the order of 75 nm. However, all these eyes pale in comparison with that of Typhoon Carmen in 1960. Like Kirk and Jelawat, Typhoon Carmen passed over Okinawa. The Annual Typhoon Report for 1960 states that photographs taken from a radar at Kadena AB show quite clearly that, with respect to wall clouds surrounding the eye, Carmen's eye had a diameter of approximately 200 nm. (The information in this paragraph was taken from Eastern Hemisphere Tropical Cyclones of 1996 by Mark Lander, Eric Trehubenko and Chip Guard, although the author has somewhere in a box buried in a closet the typhoon summary for 1960 (as published in Climatological Data, National Summary) which describes Carmen's gargantuan eye.) Tropical Cyclone (TC-14W) 6 - 12 August -------------------------- This system was carried as a tropical depression by JTWC and JMA, but an alternate track provided by Dr. Mark Lander assigns tropical storm intensity to the system coincident with the time span it was in warning status from JTWC; hence, the designation above as a generic tropical cyclone. Mark's track begins a couple of days prior to the issuance of the first JTWC warning, locating a weak 20-kt circulation about 650 nm east-southeast of Iwo Jima--very near the location where Jelawat had reached typhoon intensity six days earlier. The system moved northwestward for a couple of days and slowly became better organized. During the morning of 7 Aug (local time) a MCS which had been associated with the disturbance collapsed, but a new one formed late in the day and covered the LLCC, as a visible image taken at 07/0730 UTC shows. Maximum winds were estimated near 30 kts at this time (per Mark's track). An enhanced infrared image taken also at 0730 UTC reveals the small cyclone with Typhoon Jelawat to the west and a monsoon depression (which developed into Typhoon Ewiniar) to the south-southeast. Early on 8 Aug the small system began to curve to the north. A visible image taken at 08/0130 UTC reveals a partially-exposed center due to some shearing. JTWC issued a Formation Alert about this time, followed by the first warning at 0600 UTC on the 8th. The depression was located about 300 nm north-northeast of Iwo Jima at this time and also roughly 850 nm east of Typhoon Jelawat. This location was essentially the point of recurvature for TC-14W--it did jog slightly to the north-northwest for awhile on 8 Aug, but soon turned to a north- northeasterly course and later to an east-northeasterly one. JTWC assigned a MSW of 25 kts while Mark Lander believes the disturbance reached minimal tropical storm intensity about this time. A visible image taken at 0630 UTC reveals a partially-exposed LLCC with deep convection sheared about 10 nm to the northeast. Animated water vapor imagery indicated that the system was located at the southern end of a mid-latitude trough. JTWC increased the MSW estimate to 30 kts at 1800 UTC where it remained for the duration of the period the system was in warning status. A visible image taken at 08/2330 UTC revealed that the center was either under or along the southeastern side of a tiny CDO with good banding evident. The 09/0000 UTC warning from JTWC reported that animated satellite imagery depicted weakening of the system although it mentioned that the LLCC had moved beneath the convection. The 0600 UTC warning noted that there had been an increase in convection during the previous six hours. Mark estimates that the peak intensity of 50 kts was reached at this time with fairly rapid weakening there- after. By 09/1200 UTC the cyclone was moving around the top of a ridge to its southeast and was approaching an old frontal boundary. The final JTWC warning at 10/0000 UTC indicated that the deep convection associated with the system was sheared over 60 nm to the east of the LLCC and the depression was deemed extratropical. Mark, however, contends that the system did not become extratropical but rather moved quickly to the east-northeast as a weak, sheared tropical cyclone. His track continues to follow the weakening system through 0000 UTC on 12 Aug at which time it was located about 1200 nm north of Wake Island. Typhoon Ewiniar (TC-15W / TY 0009) 9 - 19 August ----------------------------------- Ewiniar: submitted by the Federated States of Micronesia, is the name of a Chuuk traditional storm god In contrast with Typhoon Jelawat which formed near the northern edge of the tropics, was initially a midget cyclone, and intensified very rapidly, Typhoon Ewiniar formed deeper in the tropics, was initially very large in areal extent, and intensified more slowly. The storm formed near the Mariana Islands, moved generally northward to a point a few hundred miles south of Japan, then moved on an east-northeastward heading for several days, finally turning back to the north as it slowly weakened east of northern Honshu. The first mention of the disturbance from which Ewiniar developed was in a STWO issued by JTWC on 6 Aug. An area of convection had developed about 600 nm east of Guam. Convection was persistent but unorganized and there was an indication of a weak LLCC just to the west of some isolated deep convection. The disturbed area moved steadily to the west and by 8 Aug was centered approximately 150 nm east of Guam. The area had the appearance of a large monsoon depression with light winds near the center. A scatterometer pass indicated the possibility of multiple LLCCs. A Formation Alert was issued at 0251 UTC on 9 Aug, indicating that a large monsoon gyre had formed near the Mariana Islands. Upper-level analysis indicated that an anticyclone aloft was providing good outflow over the area. The first JTWC warning was issued on TD-15W at 09/0600 UTC with the depression's center roughly 200 nm west-northwest of Guam. The system intensified rather slowly initially under weak to moderate north- easterly vertical shear. JMA upgraded the depression to Tropical Storm Ewiniar at 1800 UTC, and JTWC followed suit at 10/0000 UTC when the center was located about 375 nm west of Saipan in the Marianas. Satellite intensity estimates were 30-35 kts and there were some synoptic ship reports of pressures to 996 mb. Organization had improved with some deep convection near the LLCC. At 10/0600 UTC a SSM/I pass depicted a partially-exposed, elongated LLCC with convection primarily south of the center. An ERS2 scatterometer pass indicated an