Monthly Global Tropical Cyclone Summary September 1999 (original) (raw)


             MONTHLY GLOBAL TROPICAL CYCLONE SUMMARY

                         SEPTEMBER, 1999

(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.)


                      SEPTEMBER HIGHLIGHTS

--> Two more Category 4 hurricanes roam the North Atlantic --> Severe hurricane damages Bahamas and causes extremely devastating floods in North Carolina --> First Northwest Pacific super typhoon of year damages Okinawa and Japan


                       ACTIVITY BY BASINS

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

Activity for September: 2 tropical depressions (a) 2 tropical storms 1 possible tropical storm (b) 2 hurricanes

(a) - One of these systems was not carried operationally as a tropical depression. Dr. Jack Beven of TPC/NHC has indicated that it will likely be treated as an unnumbered depression in the annual Atlantic Tropical Systems article prepared by the staff of NHC.

(b) - This system will possibly be written up as an unnamed tropical storm. Jack Beven is currently seeking surface observations from the vicinity of the depression to see if a strong case can be made for including it as an official tropical storm. If not, it too will likely be included as an unnumbered tropical depression.

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. A special thanks to John Wallace, a tropical cyclone enthusiast and college student from San Antonio, for providing me with a log which he had kept of all Atlantic/Northeast Pacific tropical waves that proved to be very valuable in helping to trace the pre- depression history of some of the cyclones.

                  Atlantic Activity for September
                  -------------------------------

 The Atlantic tropics continued in an active mode during September.

Two large, severe Category 4 hurricanes (Floyd and Gert) formed while Hurricane Dennis was active during the first week of the month. One additional tropical storm, Harvey, formed in the Gulf of Mexico while three tropical depressions developed. Two of these were not treated operationally as tropical depressions but will likely be considered as unnumbered depressions in the annual Atlantic Tropical Systems article prepared by the staff of TPC/NHC. One of these systems will possibly be written up as an unnamed tropical storm if Jack Beven can gather enough data to establish that it contained surface winds of tropical storm intensity. In accordance with the procedure I established last month, I have dubbed these depressions "Delta" and "Epsilon" in order to reduce confusion in any discussion of the systems in this summary.

 Hurricane Floyd flirted with Category 5 intensity and caused severe

damage on some of the northwestern Bahama Islands. Although it had weakened to a strong Category 2 hurricane by the time it made landfall in North Carolina, very heavy rainfall over the state led to extensive widespread flooding which turned out to be the state's costliest natural disaster to date. Hurricane Gert remained well out in the Atlantic, brushing Bermuda and Newfoundland's Avalon Peninsula as it swept northeastward.

 Tropical Depression #4 formed in the extreme southwestern Gulf of

Mexico on 5 Sep about 75 nm southeast of Tampico, Mexico. The system drifted northward, then moved slowly northwestward and crossed the Mexican coastline about 70 nm north of Tampico in southern Tamaulipas state with 30-kt MSW. Satellite and radar imagery suggested that the depression was close to tropical storm intensity as it made landfall but there were no observations of tropical storm-force winds. The center remained quasi-stationary just after it had moved inland for about 12-18 hrs before resuming a northwestward drift. After the center had made landfall there was a large convective band present over the Gulf east of the center and Doppler radar indicated winds of tropical storm intensity aloft in the band between 1500 and 2100 m elevation. The depression subsequently moved farther inland and weakened, moving into southern Nuevo Leon state by 1200 UTC on 7 Sep. Heavy rainfall with rates up to 75 mm per hour were reported over a large area of Mexico, but the author has not received any reports regarding possible casualties or damage resulting from the rains.

                     Hurricane Dennis  (TC #5)
                      24 August - 7 September
                     -------------------------

 Although forming in August, Hurricane Dennis was still on the charts

off the North Carolina coast in early September. The hurricane was downgraded to a tropical storm at 0000 UTC on 1 Sep and made landfall on the North Carolina coast during the afternoon of 4 Sep. The remnants subsequently moved northward to the vicinity of Lake Ontario by 7 Sep. For the full write-up of Hurricane Dennis, see the Global Tropical Cyclone Summary for August, 1999.

                      Hurricane Floyd  (TC #8)
                          7 - 17 September
                      ------------------------

 A tropical wave left the coast of west Africa on 2 Sep.    As the

system continued westward across the Atlantic it gradually became better organized. By 7 Sep a broad circulation had developed with sufficient organized convection, so advisories were initiated on Tropical Depression #8, located at 1800 UTC about 900 nm east of the island of Martinique. The depression was upgraded to Tropical Storm Floyd only 12 hours later when Dvorak ratings from TAFB and SAB had reached T2.5--35 kts. Banding features with very cold cloud tops were wrapping halfway around the LLCC and there was a very impressive upper-tropospheric outflow pattern associated with the storm. Floyd initially was moving on a west-northwesterly track which became more northwesterly after a couple of days and paralleled the Leeward Island chain, keeping the storm well north of the islands.

 Even though Floyd appeared quite well-organized in satellite

imagery, it was somewhat slow to intensify. On the morning of 8 Sep position estimates based on visible pictures were 90 nm apart, and the low- and mid-level circulation centers were not well-linked at the time. Floyd's MSW had been increased to 50 kts in the 09/0300 UTC advisory, but when the first flight by the U. S. Air Force Reserves 53rd Weather Reconnaissance Squadron (Hurricane Hunters) reached the storm very early on the 9th, they found that the storm was not quite as strong as had been previously thought. Interestingly, by mid- morning of the 9th Dvorak estimates had reached 65 kts but Floyd was not upgraded to a hurricane until 1500 UTC on 10 Sep due to the findings of the Hurricane Hunters. A reconnaissance flight during the afternoon (9 Sep) fixed the center somewhat northwest of the previous satellite estimates and removed from the deep convection; yet, a FLW of 68 kts was found in the northeast quadrant. Based on this information, the official MSW was increased to 60 kts.

 Floyd's appearance in infrared imagery was very impressive with

excellent outflow, and Dvorak numbers continued to indicate hurricane intensity, but several passes through the storm by a reconnaissance plane indicated that the LLCC was decoupled from the mid-level rotation observed in infrared imagery. The reasons for Floyd's slowness to intensify are not altogether clear. Early on the morning of 10 Sep the LLCC seemed to jump northward a bit into a mass of deep convection, and a flight by the Hurricane Hunters into the storm later in the morning found FLW of 88 kts with an attendant CP of 985 mb; therefore, Floyd was upgraded to a 70-kt hurricane at 1500 UTC centered about 425 nm east of San Juan, Puerto Rico. A flight during the evening hours found that the pressure had fallen to 971 mb, and a GPS dropwindsonde measured winds of 99 kts at 925 mb. By very early on the 11th the CP had fallen further to 963 mb and FLW of 107 kts were measured. Very deep convection with cloud tops of -70 to -80 C surrounded a 25 nm-wide eye. The official MSW was increased to 90 kts at 0900 UTC.

 Floyd's intensification levelled off on 11 Sep as it passed through

a mid-level trough which induced some southwesterly shearing over the hurricane. A GPS dropwindsonde during the morning did measure winds of 100 kts at the surface, but this data from a single drop was not considered representative of the sustained wind. The MSW was upped to 95 kts at 11/1500 UTC and remained pegged there for 24 hours.

 Hurricane Floyd began to intensify in earnest on 12 Sep after it had

passed through the trough. A reconnaissance flight early in the morning found a CP of 955 mb and peak FLW of 120 kts; hence, Floyd was upgraded to a Category 3 hurricane with the surface MSW estimated at 105 kts. The hurricane was located about 265 nm north of San Juan at this time. Floyd by now exhibited good outflow in all quadrants, and light vertical shear and warm SSTs lay ahead on its projected path. The CP continued to fall through the afternoon and evening, and at 12/2200 UTC a special Tropical Cyclone Update was issued by TPC/NHC upgrading Floyd to a Category 4 hurricane--the third one of the season. The MSW was estimated at 120 kts and the most recent CP found by a reconnaissance plane was 935 mb. The 13/0300 UTC advisory further increased the MSW to 125 kts, and the CP had fallen to 931 mb. Dvorak estimates from TAFB and SAB of T6.5 (127 kts) supported the MSW of 125 kts.

 A GPS dropwindsonde released by a Hurricane Hunters reconnaissance

aircraft at 12/2315 UTC measured winds of 166 kts at 900 mb (or about 500 m). A dropwindsonde released by a NOAA P-3 research aircraft about an hour earlier, at 2203 UTC, measured winds of 174 kts at 880 mb. This particular drop was made in the northeastern eyewall but had been blown around to the northwest quadrant of the storm before the peak winds were measured. A few hours later (at 13/0528 UTC) another dropwindsonde reported winds in the eyewall of 152 kts at 919 mb. Given that the last reported pressure before the instrument fell into the ocean was 939 mb, this wind measurement was made at only 20 mb (or approximately 200 m) above the surface. A drop made at 0927 UTC also measured winds of 153 kts at 919 mb. (A special thanks to Rich Henning for giving me all this information.)

 Early on 12 Sep, at about the time that Floyd was approaching

Category 3 intensity, the hurricane had turned to more of a westerly course. By the morning of 13 Sep, Floyd was a large, intense, classically-formed hurricane approaching the central Bahamas. A SST analysis performed by John Hopkins University indicated that the hurricane was traversing a warm eddy with temperatures of about 32 C. Floyd's CP continued to fall and by 1200 UTC on 13 Sep had reached 921 mb, which was the minimum measured during the storm's history. A reconnaissance flight found FLW of 146 kts at 700 mb. The official MSW was increased to 135 kts in the 0900 UTC advisory with Floyd being located about 250 nm east of San Salvador Island in the Bahamas. The 921-mb pressure would correlate with 140-kt winds--a Category 5--but it seems that the hurricane specialist on duty at TPC/NHC was understandably reluctant to raise that flag unless Floyd continued to intensify. The discussion bulletin for 1500 UTC contained the statement that "135 knots should be strong enough to get everyone's attention."

 It would be one of the understatements of the year to say that 

Hurricane Floyd got everyone's attention! The approach of the hurricane to the central Florida Atlantic coast led to the largest mass evacuation in United States' history. More than 1.5 million people fled inland as the giant storm approached. Major highways were practically turned into parking lots. The author remembers one media report that all gasoline stations along Interstate Highway 10 between Jacksonville and Tallahassee were out of fuel.

 Floyd's CP remained near or below 930 mb for a period of about

48 hours and the official MSW estimate remained at 135 kts for 30 hours. A GPS dropwindsonde in the northwest eyewall during the afternoon of the 13th measured a wind of 145 kts about 50 mb above the surface. Reconnaissance data on the morning of 14 Sep showed a double wind maxima structure which indicated that a concentric eyewall replacement cycle might be underway. Floyd was a large hurricane at this stage with hurricane-force winds extending outward 110 nm from the center and tropical storm-force winds out to 250 nm. As Hurricane Floyd approached the northwestern Bahamas it began to make an anticipated turn to the northwest. The eye of Floyd was about 30 nm northeast of San Salvador at 14/0000 UTC. It subsequently passed north of Cat Island, just northeast of Eleuthera Island, and passed directly over Great Abaco Island between 14/1800 and 15/0000 UTC.

 During the forenoon of 14 Sep, several hours before the arrival of

Floyd's eye, wind gusts to 100 kts were being reported on Great Abaco. Andros Island, located well to the southwest of the hurricane, experienced gusts to 60 kts with some trees down. As the eye of Floyd neared the Abacos during the afternoon, a pressure of 929 mb was measured at Little Harbour. Floyd's estimated MSW had decreased slightly to around 120-125 kts by the time it struck the Abacos Islands, possibly due to some interaction with the islands and also probably due to the eyewall replacement cycle. During the evening hours of the 14th a reconnaissance flight measured FLW of 137 kts and reported a large eye 50 nm in diameter. Radar observations suggested that a new eyewall was trying to form inside the larger one.

 While passing through the northwestern Bahamas, Floyd began to turn

to an increasingly north-northwestward track, roughly parallel to the east coast of Florida. The hurricane was about 120 nm east of Melbourne at 0600 UTC on 15 Sep and about 140 nm east of Jacksonville at 1800 UTC. The MSW had fallen to 100 kts by the time the storm was abreast of Jacksonville. Early on the morning of the 15th NOAA buoy 41009, located about 20 nm off Cape Canaveral, reported wind gusts to 66 kts and 10.4 m seas. NOAA buoy 41010 reached a peak wind of 70 kts with gusts to 85 kts and seas to 16.5 m. Given the high sea states, these winds were probably too low. Around 0900 UTC buoy 41010 was in the eye of Hurricane Floyd and was reporting a pressure of 938.8 mb. Daytona Beach reported sustained winds of 35 kts with gusts to 52 kts. A reconnaissance flight by the Hurricane Hunters found peak FLW of 134 kts and a CP of 935 mb. The official MSW estimate remained at 120 kts for the 0900 UTC advisory, but as already mentioned above, Floyd's winds began to slowly decrease as the day progressed.

 After 1800 UTC on 15 Sep Floyd began to recurve to the north and

then accelerate north-northeastward toward North Carolina. Charleston, South Carolina, reported gusts to 58 kts as the storm passed about 100 nm or so to the east. A NOAA buoy (ID 42004) located about 40 nm offshore from Charleston reported gusts to 72 kts. The weather station at Frying Pan Shoals (33.5N, 77.5W) reported sustained winds of hurricane force from 16/0300 UTC through 0800 UTC, reaching a peak of 120/87 kts at 0500 UTC accompanied by a peak gust of 97 kts. The minimum pressure reading of 958.6 mb occurred at 0600 UTC. The station with ID of FBIS1 (32.7N, 79.8W) reported a peak sustained wind of 360/47 kts with gusts to 63 kts at 15/2300 UTC, and NOAA buoy 41004 (32.5N, 79.1W) reported its peak wind of 320/54 kts with gusts to 72 kts at 16/0200 UTC. (Most of the buoy observations given above were sent to me by Eric Blake at Colorado State University. A special thanks to Eric for sending me this information. If anyone would like to receive the entire set of buoy reports which Eric sent, please e-mail a request to me and I'll be happy to send a copy.)