extensive area of gale-force winds south and east of the system. By this time Ewiniar was moving to the north-northwest, a motion which continued for a couple of days due to the steering influence of a subtropical ridge to the east-northeast of the cyclone. The JTWC warning at 11/0000 UTC mentioned that some tight convective banding had appeared, and outflow in the northwestern sector was improving as an upper-level LOW to the northwest shifted to the south- west along the Ryukyu Island chain. The MSW was increased to 50 kts, and six hours later JTWC upgraded Ewiniar to a typhoon with 65-kt winds when it was located approximately 200 nm west-northwest of Iwo Jima. The storm was moving north-northwestward very quickly at 29 kts, but this rapid forward motion soon began to subside very quickly. At 12/0000 UTC Ewiniar's center had become fully-exposed and the system was downgraded to a 55-kt tropical storm. An upper-level LOW to the southwest and an upper-level HIGH to the southeast had combined to create an enhanced vertical shear environment. The 0000 UTC warning also required a relocation of the center to a point about 90 nm south of the previous forecast position. (This location was approximately 450 nm southwest of Tokyo.) By 0600 UTC the MSW had dropped further to 45 kts and Ewiniar was moving to the east-northeast at 4 kts. (It should be noted that JMA did not upgrade Ewiniar to a typhoon at this point, with the maximum 10-min avg winds being estimated at 55 kts. However, JMA maintained the 10-min avg winds at 50-55 kts during the next couple of days when JTWC had dropped their reported MSW value to 45 kts.) By 1200 UTC on 12 Aug the upper-level LOW to the southwest of the tropical storm had moved farther to the west and an anticyclone was building to the north of the system; thus, the vertical shear began to lessen. Satellite intensity estimates were primarily 35 kts on the 12th, but based upon some synoptic ship reports of winds to 40 kts, the MSW remained at 45 kts. For the next couple of days Ewiniar was steered by a subtropical ridge to its southeast on a fairly brisk east-northeasterly course, passing about 250 nm south of Tokyo around 0600 UTC on the 13th. A 13/1123 UTC TRMM pass indicated that a deep convective band had formed to the southeast of the system. By 14/1200 UTC the storm was still maintaining its 45-kt intensity (JMA's 10-min avg winds remained at 50 kts) and was forecast to soon begin weakening since it appeared to be moving into an increasingly unfavorable vertical shear environment. However, the warning for 14/1200 UTC did mention that a new band of deeper convection was developing in the northeastern quadrant. Then at 1800 UTC--surprise! Ewiniar had moved under a region with good upper-level diffluence and had developed a 13-nm diameter eye. So the storm was upgraded to a 65-kt typhoon centered about 390 nm east of Tokyo. Six hours later the MSW estimate was upped to 75 kts based upon satellite intensity estimates of 77 kts. The forward motion had slowed somewhat as a mid-latitude ridge to the northeast of the typhoon had strengthened some. At 15/0600 UTC JMA upgraded Ewiniar to a typhoon. The storm's ragged eye at this time was 20 nm in diameter, and had increased to 30 nm six hours later. Ewiniar's northeastward motion came to a halt around 1800 UTC, and by 16/0000 UTC the typhoon was moving northward at around 7 kts. Water vapor imagery indicated that southwesterly vertical shear was impinging on the storm and an eye was no longer visible in satellite imagery. Although outflow in the northern and northeastern sectors seemed to improve somewhat, the shearing took its toll and Ewiniar was downgraded to a tropical storm at 16/1200 UTC by both JTWC and JMA. The weakening storm became caught in a region of weak steering flow and drifted very slowly generally in a northerly direction for the next few days, remaining quasi-stationary at times. Animated satellite imagery revealed a tightly-wrapped LLCC with usually sparse weak convection. JTWC dropped the MSW to 40 kts at 16/1800 UTC and to minimal tropical storm intensity of 35 kts at 17/1200 UTC. JMA's 10-min mean winds were generally about 10 kts higher than JTWC's MSW during this period. The storm continued to slowly spin down due to the effects of moderate vertical shear and cooler SSTs. By 1200 UTC on the 18th there was no longer any convection associated with the system and JTWC downgraded Ewiniar to a depression as it drifted slowly northwestward southeast of Hokkaido. The final warning issued by JTWC at 19/0600 UTC indicated that based on upper-air analysis and satellite imagery, Ewiniar appeared to be an occluded LOW along a weak frontal boundary and hence extratropical, and was drifting southeastward. JMA still reported the winds at 40 kts, but downgraded the storm to 30 kts and issued their final bulletin at 1800 UTC. Super Typhoon Bilis (TC-18W / TY 0010 / Isang) 18 - 24 August ----------------------------------------------- Bilis: submitted by the Philippines, means speed or fleetness Bilis was the second typhoon of the year to reach the super typhoon threshold of 130 kts--the first being Super Typhoon Damrey in May. The storm began deep in the tropics well east of the Philippines, and after moving northward for a couple of days, took off on a remarkably straight course to the northwest which it followed until its final landfall in China. The first mention of the pre-Bilis disturbance was in a STWO issued by JTWC at 15/1700 UTC. An area of convection had developed about 500 nm south-southeast of Guam. Scatterometer data indicated a broad LLCC within a region of troughing and low vertical wind shear. Convection increased and the disturbance was given a development potential of Fair at 0000 UTC on 16 Aug. The area moved westward on the 16th and 17th and while convection was still fairly disorganized, animated satellite imagery showed good cross-equatorial flow into the monsoon trough. A Formation Alert was issued early on 18 Aug, and the first warning on TD-18W was issued at 0600 UTC placing the center approximately 75 nm west-northwest of Yap or about 500 nm southwest of Guam. Convection was beginning to organize rapidly around the circulation and the system was in a region of weak vertical shear. The depression initially moved northward, gradually turning to the north-northwest. Rapid intensification