 Hurricane Floyd made landfall near or just east of Cape Fear, North

Carolina, around 0700 UTC on 16 Sep with an estimated MSW of 95 kts and a CP of 956 mb. The University of Oklahoma Doppler-on-Wheels (DOW) team recorded a MSW of 70 kts with a peak gust of 107 kts near Topsail Beach. In New Hanover County a peak sustained wind of 85 kts was recorded with the peak gust reaching 113 kts. The attendant pressure measured was 971.9 mb. The Cherry Point Naval Air Station reported a peak gust of 71 kts while a C-MAN buoy at Cape Lookout reported a sustained wind of 60 kts with gusts reaching 79 kts. The station at Diamond Shoals reported a MSW wind of 70 kts at 1300 UTC, which translates to about 60 kts at 10 m elevation. Floyd was well inland at this time and moving rapidly north-northeastward. The last surface report of hurricane force winds was from Duck Pier, North Carolina. At 1400 UTC the C-MAN station reported sustained winds of 65 kts with a peak gust of 83 kts during the preceding hour.

 After gradually making landfall between Cape Fear and New River

Inlet, Hurricane Floyd moved rapidly north-northeastward across the
tidewater region of North Carolina and extreme southeastern Virginia. By 16/1800 UTC Floyd's center was moving offshore near Chincoteague, Virginia. A reconnaissance aircraft measured FLW of 70 to 80 kts over the Atlantic, but a GPS dropwindsonde found winds of only 35 to 40 kts at the surface; therefore, Floyd was downgraded to a tropical storm at 2100 UTC. At 17/0000 UTC the center of Floyd was located about 20 nm east-southeast of New York City. Wind gusts to about 45 kts were reported at Buzzards Bay and New Bedford, Massachusetts.

 Floyd continued moving northeastward near 30 kts and at 0600 UTC on

17 Sep was near Worcester, Massachusetts. The storm was beginning to lose its tropical characteristics and NHC issued the last advisory at 0900 UTC. As it transitioned into an extratropical cyclone Floyd's forward motion slowed somewhat and the storm moved along the coast of Maine in the direction of the Canadian Maritimes. The last position gleaned by the author from a High Seas Forecast placed the storm near the mouth of the Bay of Fundy at 1800 UTC, moving eastward at 15 kts.

 The Bahamas were the first land area to be battered by Hurricane

Floyd. The islands of Eleuthera, Great Abaco, Cat, San Salvador and Grand Bahama were most severely affected. Floyd disrupted electrical power, water and communications throughout the Bahamas, and damage to agricultural crops was also significant. The Bahamas were the first Caribbean nation to establish mandatory building codes, so damage to houses was not as catastrophic as one might expect from a severe Category 4 hurricane. On Eleuthera about 25% of the houses sustained some roof damage but less than 1% could be considered destroyed. On Great Abaco, which Hurricane Floyd's eye passed directly over, about 10% of the island's homes were destroyed and 40% severely damaged. Fortunately, it appears that only two lives were lost in the Bahamas and no severe injuries were reported. More information on Hurricane Floyd's impact on the Bahamas and associated relief efforts can be found at the following website:

 http://wwwnotes.reliefweb.int>

 Although Floyd's winds had abated somewhat by the time the storm

reached North Carolina, the incredibly heavy rains deposited by the hurricane led to what has become the state's greatest natural disaster. Even though the hurricane moved quickly through the eastern coastal plain, rainfall amounts ranging from 150 to 500 mm fell over a large area, leading to widespread and severe flooding. Floodwaters virtually shut down the eastern one-third of the state--more than 45,000 square kilometers and home to 2.1 million people. Over 30,000 homes were flooded and 1600 damaged beyond repair. To make matters worse the floodwaters became polluted by fuel, farm chemicals and manure. Losses to farm animals were staggering: an estimated 100,000 hogs; 500,000 turkeys; and 2.4 million chickens perished in the floods. The rotting carcasses and an explosion in the mosquito population posed an increased threat of disease. Several media reports indicated that the total damage to North Carolina would likely exceed the $5-6 billion caused by Hurricane Fran in 1996. The largest damage estimate seen by the author was around $7 billion.

 New Jersey also experienced some significant flooding from Floyd's

rains, but the magnitude was much less than that which occurred in North Carolina. One media report quoted the Governor of New Jersey as stating that losses would likely $100 million. The flooding in some areas of New Jersey was the worst experienced since the floods caused by Tropical Storm Doria in August, 1971.

 One media source stated that between 2 and 3 million persons all up

and down the Eastern Seaboard evacuated in advance of Hurricane Floyd. Transportation was disrupted--tens of thousands of air travellers were stranded in airports as flights were delayed or cancelled. Over 2.9 million students had a day off as hundreds of schools closed in Washington, Baltimore, Philadelphia, New Jersey and New York City. It was the first time that anyone could remember all the schools in the state of New Jersey and New York City being closed due to a hurricane. More than 1.4 million people lost electrical power from South Carolina to New Jersey.

 The Monthly Summary for September prepared by the staff of TPC/NHC

lists the number of deaths at 69. At least 40 or more of these occurred in North Carolina, making Floyd the second deadliest hurricane in the state's history. (A hurricane in 1883 was responsible for 53 deaths in the state.) A tragedy at sea was narrowly avoided when a tugboat with eight crew members aboard sank in 11 m seas about 300 nm off Jacksonville. A Navy helicopter spotted and picked up three of the men in life jackets floating in the water and holding on to a broom handle. Later, another chopper picked up the remaining five who had managed to board one of the boat's liferafts.

 Hopefully, more statistics on the aftermath of Hurricane Floyd will

become available later and if so, will be included in a future summary. Additional information on Floyd's effects in the United States can be found at the following websites:

 http://www.floydhelp.org>
 http://www.weatherwatchers.org/tropical/1999/08/post/greatest.html>


                     Hurricane Gert  (TC #9)
                        11 - 24 September
                     -----------------------

 An active tropical wave with an associated LOW center left the

African coast around 10 Sep. By 1200 UTC on 11 Sep the system had developed sufficient convective organization so that advisories were initiated on Tropical Depression #9. The large, poorly-defined center was located about 175 nm south of the Cape Verde Islands. The depression had good banding, but there was not much convection near the center and outflow was somewhat restricted on the east side due to some modest easterly shear. Satellite intensity estimates from TAFB and SAB were 30 kts and 25 kts, respectively, so the initial MSW was set at 30 kts. Imbedded within a strong, deep layer of easterlies accompanied by a strong 700 mb jet to the north, the depression moved westward rather quickly at around 17 kts.

 Over the next 24 hours the deep convection became more centralized

and satellite intensity estimates had reached T2.5 from all agencies by the morning of the 12th, so the depression was upgraded to Tropical Storm Gert at 1500 UTC. Gert was centered roughly 450 nm west of the Cape Verdes at this time. The tropical storm continued moving quickly on a course just north of due west and steadily intensified. Gert was upgraded to a hurricane at 1500 UTC on 13 Sep when it was located over 1200 nm east of the northern Windward Islands. By the afternoon of 14 Sep Dvorak estimates from TAFB and SAB had reached T5.5 and Gert's eye had become better defined, so the MSW estimate was increased to 100 kts, making Gert the fourth major hurricane of the season. The hurricane's appearance continued to improve during the evening with an intensity estimate from SAB reaching 115 kts and objective Dvorak estimates from the University of Wisconsin reaching 120 kts. The official MSW was increased to 110 kts at 15/0300 UTC and to 115 kts at 1500 UTC, making Gert the fourth Category 4 hurricane of the season on the Saffir/Simpson scale. By afternoon T-numbers from TAFB and SAB had reached T6.5 (127 kts) and T7.0 (140 kts), respectively, so Gert's MSW was increased to 130 kts on the 2100 UTC advisory. The center of this strong Category 4 hurricane was located about 650 nm east of the island of Guadeloupe at this time. The CP was estimated to be around 930 mb, based upon satellite intensity estimates.

 The satellite appearance of Gert became slightly less impressive

during the night of 15-16 Sep. The first reconnaissance flight by the Hurricane Hunters very early on the 16th measured a CP of 941 mb and FLW of 119 kts. A GPS dropwindsonde released in the eyewall reported winds of 141 kts near the surface. Based on these findings, plus a slightly degraded satellite signature, the MSW was decreased to 125 kts. The aircraft also reported concentric eyewalls with diameters of 18 and 30 nm. The next flight during the afternoon found the CP still near 940 mb. A GPS dropwindsonde measured mean boundary layer winds of 110 kts and peak winds of 138 kts near 900 mb, so the surface MSW estimate was reduced to 115 kts on the 2100 UTC advisory. A subsequent reconnaissance flight very early on 17 Sep reported winds of 149 kts at 868 mb, and a GPS dropwindsonde measured winds of 123 kts at the surface, so the MSW was bumped up once more to 120 kts.

 As Gert approached the 20th parallel on 17 Sep the hurricane began

to turn to the northwest on a track which, over the next few days, would carry it on a broad recurve through the west-central Atlantic. During the day the cloud pattern showed a gradual weakening, possibly due to interaction with an upper-level trough located west of the storm. The MSW estimate dropped to 105 kts by 1500 UTC on 18 Sep but increased to 115 kts once more at 19/0900 UTC as Gert's satellite appearance indicated that the hurricane had once more strengthened.
Dvorak estimates from TAFB and SAB reached T6.5 and T6.0 early on the 19th and Gert displayed a 25 nm-diameter eye surrounded by very cold cloud tops. The hurricane maintained this intensity through 20/0900 UTC, and microwave data indicated that Gert underwent an eyewall replacement cycle early on the 19th.

 After a three-day hiatus, a reconnaissance flight was made into

Gert around midday on 20 Sep. The airplane found a CP of 948 mb and peak FLW of 105 kts about 70 nm northeast of the center. The deep convection associated with Gert had begun to decrease once more by this time. Using data from the reconnaissance flight and 21/0000 UTC Dvorak classifications, a surface wind analysis was performed by HRD. Based on this, the MSW was lowered to 95 kts on the 21/0300 UTC advisory. Gert reached the westernmost point of its track about this time and began to accelerate to the north-northeast on the eastern side of a large 500 mb trough. The storm passed about 150 nm east of Bermuda around 1800 UTC on the 21st where gusts to hurricane force were recorded.

 Hurricane Gert continued to steadily weaken as it moved quickly

into the cooler waters of the North Atlantic. By the morning of 22 Sep satellite estimates indicated that Gert was barely a hurricane; however, buoy 41505, located near the storm's eye, reported a minimum pressure of 967.9 mb, so the MSW was kept at 75 kts. AT 23/0000 UTC the Canadian buoy 44141 reported a pressure of 966.2 mb with winds of 200/23 kts and seas to 13.7 m as Gert passed just to its west. Gert was downgraded to a 60-kt tropical storm at 23/0900 UTC when it was centered about 140 nm south-southwest of Cape Race, located on the tip of Newfoundland's Avalon Peninsula. By 1500 UTC Gert's center was about 50 nm east-southeast of Cape Race and rapidly losing its tropical characteristics. However, reports from the New Hibernia buoy (ID 44145), located at 46.7N, 48.7W, indicated that surface winds were still near 60 kts. Gert was declared extratropical at 2100 UTC, and by 1200 UTC on 24 Sep had merged with another developing extra- tropical cyclone approaching from the west.

                 Tropical Storm Harvey  (TC #10)
                       19 - 22 September
                 -------------------------------

 The progenitor of Harvey seems to have been a tropical wave which

was first mentioned in the Tropical Weather Discussions on 15 Sep when it was traversing the central Caribbean Sea. By the 17th a weak LOW had formed in the northwestern Caribbean which subsequently drifted northward into the eastern Gulf of Mexico. Early on 19 Sep satellite imagery indicated that convection had increased near and east of the broad center. Observations from two buoys (42001 and 42003) suggested a minimum CP of 1005 mb, and buoy 42003 reported 30-kt winds; therefore, advisories were initiated on Tropical Depression #10 at 0900 UTC. The center of the poorly-organized depression was estimated to be about 350 nm south of Pensacola, Florida.

 The center was relocated slightly farther to the north on the next

advisory--something which is not unusual in a broad circulation. A ship towing a barge south of the center and located underneath the deep convection reported winds of approximately 35 kts and a pressure of 1015 mb at 1600 UTC. A reconnaissance aircraft reached the area during the afternoon and confirmed the location of the center and reported an extrapolated CP of 1002 mb. The depression continued to move slowly on a track just east of due north throughout the afternoon and evening of 19 Sep. By 20/0000 UTC the center was located about 260 nm south of Pensacola. The LLCC was still located north and west of the deepest convection. During the evening another reconnaissance flight visited the depression and found peak FLW of 56 kts, while a short time later a ship directly underneath the location of the reconnaissance observation reported south winds of 40 kts and a pressure of 1004 mb. In addition buoy 42003, located about 40 nm east of the ship, reported winds to 33 kts. Based on these pieces of information, the depression was upgraded to Tropical Storm Harvey at 0300 UTC on 20 Sep.

 Tropical Storm Harvey reached a point about 200 nm south of

Ft. Walton Beach, Florida, around 1200 UTC. The northward motion came to a halt and the storm began to drift to the east. Although the LLCC remained just northwest of the deep convection, Harvey managed to intensify some. The MSW reached a peak of 50 kts at 1800 UTC while the storm was centerd about 200 nm south of Panama City. The storm however was beginning to entrain some drier air from the northwest, so the intensification trend seen during the day was halted. Ship reports indicated that the extent of tropical storm-force winds was larger than previously estimated--about 150 nm in the southeast quadrant. Throughout Harvey's lifetime all the tropical storm winds were to be found only in the eastern semicircle, and late in the storm's life were occurring only in the southeast quadrant.

 With a mid-latitude trough approaching from the west, it was thought

that Harvey would soon scoot off to the northeast in advance of the trough. However, the storm had another trick or two up its sleeve. After reaching a point about 120 nm west-southwest of Tampa at 0000 UTC on 21 Sep, Harvey began to move steadily toward the southeast, a move which necessitated that tropical storm warnings be hoisted for south Florida and the Keys. The storm also weakened slightly with the MSW dropping to 45 kts; and as mentioned above, all the gale-force winds were confined to the southeast quadrant of the storm. By around 1200 UTC Harvey's track began to turn to the east and the storm made landfall around midday just south of Naples or about 40 nm south of Ft. Myers. Gusts to tropical storm force were reported from the Keys with Sand Key reporting gusts to 40 kts and Sombrero Key experiencing gusts to 41 kts.