ensued and at 19/0000 UTC the system was abruptly upgraded from a 30-kt depression to a 55-kt tropical storm based upon satellite intensity estimates. TS-18W was then centered about 260 nm north-northwest of Yap or about 700 nm east of Catanduanes Island in the Philippines. (JMA upgraded the system and assigned the name Bilis at 0600 UTC.) By 0600 UTC Bilis was tracking to the north- west, guided by a mid-level ridge to its northeast. Based upon satellite intensity estimates, JTWC upgraded Bilis to a typhoon at 19/1200 UTC when it was centered roughly 550 nm east-northeast of Catanduanes Island. Animated water vapor imagery depicted good outflow aloft in all quadrants and shear was low as the young typhoon tracked northwestward at 14 kts. The MSW was increased to 75 kts at 20/0000 UTC and a banding eye had become visible by 0600 UTC. Typical of most NWP tropical cyclones originating in the monsoon trough deep in the tropics, Bilis was a fairly large storm in areal extent--gales covered a zone almost 250 nm in diameter. A 30-nm irregular eye was visible by 1200 UTC and the MSW was increased to 90 kts. (JMA at this point upgraded Bilis to a typhoon with 70-kt maximum 10-min avg winds.) Located in a very favorable environment and with good outflow, Bilis steadily increased in intensity, reaching super typhoon status at 1200 UTC on 21 Aug when located approximately 325 nm east-southeast of the southern tip of Taiwan. At this point Bilis exhibited a 25-nm round, cloud-free eye with an eye temperature of 17 C. Winds to 50 kts extended outward from the center over 60 nm to the northeast, and gales covered an area nearly 300 nm in diameter. (JMA's 10-min mean wind estimate had reached 100 kts by this time.) Super Typhoon Bilis reached its estimated peak MSW of 140 kts at 21/1800 UTC and maintained this intensity until it made landfall in Taiwan around 24 hours later. JMA's maximum 10-min mean wind estimate for the storm reached 110 kts at 22/0600 UTC with an estimated CP of 915 mb. Microwave imagery from a TRMM pass at 22/0347 UTC indicated possible concentric eyewalls. By 22/1200 UTC the center of Bilis had just about reached the coast of southeastern Taiwan. The eye diameter had shrunk to 12 nm, and while satellite intensity estimates were still 140 kts, the storm was beginning to show signs of weakening, likely due to some interaction with the mountainous island it was approaching. A 22/1009 UTC TRMM pass revealed concentric eyewalls with spiral convective bands wrapping into the center of the storm. Bilis made landfall along the south- east coast of Taiwan around 1500 UTC and by 1800 UTC was located over the west-central part of the island about 155 km south-southwest of Taipei. The typhoon was moving at around 15 kts as it made landfall and its forward motion increased to 18 kts while crossing Taiwan, so the center was over the island for no more than five or six hours. By 0000 UTC on the 23rd Bilis' center was well out into the Taiwan Strait about 155 nm west-southwest of Taipei. There was a significant discrepancy in MSW estimates between the warning agencies. JTWC's 1-min avg MSW was 120 kts while JMA's maximum 10-min avg wind estimate was 65 kts. PAGASA was in the middle with 80 kts, so that was the value I reported for 23/0000 UTC in the cyclone tracks file. Typhoon Bilis made landfall in China around 0300 UTC, and by 0600 UTC was inland about 175 km southwest of Fu Chau and weakening. Again, a large difference in the reported intensity between JTWC (90 kts) and JMA (45 kts). The weakening cyclone slowed in its forward motion and became very slow-moving in a region roughly 310 km west-southwest of Fu Chau. Around 1800 UTC some new convection developed over the north- east coast of Guangdong province well to the southeast of the low-level center, but this had weakened considerably six hours later. The final JTWC warning on Bilis at 24/0000 UTC, still reporting the MSW at 35 kts, placed the dissipating center about 330 km west of Fu Chau. (JMA had downgraded the system to a depression and issued their final bulletin at 23/1800 UTC.) The exact center of the eye of Typhoon Bilis crossed the Taiwanese coast just north of the surface weather station at Cheng-Kon (WMO 46761) located at 23.1N, 121.37E. The station was in the eye of the typhoon and recorded a minimum pressure of 931.2 mb, a maximum 10-min avg wind of 102 kts, and a peak gust of 152 kts. It should be noted that the anemometer was destroyed after the maximum wind was recorded. Another station, Lan-Yu (WMO 46762), located 117 km south-southeast of Cheng-Kon on an offshore island (22.03N, 121.53E), recorded a peak gust of 148 kts. The Lan-Yu station is located on top of a hill at about 300 m elevation. (These observations were taken from an e-mail by Chun-Chieh Wu, Associate Professor in the Department of Atmospheric Sciences at National Taiwan University.) Mark Lander performed an analysis of the Cheng-Kon wind measurements in order to estimate the 1-min avg MSW at landfall. Using a conversion factor of 1.15 to convert 10-min mean winds to 1-min mean winds yields a MSW of approximately 117 kts. Working from the 152-kt gust and using a gust-reduction factor of 1.22 (used by JTWC) yields a MSW of 125 kts, which is in reasonably close agreement. Just prior to landfall Mark reports that Dvorak T-numbers from the satellite reception ground station at the University of Guam were running around T6.5--127 kts, which is in excellent agreement with the MSW value derived from the peak recorded gust. Also, Roger Edson has pointed out that Cheng-Kon lay in the southern half of the eyewall, and given that the strongest winds in an intense typhoon are most likely to be occurring in the northern eyewall, and considering the fairly brisk forward motion, he feels that winds very possibly a good deal stronger than 125 kts could have occurred to the north of the station. In fact imagery from the Doppler radar at Green Island at 22/1130 UTC indicated that the strongest winds were occurring in the northern eyewall. (Incidentally, according to Mark Lander, the maximum wind gust ever recorded in a typhoon was 166 kts. This has happened twice: once at Miyako Jima (WMO 47927) near the eye of Typhoon Cora in September of 1966, and the other on the Taiwanese island of Lan-Yu near the eye of Typhoon Ryan in September, 1995.) According to press reports and other information available to the author, Bilis' death toll in Taiwan stands at eleven. Eight of these were orchard workers buried by a mudslide in the central county of Nantou. Eighty persons were reportedly injured, and as of Friday, 25 Aug, five persons were still missing--one of those being a doctor who was mountain climbing when the typhoon struck. To make matters worse, a 5.7-magnitude earthquake struck the island the day following the typhoon's arrival. About 250 houses collapsed in eastern Hualien county due to the strong winds of Bilis, and more than 600,000 homes lost electrical power in the typhoon. Rains from the storm flooded approximately 40,000 hectares of rice paddies and orchards, resulting in agricultural losses of $48 million. Typhoon Bilis was much weaker when it made its final landfall in China, but the storm still packed quite a punch. In the coastal city of Jinjiang near where the center came ashore, more than 300 farmers' homes were destroyed and a 100-m seawall collapsed. Bilis dropped 220 mm of rain on coastal Fujian province, resulting in flooding which sent several rivers 2-3 m above warning levels. In Quanzhou 2300 hectares of crops were destroyed while 310 houses collapsed and 200 businesses were affected. Damage estimates for the coastal city of Fuzhou were 500 million yuan, equivalent to 69 million U. S. dollars. There, 845 homes were destroyed and more than 4500 trees uprooted. A Beijing newspaper also reported that a tornado triggered by Typhoon Bilis struck villages near the city of Yueqing in the eastern Chinese province of Zhejiang on 23 Aug. This tornado reportedly destroyed more than 50 homes and damaged 612 buildings. There were 16 serious injuries and the damage was estimated at $443,000. Finally, Patrick Hoareau sent me some information on the heavy rains which fell on South Korea in late August. Kunsan, in the southwestern portion of the country, recorded a 48-hour total of 468 mm ending at 0000 UTC on 27 Aug. The monthly average for Kunsan is 150 mm. Patrick notes that, based on some animation of GMS satellite imagery, it appears that the Korean rains resulted from the remnants of Typhoon Bilis. Tropical Storm Kaemi (TC-19W / TS 0011) 20 - 23 August ---------------------------------------- Kaemi: submitted by the Republic of Korea (South Korea), means ant. The ant often appears in Korean fairy tales as a symbol of diligence. Tropical Storm Kaemi was typical of many systems which form in the South China Sea (the "Gulf of Mexico" of the North Pacific), developing off the coast of Vietnam and moving westward and inland without significant intensification. An area of convection developed on 17 Aug off the southeast coast of Vietnam in association with the monsoon trough extending from Vietnam into the Philippine Sea. No LLCC was evident and the area was under moderate to strong vertical wind shear. By the 18th a weak, elongated LLCC had become evident in synoptic data and a QuikScat pass. A steady pressure drop and increasing convective organization led to the issuance of a Formation Alert at 19/0000 UTC. The area was embedded in the monsoon trough with linear convergent flow south of 11N. A 18/2201 UTC QuikScat pass revealed 20-30 kt winds around the periphery of the LOW with winds of only 10-15 kts near the center; hence, the system had the characteristics of a monsoon depression. By early on the 20th the system had shown improved organization with convection developing around the LLCC, so JTWC initiated warnings on TD-19W at 0600 UTC. The center of the 25-kt depression was located about 330 nm east-southeast of Da Nang and was tracking west- northwestward at 5 kts. The system was still under moderate vertical shear with outflow fair to the west and northeast of the LLCC. Guided by a subtropical ridge to the north, TD-19W moved generally in a northwesterly direction for most of its life, becoming more west- northwesterly as it approached the coast of Vietnam. By 21/0600 UTC deep convection, which had been most persistent west of the center, had increased in organization with a spiral band west of the LLCC wrapping into the southern quadrant; therefore, the MSW was increased to 30 kts. The system at this stage was still a large monsoon depression with the 1000-mb isobar about 500 nm in diameter. The 21/0600 UTC synoptic reports from sites WMO 59985 and 59981, located about 2 degrees north of the LLCC, indicated sustained winds of 20-25 kts with a pressure of 993 mb. By 1200 UTC on the 21st the system was centered approximately 170 nm east-southeast of Da Nang. Spiral bands over land west of the center had weakened, but convection west and south of the center had intensified and increased in areal extent. Synoptic ship reports approximately 40-50 nm southeast of the LLCC indicated sustained winds of 37 kts; hence, the depression was upgraded to a tropical storm. JMA also upgraded the system to a tropical storm at 1200 UTC, naming it Kaemi. Tropical Storm Kaemi reached its estimated peak intensity of 45 kts (from both JTWC and JMA) at 22/0000 UTC based upon satellite intensity analysis. The system was very near the coast of Vietnam about 60 nm east-southeast of Da Nang, and satellite imagery revealed an area of deep convection sheared to the southwest of a partially- exposed center. A 200-mb analysis indicated fair diffluence aloft with moderate vertical wind shear over the system. Kaemi made landfall at its peak intensity of 45 kts around 0600 UTC on 22 Aug with the associated deep convection sheared well to the south of an exposed LLCC. By 1200 UTC the system was inland about 55 km west of Hue, Vietnam, and dissipating. The final warning from JTWC at 23/0000 UTC placed the diffuse LLCC roughly 150 km west of Hue, just north of the Bolaven Plateau of southern Laos. The passage of Kaemi brought some fairly heavy rains to Vietnam. Da Nang recorded 140 mm in the 24 hours ending at 21/1200 UTC, and Hue measured 89 mm for a 24-hour total from 21/0000 UTC through 22/0000 UTC--more than half the monthly average of 137 mm. Various media sources reported that Kaemi brought more than 300 mm of rain to central Vietnam which caused extensive flooding. Three persons died in Quang Ngai province, one death occurred in the province of Danang, and another fatality was reported in the Thua Thien Hue province where over 12,000 hectares of rice fields were inundated. Also, according