 After moving inland Harvey finally began to accelerate to the east-

northeast as had been anticipated. The storm also began to rapidly lose its tropical characteristics as it began to merge with a developing extratropical LOW located southeast of the Carolinas. The final advisory on Harvey from TPC/NHC, issued at 22/0300 UTC, placed the center (if there was a distinct one left) about 80 nm east- northeast of West Palm Beach or near Grand Bahama Island. The remnants of Harvey were absorbed by the developing LOW southeast of Charleston which was forecast to accelerate off to the east-northeast. Although Tropical Storm Harvey was very poorly-organized as it approached and moved inland into southwest Florida, the associated convection was very strong and Doppler and satellite estimates of rainfall indicated accumulations of around 250 mm in some locations.

                         NOTE by AUTHOR
                         --------------

 The following two systems are the unnamed/unnumbered depressions

mentioned above. Although it now seems unlikely that the first system was of tropical storm intensity, I had already made the decision to dub the system "Delta" in some discussions with others. (I also did this in the companion track files document.) Most of the information contained in the discussion of "Delta" comes from the author's recollection and from Steve Filoso, a young tropical cyclone enthusiast from Phoenix who plans perhaps to study tropical meteorology one day. The other system, which (as noted) is being studied for possible inclusion as an unnamed tropical storm, I have referred to as "Epsilon". The nice write-up on this storm was supplied by John Wallace of San Antonio, and I have for the most part used his summary verbatim. A special thanks to Steve and John for their assistance in studying and preparing tracks for these interesting weather systems.

                    Tropical Depression "Delta"
                         7 - 11 September
                    ---------------------------

 A small low-level swirl of low clouds could be seen in satellite

imagery about 325 nm west-southwest of the Azores around 1800 UTC on 7 Sep. The system was moving slowly in a general southwesterly direction at the time. A small area of deep convection had appeared by the afternoon of 8 Sep, and it was on 9 Sep that the author took note of the system in satellite imagery. The area of deep convection was quite small, but it persisted throughout the day. On the evening of the 9th I sent an e-mail to Mark Lander requesting that he take a look at the system. Mark examined some satellite imagery available on the internet and wrote me back the next morning with his opinion that the system was indeed a midget tropical cyclone and could possibly have winds to 35 kts or higher.

 By late on 9 Sep "Delta" was located about 700 nm west-southwest of

the Azores and from this point moved slowly westward throughout the remainder of its life. On 10 Sep the convection was persisting so I sent an e-mail to James Franklin who happened to be on duty at NHC. James answered me promptly and stated that SAB had also called them inquiring about the nature of the system. James had been studying imagery of the disturbance and was unable to see any clear evidence of rotation in the low-level cloud lines. I looked closely at some animated satellite imagery during the afternoon and was also unable to see any evidence of a circulation in the low-level clouds. However, there was clearly some rotation in the upper-level convective clouds.

 TPC/NHC mentioned the system as a possible candidate for tropical

cyclone development beginning with the Tropical Weather Outlook issued at 2130 UTC on 10 Sep. "Delta" continued to move slowly westward and by the morning of 11 Sep the convection was beginning to decay. As the convection disappeared there was evidence of a weak surface circulation left behind, so it appears likely that "Delta" had indeed been a small depression. The final position in John's track, at 11/1415 UTC, places the center about 900 nm west-southwest of the Azores. The highest MSW given by John in his track was 25 kts.

                  Tropical Depression "Epsilon"
                    (Possible Tropical Storm)
                        13 - 17 September
                  -----------------------------

 A strong tropical wave left the African coast on 6 Sep, shortly

after the wave which had spawned Hurricane Floyd. There was an associated vigorous 1011 mb LOW along the wave axis which boasted a well-defined LLCC with bursting convection--indeed, the system could perhaps have been classified as a depression at this time. However, as the wave tracked westward under the influence of the subtropical ridge to the north, its organization decreased markedly--probably due to some interference from the large Floyd to the west and the developing Gert to the east. Also, the system was underneath an upper-level trough with its associated dry air. The LOW degenerated into an open wave on 11 Sep as it crossed 40W but continued to be accompanied by a large area of scattered, disorganized strong convection.

 The area of disturbed weather showed little change from the 10th

through early on the 12th as it tracked roughly west-northwest between Floyd and Gert. The parent wave dissipated on the 12th as the main area of disturbed weather began to acquire a more northerly component, its motion being influenced by the north-south trough that continued to inhibit development. On 13 Sep the system began to move on a north-northwesterly track. There was a small flare-up of convection at the southern end of the disturbance--which was elongated along a north-south axis--at 0100 UTC. This convection blossomed and there was a rapid, dramatic increase in organization between 0600 UTC and 1600 UTC on 13 Sep as convection expanded and intensified, and the system acquired strong cyclonic curvature. At this time atmospheric conditions had become more favorable with a minimum of shear. At 13/0600 UTC a weak center was located approximately 700 nm east- northeast of San Juan, Puerto Rico, but by 1200 UTC the center had jumped north (or more likely re-formed) to a position about 775 nm southeast of Bermuda.

 The cyclonic disturbance did not present a fully tropical appearance

at this juncture, however, with a "comma cloud" formation and what looked like a very weak baroclinic boundary to the south. There was also no solid CDO-like mass of convection. Water vapor imagery at 13/1600 UTC showed that the disturbance was entangled with an upper- level trough which extended to the south (the same trough which later helped to steer Gert on a more northerly track). Around 2000 UTC there was an intense flare-up of convection over what appeared to be a LLCC. The convection expanded, intensified, and consolidated into a CDO by 14/0000 UTC. Outflow was fairly well-developed at this time, and TPC/NHC mentioned this system in the Tropical Weather Outlook issued at 1530 UTC on the 13th. The north-northwesterly track curved more to the northwest early on 14 Sep. Around 0100 UTC the depression was centered about 550 nm east-southeast of Bermuda.

 "Epsilon" was likely at its peak strength from 13/2000 UTC through

14/0400 UTC, after which time it began to experience significant vertical shear. Visible imagery on the 14th revealed a tight, well- defined LLCC. Low-level visible wind analyses from the University of Wisconsin indicated maximum low-level winds of 30 kts within the circulation but at some distance from the center. There was a possibility that the MSW near the center could have been 35 kts, as John's track reflects. The cyclone continued to experience shear, due in part to outflow from Hurricane Floyd. By 1200 UTC most of the deepest convection had disappeared and only scattered moderate to weak convection remained, displaced well east of the center. The LLCC, however, remained well-defined. The appearance of the system at this stage reminded me very much of Tropical Storm Ana in July, 1997, which managed to retain its intensity even after it had become sheared.

 The cyclone came to within about 120 nm east of Bermuda around 0200

UTC on 15 Sep, after which time it turned toward the north. No strong winds, or even any winds out of the north, were felt at Bermuda. The wind analyses on the 15th from the University of Wisconsin indicated low-level winds on the order of 35-40 kts in the vicinity of "Epsilon", so a MSW of 35 kts seems plausible for this date. The system continued to track slowly northward as its LLCC became increasingly diffuse on visible imagery late on 15 Sep. By 16/0000 UTC it was devoid of any significant convection and had been caught up in the westerlies. The weakened circulation tracked north-northeast south of Nova Scotia, being located about 200 nm southeast of Halifax at 16/1800 UTC. Wind analyses taken at 1645 UTC showed only weak surface winds with much stronger winds aloft well east of the center. The remnants of the cyclone continued northeastward until they were absorbed into the frontal band of a large, powerful gale in the North Atlantic on 17 Sep about 175 nm southeast of Cape Race, Newfoundland.


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

Activity for September: 2 hurricanes

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.

 A special thanks to John Wallace, a tropical cyclone enthusiast

and college student from San Antonio, for providing me with a log which he had kept of all Atlantic/Northeast Pacific tropical waves that proved to be very valuable in helping to trace the pre-depression history of the cyclones.

             Northeast Pacific Activity for September
             ----------------------------------------

 September was a relatively quiet month in the Northeast Pacific

with only two tropical cyclones, both of which briefly became minimal hurricanes. One of these, Greg, made landfall on the southern tip of the Baja California peninsula shortly after being downgraded from a hurricane back to a tropical storm. Hilary trekked northward about 250 nm west of Cabo San Lucas and weakened in the cooler waters west of the Baja.

                     Hurricane Greg  (TC-12E)
                         5 - 9 September
                     ------------------------

 Hurricane Greg developed from a large, monsoon-type disturbance

which persisted off the Mexican coast for several days in early September. Its development may also have been enhanced by a tropical wave which was first identified in the Western Caribbean on 29 Aug and had tracked into the Eastern North Pacific by 1 Sep. By the morning of 5 Sep satellite intensity estimates had reached 30 kts so advisories were initiated on the depression. At 1200 UTC the center was roughly 65 nm southwest of Manzanillo, Mexico, where a wind gust of 35 kts was reported during the morning. The system was upgraded to Tropical Storm Greg with 40-kt winds at 2100 UTC based upon Dvorak estimates of 35 kts and a report from ship 3EJ06 of 42-kt winds and a pressure of 1006.5 mb about 90 nm from the center. Greg initially moved north-northwestward just offshore of the Mexican coast between Manzanillo and Cabo Corrientes. After passing Cabo Corrientes the storm turned more to the northwest and traversed the mouth of the Gulf of California on a beeline for Cabo San Lucas.

 During the night there was a tremendous bursting CDO with very cold

cloud tops. This persisted and by the next morning the radar at Los Cabos indicated that an eye was forming. By early afternoon of 6 Sep satellite intensity estimates from TAFB and SAB had reached 65 kts so Greg was upgraded to a hurricane at 1800 UTC. The center of the hurricane was located approximately 150 nm southwest of the city of Mazatlan, which lies on the Mexican mainland almost due east of Cabo San Lucas. However, even as Greg reached hurricane intensity, there was some evidence that shearing was beginning to take place. There was no outflow to the northeast and some of the low-cloud lines northeast of the eye were exposed. Strangely, water vapor imagery and vertical shear analyses from the University of Wisconsin indicated shear of less than 20 kts, so the ragged appearance was a bit puzzling. The 06/2100 UTC discussion bulletin mentioned that Colima State had reported as much as 380 mm of rain during the past 24 hours.

 During the night of 6-7 Sep the deep convection waned somewhat and

was not very impressive, being mostly confined to the southern semi- circle, but the radar at Los Cabos continued to show a fairly well- defined eye and Dvorak estimates remained at 65 kts. The first visible satellite images on the morning of the 7th revealed a LLCC exposed on the northeast side of the convection and cirrus clouds from a convective band to the northeast blowing across the center-- a sign of moderate to strong shear. Greg was downgraded to a 50-kt tropical storm at 1800 UTC when the center was just southeast of Cabo San Lucas. Satellite intensity estimates at the time from TAFB and SAB had decreased to 55 kts and 45 kts, respectively.

 The tropical storm made landfall on the very tip of the Baja

peninsula around 2100 UTC and turned to more of a westerly course which carried the weakening cyclone into progressively colder SSTs. During the evening of 7 Sep the center was still located under a small area of deep convection which was moving away from the Baja. By the morning of 8 Sep Dvorak estimates from TAFB and SAB were down to 35 kts and 30 kts, respectively, and the deep convection was diminishing. Greg was downgraded to a tropical depression in an intermediate advisory issued at 1800 UTC. The weakening system continued moving westward out into the Pacific and by late afternoon of 8 Sep was only a swirl of low- and mid-level clouds with minimal deep convection about 80 nm to the west of the center. The final advisory placed the dissipating center about 100 nm west of Cabo San Lucas at 09/1200 UTC.

                   Hurricane Hilary  (TC-13E)
                        17 - 21 September
                   --------------------------

 The precursor of Hurricane Hilary was likely a tropical wave which

left the coast of west Africa on 29 Aug. The wave propagated westward across the Atlantic, reaching the eastern Caribbean by 6 Sep and moving into the Eastern North Pacific by 12 Sep. On 17 Sep a tropical LOW was located several hundred miles south of Baja California and was showing signs of increased organization. A special advisory was issued at 17/1600 UTC upgrading the disturbance to a tropical depression. The LLCC had moved closer to deep convection and ships in the area reported that pressures were low enough to classify the system as a tropical depression. At 1200 UTC the center was located about 325 nm southwest of Manzanillo. During the evening a strong burst of convection occurred near the estimated center, but banding features were not well-developed so the system remained classified as a depression. Some conflicting positioning information was received during the early morning of 18 Sep. Satellite fixes from TAFB and SAB placed the center south and east of the previous advisory position, but information from a Tropical Rainfall Measurement Mission (TRMM) overpass suggested a more northerly location. There were indications that the center might be trying to move under the CDO. A strong, persistent area of convection with tops to -80 C was located west or southwest of the center.

 The depression had reached tropical storm strength by 18/1200 UTC

when it was located approximately 400 nm south of Cabo San Lucas. Hilary was moving slowly westward at the time but turned to a north- northwesterly course early on the 19th. The center gradually became better defined throughout the afternoon and evening of 18 Sep and the estimated MSW had reached 50 kts by 19/0000 UTC. By the afternoon of 19 Sep Dvorak estimates had reached T4.0--65 kts--and Hilary was upgraded to a hurricane at 1800 UTC when it was centered about 250 nm southwest of Cabo San Lucas. However, since much colder water lay ahead of the storm on its projected path, weakening was forecast to begin soon. Early on 20 Sep the Dvorak numbers began to decrease as Hilary began to show signs of weakening with the center to the south of the deepest convection. The storm passed about 250 nm to the west of Cabo San Lucas around 1800 UTC with winds down to 40 kts.