to media sources, a whirlwind moving across the Halong Bay on the 22nd sank two wooden boats carrying tourists, resulting in two deaths with three persons reported as missing. (I assume that "whirlwind" here more than likely means a waterspout or small tornado which was possibly related to the passage of Tropical Storm Kaemi.) On 24 Aug there were reports of more whirlwinds and a tornado affecting the southern provinces, primarily in the Mekong River delta region, with significant damage and injuries; however, these local storms would likely not have been directly attributable to the tropical storm. According to the Bangkok Post newspaper, the remnants of Kaemi were responsible for heavy rains which caused flash floods in northeastern Thailand. Two residents of the village of Surin were drowned, and there were also two fatalities in the Trat province. Substantial amounts of farmland, fruit orchards, and rubber plantations were ruined or damaged. In the Warin Chamrap district four homes were reportedly swept away while ten other homes and two temples were damaged. In the Trat province the total damage estimate was placed at 179,355,460 baht (Thai unit of currency). Typhoon Prapiroon (TC-20W / TY 0012 / Lusing) 25 August - 1 September ---------------------------------------------- Prapiroon: submitted by Thailand, is the Thai god of rain An area of convection was noted deep in the tropics southeast of Guam on 22 Aug. The area was within a region of weak vertical shear, and a partial QuikScat pass revealed a possible broad LLCC located in the monsoon trough. The disturbed area migrated westward and was assigned a Fair development potential on 23 Aug. By early on the 24th the disturbance was located approximately 70 nm southeast of Yap. Animated visible satellite imagery indicated a fully-exposed LLCC with deep convection to the northwest. Synoptic observations from Yap indicated a 24-hour pressure fall of 3 mb with east-southeasterly sustained winds of 10 kts. A Formation Alert was issued at 2230 UTC, upgrading the development potential to Good. On 25 Aug satellite imagery indicated improved organization with deep convection located primarily to the north of the center. PAGASA initiated advisories on the system at 25/0000 UTC, naming it Tropical Depression Lusing. Lusing's center was located about 600 nm east- southeast of Catanduanes Island, or roughly 450 nm west-northwest of Yap, with maximum winds (10-min avg) estimated at 30 kts. JMA also began referring to the system as a tropical depression at 0600 UTC. There was a fair amount of difference between PAGASA's and JMA's coordinates (as might be expected for a weak, diffuse system), but Lusing generally moved slowly northward. JTWC issued a second Formation Alert at 2230 UTC followed by the first warning on TD-20W at 26/0000 UTC. Lusing/20W was centered approximately 400 nm east of Catanduanes Island at that time, moving northward at around 9 kts, but shortly thereafter underwent a significant acceleration to the north-northeast. At 26/1200 UTC Lusing was moving north-northeastward at 23 kts from a location about 560 nm east-northeast of Port San Vicente in the Philippines. The JTWC warning indicated that the rapid motion might be due in part to convection reconsolidating over a new LLCC. By 1800 UTC Lusing/20W was located roughly 375 nm south-southeast of Okinawa, and the system's motion had turned to the north-northwest, but still at a fairly fast clip of 20 kts. Based on satellite intensity estimates and a ship report of 35 kts, the depression was upgraded to a tropical storm at 1800 UTC and named Prapiroon by JMA. Deep convection was sheared approximately 15 nm to the north and east of a partially- exposed LLCC. Over the next 24 hours Prapiroon's forward motion slowed considerably and the storm turned increasingly toward the west- northwest and west as a subtropical ridge to the north strengthened. Vertical shear was weak and outflow aloft favorable, but the storm remained rather disorganized with multiple LLCCs evident. Early on the 27th satellite imagery revealed a large band of convection extending approximately 650 nm to the southeast of the center, but by 1800 UTC this feature was weakening and Prapiroon's center had separated from the band. At 0000 UTC on 28 Aug the storm was centered roughly 175 nm south of Okinawa and moving westward at 8 kts. JTWC's MSW estimate was still 35 kts based upon satellite intensity estimates of 30 kts and synoptic reports of 35 kts, and the remarks indicated that animated visible and infrared imagery still showed a broad circulation with smaller multiple LLCCs orbiting the center of the broader circulation. Interestingly, JMA had increased their maximum 10-min avg wind estimate to 50 kts at 27/1800 UTC. By 28/1800 UTC Prapiroon was moving slowly westward from a position about 240 nm east-southeast of Taipei. The MSW (per JTWC) had increased to 45 kts and convection was consolidating over the LLCC. At 0000 UTC on the 29th JTWC increased the MSW further to 55 kts. A SSM/I pass at 28/2202 UTC depicted a partially-exposed LLCC at the northern edge of some deep convection. Water vapor imagery indicated an upper-level HIGH to the northwest of the storm, resulting in poor outflow, and CIMSS shear analysis charts indicated weak to moderate northeasterly shear over Prapiroon. The tropical cyclone began to turn more to the northwest on 29 Aug and slowly became better organized. A SSM/I pass at 29/0919 UTC showed a convective band wrapping in toward the LLCC from the south- west, and 200-mb analysis indicated that an upper-level HIGH had developed over the system. The JTWC warning at 29/1200 UTC noted that a ship located approximately 20 nm east of the center had reported south-southeast winds of 50 kts (10-min avg) with a pressure of 981 mb. JTWC upgraded Prapiroon to a 65-kt typhoon at 1800 UTC when the storm was centered about 140 nm east-northeast of Taipei, moving northwest- ward at 10 kts. The MSW was increased to 70 kts six hours later as a 29/2149 UTC SSM/I pass revealed a ragged eye 50 nm in diameter with convective banding in the eastern semicircle. By 0600 UTC on the 30th Typhoon Prapiroon was located 330 nm south- southwest of Cheju Do (off the coast of Korea) and moving northward at 17 kts. A mid-level ridge to the north was weakening and a major short wave trough was approaching from the