 After passing west of the tip of the Baja peninsula, the weakening

Hilary turned to more of a northward course. By the afternoon of 20 Sep the storm had become a tight swirl of low clouds with a few patches of deep convection to the north of the center. Deep convection continued to dwindle and by the afternoon of 21 Sep the system was completely devoid of any deep convective activity. The final advisory was issued at 2100 UTC, placing the center about 150 nm south-southeast of Punta Eugenia. Mid- and high-level moisture from the storm spread northward and eastward into the southwestern United States.


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

Activity for September: 4 tropical storms 1 typhoon (a) 1 super typhoon (b)

(a) - this storm treated as a typhoon by JTWC only

(b) - this category used by JTWC only

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. Also, as announced earlier in a separate posting, a column of 10-min avg MSW is included--the values being obtained from either PAGASA's or JMA's advisories. A special thanks to Michael V. Padua, owner of the Typhoon '99 webpage, 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 and 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 Activity for September
             ----------------------------------------

 Tropical cyclone activity in the Northwest Pacific basin was down

somewhat from the level seen during August. Five tropical depressions formed and all were upgraded to tropical storms by JTWC, JMA, and PAGASA (for those in that center's AOR). Only Bart was recognized as a typhoon by all three agencies, and it was the first Northwest Pacific cyclone of the year to reach JTWC's super typhoon status. York was treated as a typhoon by JTWC but only as a severe tropical storm by JMA. York/Neneng followed a very similar track to earlier storms Sam/Luding and Wendy/Mameng, forming east of Luzon, either crossing or passing near the northern end of the island, and then making landfall in southern China. Super Typhoon Bart also formed off to the east of Luzon but took a northerly track, striking Okinawa and southwestern Japan. Tropical Storms Ann and Zia both formed at higher latitudes and affected Korea and Japan, respectively; while Tropical Storm Cam formed in the South China Sea and moved into southern China just east of Hong Kong.

              Typhoon York/Neneng  (TC-21W / STS 9915)
                         10 - 17 September
              ----------------------------------------

 A STWO issued by JTWC at 0100 UTC on 5 Sep mentioned an area of

convection near 6N, 149E--southeast of Guam--which had persisted for ten hours. Animated visible imagery indicated low-level convergence beneath the convection in a region of near-equatorial troughing with slowly falling surface pressures. By 07/0600 UTC the area of convection had shifted westward to near 10.5N, 135E, or northwest of Yap. The deep convection had become better organized during the past 24 hours. The system resembled a large wave but there was evidence of a weak LLCC forming, so JTWC upgraded the development potential of this disturbance to Fair. Twenty-four hours later the disturbance had moved somewhat to the north to near 15.0N, 134.3E. Although the convection had become more isolated, satellite imagery indicated that a broad circulation had formed, and a surface analysis at 08/0000 UTC indicated the existence of several LLCCs within the broad circulation.

 By 1900 UTC on 9 Sep the main area of interest had moved to near

17N, 123.5W, or just east of Luzon. Scatterometer data indicated a broad LLCC with weak winds near the center, and satellite imagery revealed an increase in deep convection across the region. JTWC issued the first of two Formation Alerts at 10/0300 UTC when animated visible imagery indicated a developing LLCC with persistent though somewhat disorganized convection. PAGASA upgraded the system to Tropical Depression Neneng at 0600 UTC and began issuing advisories. The broad center of Neneng was roughly 125 nm north-northwest of Catanduanes Island and remained quasi-stationary near and just east of northern Luzon for a couple of days. At 11/0300 UTC JTWC issued a second Formation Alert. Surface analyses indicated a LLCC under deep convection, but land interaction was inhibiting intensification, although an anticyclone aloft had strengthened and increased upper- level outflow. JTWC initiated depression warnings on the system at 1800 UTC on 11 Sep, placing the center about 150 nm north-northwest of Catanduanes Island--pretty much where it had been when PAGASA initiated advisories 18 hours earlier. Strongest winds were well southeast of the LLCC although some deep convection was beginning to wrap around the center. Vertical shear was minimal and Neneng had good outflow and was forecast to slowly intensify.

 By 0600 UTC on 12 Sep the LLCC east of Luzon had weakened and deep

convection had consolidated around a second circulation center along the northwest tip of Luzon just west of Laoag. This center became the primary LLCC for Neneng and accelerated west-northwestward at 9 kts across the South China Sea, steered by a weakening mid-level ridge over southern China. PAGASA upgraded Neneng to a tropical storm at 13/0000 UTC and JTWC christened the depression Tropical Storm York six hours later. York/Neneng was then located about 200 nm west of Laoag. A secondary circulation was located just north of Luzon and some of the associated deep convection was being drawn into York. This secondary LOW had been absorbed into the storm by 1200 UTC and the deep convection was becoming better organized, resulting in the MSW being increased to 40 kts. The center of York/Neneng remained quasi-stationary for most of the 13th. Animated visible imagery early on 14 Sep indicated several LLCCs but with the primary center beneath the most persistent deep convection. This resulted in a northeastward re-adjustment to the center position estimate.

 By 0600 UTC on 14 Sep Tropical Storm York had begun to move on a

northwesterly course, guided by a strengthening subtropical ridge over southern China. The 14/1800 UTC warning from JTWC indicated that an SSM/I pass had revealed a well-defined outflow channel to the northeast of the LLCC but that deep convection was confined to the southern semicircle of the cyclone. York continued to strengthen and at 1200 UTC on 15 Sep was upgraded to a typhoon by JTWC. York's center was only about 75 nm southeast of Hong Kong at the time. A broad, ragged eye had developed but the dominant convective shield remained southwest of the center. As Typhoon York approached the coast of China, its motion became slightly more west-northwesterly as the ridge over China continued to re-establish itself.

 York reached its peak intensity of 70 kts at 15/1800 UTC but the

western eyewall was already being eroded somewhat by the entrainment of dry air. (JMA's peak 10-min MSW estimate for York was 55 kts from 14/1200 UTC through 16/0600 UTC.) The typhoon passed just south of Hong Kong shortly after 0000 UTC on the 16th and made landfall around 0700 UTC between Macau and Zhu Hai, China, or about 25 nm west of Hong Kong.

 Typhoon York was responsible for one death in the Hong Kong area

with three others reported missing. There were also 493 injuries attributed to the storm. In the former British colony all banks, clinics, and courts closed due to the passage of the typhoon.
Over 400 persons became trapped in elevators in high-rise apartment buildings and had to be rescued. In southern China eight deaths were attributed to the storm with 10,000 people left stranded away from their homes due to flooding.

               Tropical Storm Zia  (TC-22W / TS 9916)
                          13 -15 September
               --------------------------------------

 An area of convection had developed near 15.5N, 147E--east of the

Marianas--by 9 Sep. Visible satellite imagery indicated a TUTT to the north-northwest of the convection which was helping to create favorable upper-level diffluence over an area of low-level convergence. By the 11th the disturbance had moved to a position northwest of Guam. The deep convection was near the eastern end of a monsoon trough and a LLCC was becoming better defined. The system continued to move to the north-northwest on 12 Sep, and at 1000 UTC JTWC issued a Formation Alert. One or more LLCCs were associated with a recent development of very deep convection within the strong monsoon trough extending east- northeastward from TD-21W (later to become TS York). However, by early on 13 Sep the well-defined LLCC had outrun its deep convection and the upper-level outflow had weakened due to shear from an anti- cyclone to the north; therefore, JTWC cancelled the Formation Alert at 13/0600 UTC.