west. JMA upgraded the storm to a typhoon at 0600 UTC while JTWC upped their MSW estimate to the peak value for the storm's history, 75 kts, which was maintained for 18 hrs. (JMA's peak 10-min avg wind estimate for Prapiroon was 70 kts at 0000 and 0600 UTC on 31 Aug.) SSM/I imagery at 30/0907 UTC indicated a 33-nm irregular eye with an impressive banding feature about 180 nm southeast of the LLCC. A ship transiting beneath the convective band reported 10-min sustained winds of 45 kts from the south-southeast and 7.3 m swells. At 1800 UTC Prapiroon was located about 100 nm east of Shanghai and still tracking northward. Around 31/0000 UTC the typhoon passed about 120 nm west of Cheju Do, and the forward speed had increased to 19 kts. Also, convection was beginning to weaken and the MSW was lowered to 70 kts. (JMA, however, increased their maximum 10-min avg wind estimate from 65 to 70 kts at this point.) Based on a synoptic report of 54-kt winds from Kosan-Ni (WMO 47185), located on Cheju Do, the 50-kt wind radius was increased to 85 nm in the eastern semicircle. Gales at this time covered an area approximately 275 nm in diameter. By 1200 UTC on the 31st Prapiroon was located about 90 nm west of Seoul, South Korea, and was tracking north-northeastward at 22 kts. The storm had weakened to a minimal typhoon with 65-kt winds (JMA had downgraded it to a 60-kt tropical storm), and the convection associated with the system was beginning to elongate to the north-northeast. CIMSS wind shear charts showed moderate vertical shear over the storm with increasing shear to the north. Typhoon Prapiroon made landfall in southwestern North Korea shortly after 1200 UTC, and by 1800 UTC was weakening inland in North Korea about 115 nm north of Seoul. Hamheung (WMO 47401), North Korea, reported 30-kt winds at 31/1800 UTC. The storm continued to track rapidly northeastward across the Korean peninsula, moving up the coastal region of northeastern North Korea, and by 0600 UTC on 1 Sep had entered the Sea of Japan just south of Vladivostok, Russia. Animated infrared satellite imagery showed little deep convection, and no low-level circulation was evident as the system continued to transition into an extratropical LOW. This was the final warning by JTWC, but JMA issued a couple more bulletins, finally deeming Prapiroon extratropical at 1800 UTC as it moved eastward across the Sea of Japan. Although the center of Typhoon Prapiroon remained offshore as it swept past eastern China, the storm brought heavy rains to some areas. The official Chinese news agency (Xinhua) reported that at least four persons were killed and 80 injured in the northern part of Jiangsu province. Heavy rainfall resulted in the flooding of 363,933 hectares of farmland, and over 7500 houses were destroyed. Xiangshui county reportedly received 821 mm of precipitation--a record in the history of Jiangsu province. Rainfall amounts in the Huangpu district and Chongming county were reported as 85 mm and 79 mm, respectively. While Prapiroon's center made landfall in North Korea, it was near enough to South Korea to cause winds of up to 58 kts in the coastal region of the country. At least four persons were reported dead with another 21 missing following the storm. Two boats were sunk and many power lines downed. In the southern part of the nation a dike ruptured, flooding residential areas and farmland with approximately 300 persons left homeless. In North Korea flooding in the north- eastern sections of the country seems to have been the major impact caused by the storm. The death toll stands at 42 with most of these occurring in the northeastern province of South Hamgyong when rivers burst their banks. There was also extensive damage to crops, and road and rail connections were disrupted along the east coast. The region's major city, Chongjin, was flooded with up to about a metre of water in some sections. Earlier, while Prapiroon was passing through the Ryukyu Islands, Ishigakishima recorded 112 mm of rain in the twelve hours ending at 0000 UTC on 30 Aug. The monthly average for the station is 236 mm. (Thanks to Patrick Hoareau for sending me this tidbit of information.) Tropical Storm Maria (TC-21W / TS 0013) 28 August - 1 September ---------------------------------------- Maria: submitted by the United States, is the Latin/Hispanic form of Mary and is popular as a Chamorro woman's name A STWO issued by JTWC at 2200 UTC on 26 Aug mentioned that an area of convection had developed approximately 45 nm east-southeast of Hong Kong. The convection was disorganized and was located in an area of moderate northerly vertical shear. Synoptic analysis indicated that a LLCC was located just inland and north of the convection. By 1200 UTC on the 27th animated satellite imagery revealed increasing organization of the cloud mass and decreasing vertical shear. The development potential of the system was upgraded to Fair. (More than likely a Formation Alert was issued by JTWC, but I could not locate one anywhere in my files.) The first warning on TD-21W was issued by JTWC at 0000 UTC on 28 Aug with the center located approximately 100 nm east-southeast of Hong Kong. A small area of deep convection was organizing around a LLCC, shear was weak to moderate, and a 200-mb analysis showed an upper-level ridge axis just northwest of the system. The depression commenced a slow, generally southward, drift and gradually increased in organization. By 28/1200 UTC the system was centered about 150 nm south-southeast of Hong Kong and was upgraded to Tropical Storm Maria by JMA; JTWC, however, was still estimating the MSW at 25 kts. JTWC did upgrade the system to a tropical storm six hours later, based primarily on synoptic data. Maria displayed a partially-exposed LLCC with the convection to the south. Over the next 24 hours Maria continued to drift somewhat erratically to the south and southeast. Convection gradually increased near the LLCC, and JTWC bumped the MSW up to 45 kts at 30/0000 UTC when the storm was centered approximately 300 nm south-southeast of Hong Kong. A subtropical ridge was north of the system, and northeasterly flow over the storm caused most of the convection to be sheared to the southwest quadrant of the circulation. Maria's track, when plotted, looks like a slightly distorted "figure 8". The low- and mid-level steering flow in the northern South China Sea was being influenced by Typhoon Prapiroon to the northeast. The position at 30/0000 UTC was the southernmost point in Maria's track. By 0600 UTC the storm had moved very slightly to the west-northwest, then abruptly changed course and began tracking slowly to the northeast and later north-northwest, completing the right side of the "figure 8". The MSW was decreased back to minimal tropical storm intensity of 35 kts at 0600 UTC when animated visible imagery depicted the LLCC tucked in underneath the northeastern portion of the convection but also revealed another weaker, fully-exposed circulation about 60 nm to the northeast. Convection was seen to weaken somewhat over the next few hours but had begun to increase again by 1800 UTC. A TRMM pass at 30/2312 UTC located the center of the cyclone under the deep convection and also showed that an area of deep convection had formed on the southeastern side of the LLCC. The MSW was increased to 50 kts at 31/0600 UTC based on satellite intensity estimates of 45 and 55 kts. A 31/0155 UTC microwave image placed the center under the deep convection and also indicated a tight band of strong convection wrapping into the center from the south. Maria was located about 120 nm southeast of Hong Kong and was moving northward toward the southern coast of China at around 7 kts. By 1200 UTC deep convection was evident in all quadrants, and the storm reached its peak intensity of 55 kts at this time. (JMA's maximum 10-min avg wind estimate remained at 35 kts until just before the cyclone made landfall in China when it was increased to 40 kts.) At 1800 UTC on the 31st Tropical Storm Maria was located just off the Chinese coast about 50 nm east of Hong Kong, moving north-northwestward at 9 kts. The storm had maintained its intensity, and a 31/1309 UTC SSM/I pass indicated deep convection in a band circling the LLCC in all quadrants except the eastern. Maria made landfall shortly after 1800 UTC on 31 Aug and continued to track to the northwest after moving inland. The final JTWC warning at 0600 UTC on 1 Sep placed the center of the dissipating storm about 200 km north-northwest of Hong Kong. The only report of fatalities available to the author indicated that more than 50 persons lost their lives from Tropical Storm Maria, presumably in southern China and due to flooding caused by storm-related rainfall. (This information was gleaned from the Australian Broadcasting Corporation's website by Matthew Saxby, who forwarded it to me. Thanks to Matthew for sending me this information.) *********************************************************************** NORTH INDIAN OCEAN (NIO) - Bay of Bengal and Arabian Sea Activity for August: 1 tropical depression ** ** - system was briefly mentioned as a depression by IMD only North Indian Ocean Tropical Activity for August ----------------------------------------------- During the latter part of August several LOWs formed in the monsoon trough in the Bay of Bengal and were mentioned by the IMD. One of these was classified as a depression on 23 Aug. An upper-level cyclonic circulation was noted on 19 Aug over the Bay off the South Orissa coast. This feature persisted and by 21 Aug was considered a well-marked low-pressure area. Around 0600 UTC on 22 Aug the center was near 16.5N, 85.5E, or about 150 nm east-southeast of Visakhapatnam. By 23/0300 UTC the LOW had been classified as a depression and was located near 16.5N, 83.5E, or about 80 nm south-southeast of Visak- hapatnam, moving slowly in a west-northwesterly direction. Nine hours later (1200 UTC) the depression's center lay near 16.5N, 82.5E, or about 55 nm southeast of Kakinada. The system continued to move generally westward and moved inland along the North Andhra coast without reaching cyclonic storm intensity. (No warnings on this system were issued by JTWC; in fact, I could find no TWOs which I'd saved for the period in question, which strongly suggests that the disturbance was not considered significant enough to be mentioned in JTWC's daily TWOs for the North Indian Ocean.) I did not include a tabular track for this depression in the cyclone tracks file, although in hindsight I rather wish that I had; hence, the reason for including the coordinates in the narrative. In IMD terminology, a depression usually implies maximum winds of 25 kts, while a 30-kt system is referred to as a deep depression. *********************************************************************** SOUTHWEST INDIAN OCEAN (SWI) - South Indian Ocean West of Longitude 90E Activity for August: 1 tropical disturbance ** ** - this system was treated as a minimal tropical storm by JTWC The primary sources of information upon which the narrative is based are the warnings issued by the TCWC on La Reunion Island, associated with Meteo France, and which is the RSMC for the Southwest Indian Ocean basin. However, cyclones in this region are named by the sub-regional centres on Mauritius and Madagascar with longitude 55E as the dividing line between their respective areas. La Reunion only advises these centres regarding the intensity of tropical systems. References to sustained winds should be understood as implying a 10-min averaging period unless otherwise stated. In the accompanying tracks file some position comparisons have been made with JTWC's positions, and warnings from JTWC were used as a source of 1-min avg MSW estimates. Also, the comments about satellite imagery and other sources of data such as SSM/I and TRMM were obtained for the most part from the JTWC warnings. Southwest Indian Ocean Activity for August ------------------------------------------ Weaker tropical depressions or disturbances are not all that uncommon in the South Indian Ocean during the winter months. A tropical depression occurred in that area in July, 1998, and there were a couple of systems the following September. In early August of this year a system gained enough organization that the La Reunion TCWC issued bulletins for a couple of days. JTWC also issued warnings on the disturbance. A report is included on Tropical Disturbance #1 (TC-01S per JTWC's nomenclature) . For naming purposes the tropical cyclone season in the Southwest Indian Ocean runs from 1 August through 31 July. Names for the 2000-2001 tropical cyclone season are: Ando Jakoba Suzy Bindu Kiran Tovo Charly Lanto Ursula Dera Mathieu Vimla Evariste Nancy Wenda Francois Oda Xino Gaby Premnath Yul Hans Quirin Zoe Idriss Rakoto Tropical Disturbance (TC-01S / MFR #1) 1 - 3 August --------------------------------------- An area of convection had developed by early on 1 Aug about 400 nm east-northeast of Diego Garcia. Satellite imagery depicted an increase in areal coverage of deep convection and a partially-exposed LLCC east of the convection. A QuikScat swath indicated stronger winds over the western half of the system which was moving slowly westward beneath 30-40 kt easterlies. The La Reunion TCWC (MFR) began issuing bulletins on the system as Tropical Disturbance #1 at 01/0600 UTC. The maximum central 10-min avg winds were estimated at 25 kts, but localized winds to gale force were forecast in the southern quadrant due to the pressure gradient with the subtropical HIGH to the south. By 1800 UTC the system's organization had improved to the point that JTWC began issuing warnings. A 01/1313 UTC QuikScat pass indicated 20-30 kt winds on the western side of the system, so the initial warning intensity was set at 35 kts--minimal tropical storm intensity. The system was centered about 320 nm east-northeast of Diego Garcia at the time. Located north of a subtropical ridge, the disturbance moved slowly in a general westerly direction through 02/1200 UTC, then began to track southwestward at an increased pace as a mid-latitude trough created a weakness in the steering ridge. Satellite intensity estimates reached 45 kts from at least one agency, but the system remained in a moderate vertical shear environment throughout its life and JTWC maintained 35 kts as the warning intensity. MFR's bulletins continued to report the maximum 10-min avg wind as 25 kts near the center, though as indicated earlier, some stronger winds were forecast for the southern periphery of the system. The final JTWC warning, issued at 1200 UTC on 3 Aug, indicated that a LLCC could not be found in satellite imagery, so the system was downgraded. MFR also issued their final bulletin at the same time. While MFR classified the system no higher than a tropical disturbance and JTWC treated it as a minimal tropical storm, in reality this does not represent as great a difference as may be supposed. It is simply due to a delta of one-half T-number on the Dvorak scale in combination with the difference between wind averaging periods. While satellite intensity estimates available to JTWC ranged from T2.0 (30 kts) to T3.0 (45 kts), the various forecasters utilized an average value of T2.5; i.e., 35 kts in terms of a 1-min avg MSW. Bulletins from MFR clearly state that they used a Dvorak rating of T2.0 for the disturbance, equivalent to a 26-kt 10-min mean wind. MFR procedures require that a system have 10-min mean winds exceeding 28 kts in order to be classified as a tropical depression, so that category is reserved for systems to which the agency assigns a T2.5 rating on the Dvorak scale. Out-of-season tropical cyclones are not common in the South Indian Ocean, but moreso than in the Atlantic basin. A tropical depression formed near 85E in August of 1996, and tropical depressions formed in July in both 1997 and 1998. (This is per MFR's classification-- JTWC carried these systems as tropical storms.) Based upon JTWC's database, August is the least active month for the Southern Hemisphere as a whole with only two tropical cyclones (34 kts or higher) forming over the past twenty seasons. During the 1996-1997 tropical cyclone season in the Southern Hemisphere a storm formed in every month, beginning with Lindsay in July in the Perth AOR and ending with Keli the following June in the South Pacific. It should be pointed out that JTWC, Australia and Fiji consider the tropical cyclone year to run from 1 July through the following 30 June, whereas La Reunion defines their season as 1 August through the following 31 July. In fact Philippe Caroff, Chief Forecaster at La Reunion, indicated in an e-mail that they were following the system described above on 31 July, but since it was weak, postponed initiating advisories until the date rolled over so it would be the first numbered disturbance of the new cyclone season. (Some of the above information was taken from tables of Southern Hemisphere tropical cyclone seasonal statistics compiled by Patrick Hoareau and based upon JTWC's database.) *********************************************************************** 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 OCEAN (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 July, 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 in the following manner: (a) FTP to: hrd-type42.nhc.noaa.gov [140.90.176.206] (b) Login as: anonymous (c) For a password use your e-mail address (d) Go to "data" subdirectory (Type: cd data) (e) Set file type to ASCII (Type: ascii) (f) Transfer file (Type: get remote_file_name local_file_name ) (The files will be named with an obvious nomenclature--using August as an example: aug00.tracks) (g) To exit FTP, type: quit 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. If anyone wishes to retrieve any of the previous summaries, they may be downloaded from the aforementioned FTP site at HRD. The summary files are catalogued with the nomenclature: aug00.sum, for example. Back issues can also be obtained from the following websites (courtesy of Michael Bath, Michael V. Padua, Tom Berg, Michael Pitt, and Rich Henning): http://australiasevereweather.com/cyclones/ http://www.typhoon2000.com OR http://www.geocities.com/taifun00/ http://www.hurricanealley.net/ http://www.qisfl.net/home/hurricanemike http://www.met.fsu.edu/gsc/Docs/Grads/henning/cyclones/ 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 I have discovered that JTWC now has available on its website the complete Annual Tropical Cyclone Report (ATCR) for 1999 (1998-1999 season for the Southern Hemisphere). Also, 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 1999 Atlantic and Eastern North Pacific tropical cyclones; also, preliminary storm reports for all the 1999 Atlantic and Eastern North Pacific cyclones are now available. The URL is: http://www.nhc.noaa.gov Prepared by: Gary Padgett E-mail: garyp@alaweb.com Phone: 334-222-5327 (nights & weekends) / 850-882-2594 (weekdays) *********************************************************************** ***********************************************************************