 A second Formation Alert was issued at 1730 UTC on the same day

when satellite imagery revealed a flare-up of convection over the LLCC concurrent with a reduction in the vertical shear. Although Dvorak intensity estimates were only 25 kts, there were synoptic reports from some ships of winds to 35 kts; hence, the disturbacne was upgraded directly to Tropical Storm Zia at 1800 UTC. Zia's center was located at this time about 150 nm south-southeast of the southeastern tip of Kyushu. The heaviest bands of precipitation were in the northwest quadrant. By 14/0000 UTC satellite intensity estimates ranged from 35 to 45 kts, but based on ship reports, JTWC kept the estimated MSW at 35 kts. Zia at this time was a well-consolidated symmetrical system with low-level cloud lines east of the LLCC. JMA initiated warnings on the storm at 0600 UTC, estimating the 10-min avg MSW to be 45 kts.

 Tropical Storm Zia moved northwestward to near the southeastern

tip of Kyushu, then curved northward, skimming up the eastern coast of the island. The storm then turned northeastward, crossing over the northwestern portion of Shikoku, passing near Matsuyama, and then began to accelerate east-northeastward across Honshu. At 15/0000 UTC Zia passed a short distance west of Kyoto and by 0600 UTC was weakening and losing tropical characteristics in the Sea of Japan northwest of central Honshu. While crossing over Honshu the upper-level anticyclone remained in place and Zia maintained minimal tropical storm intensity even while over land. Even when the storm was downgraded to a depression at 15/0600 UTC the satellite intensity estimates were still near 35 kts.

 Tropical Storm Zia brought heavy rains to western Japan.  Some areas

received almost 500 mm in 24 hours. The heavy rains led to flooding and 85 reported mudslides. One slide near Kamikachi, a popular resort in the mountains of central Japan, left 1300 tourists stranded. More than 14,000 persons had to be evacuated from their homes in Kobe when a nearby river burst its banks. In all, nine deaths resulted from the passage of Zia across west-central Japan.

              Tropical Storm Ann  (TC-23W / STS 9917)
                          15 - 20 September
              ---------------------------------------

 The STWO issued by JTWC at 13/0600 UTC mentioned an area of

convection located southeast of Okinawa near 22N, 136E which had persisted for 24 hours. The convection was in association with strong southwesterly convergence occurring on the eastern side of a strong monsoon trough which extended northeastward from newly-named Tropical Storm York in the South China Sea. Animated visible satellite imagery indicated abundant convection with good outflow to the south but with no discernible LLCC. On the 14th the area moved slowly to the west-northwest with persistent but very disorganized convection. Animated satellite imagery suggested that the thunderstorms were primarily associated with linear surface convergence. At the same time a second area of convection had developed a few hundred miles to the west with some indications of a LLCC. By 15/0600 UTC the two disturbed areas had merged east of Okinawa near 27N, 130E. While the convection appeared weak, visible imagery indicated a LLCC with atmospheric conditions favorable for intensification, so the potential for development was upgraded to Fair.

 The system continued to track generally northward and by 1800 UTC

on the 15th had become sufficiently organized that both JTWC and JMA initiated tropical depression warnings. The depression was centered at that time about 250 nm south-southwest of Nagasaki. TD-23W at first tracked northwestward but later turned to a westerly track as a low-level ridge to the north strengthened. The depression was initially forecast not to strengthen due to the inhibiting effects of vertical shear, but the shearing eventually lessened and the system did intensify, becoming Tropical Storm Ann at 16/0600 UTC when it was centered approximately 225 nm south-southeast of Cheju Island, which lies south of the Korean peninsula. Again, as was the case with Zia, satellite intensity estimates were below tropical storm strength but synoptic reports from ships indicated winds of 35+ kts. (JMA had upgraded the system to Tropical Storm 9917 six hours earlier with a 10-min MSW estimate of 35 kts.)

 Tropical Storm Ann experienced vertical shear throughout most of its

life which kept the deeper convection usually pushed to the north and east quadrants of the storm, leaving the center partially exposed most of the time. A well-defined jet just to the north of the storm combined with a surface ridge to create the shearing environment. At 0000 UTC on 17 Sep visible imagery indicated low-level cloud lines wrapping into the LLCC within the convection, but rotation was not evident in the convective clouds. Ann intensified slightly on the 17th, reaching its estimated peak intensity of 45 kts. JMA, however, estimated the 10-min MSW to be 50 kts, thereby ranking Ann as a severe tropical storm. Intensity fluctuated slightly over the next couple of days, but Ann basically held its own. Around 18/0000 UTC animated satellite imagery revealed low-level cloud lines curving into a fairly small CDO with overshooting tops near the center.

 Ann moved westward across the East China Sea on the 17th and 18th,

passing about 180 nm south of Cheju Island around 17/1800 UTC. Later on 18 Sep the storm began to turn more northwestward as it approached the coast of China. At 1200 UTC on 18 Sep the center was located about 100 nm east of Shanghai, and at 19/0000 UTC Ann reached the westernmost point of its track as it recurved to the northeast just off the Chinese coast at a point roughly 100 nm north-northeast of Shanghai. By late on the 18th some drier air had begun intruding into the storm on the west side and Ann began to slowly weaken. The storm's recurvature was largely effected by a mid-latitude trough approaching from the west. By 1200 UTC on 19 Sep, Ann was a minimal tropical storm with 35-kt winds moving slowly northeastward across the Yellow Sea. The LLCC had become completely exposed, although there was an increase in convection due to enhanced divergence aloft.

 Based on scatterometer data, at 19/1800 UTC the center of Ann was

relocated about 90 nm south of the previous warning position. All the deep convection was displaced well to the northeast of the LLCC and was being rapidly advected along a shear line which lay over central Korea. Ann was downgraded to a 25-kt depression at this time. Six hours later all the convection associated with the system had dissipated and the final warning was issued, placing the weakening center in the Yellow Sea just off the west coast of South Korea about 200 nm south-southwest of Seoul. The author has learned of no damage or casualties caused by Tropical Storm Ann. If any come to light later, they will be reported in a future summary.

           Super Typhoon Bart/Oniang  (TC-24W / TY 9918)
                          17 - 25 September
           ---------------------------------------------

 An area of convection was evident in satellite imagery on 14 Sep

south of Okinawa (near 24.3N, 128.3W) within an active monsoon trough. Surface data suggested that a LLCC might be forming, but this area subsequently merged with a disturbance to the north. The next day another area of convection had formed to the south, near 21.5N, 129E due to enhanced divergent flow aloft, but there was currently no LLCC associated with this new area. By the 16th the main convective area had moved to near 19N, 127.5E, or several hundred miles east of Luzon. There was good outflow to the south but synoptic data indicated predominantly linear convergence throughout the area. The environment though was becoming more favorable for tropical cyclogenesis and JTWC upgraded the development potential of the disturbance to Fair. The system then drifted back eastward a bit to near 20.5N, 129.8E by early on 17 Sep. Synoptic observations plus scatterometer data indicated that a LLCC with some associated deep convection had formed, and JTWC issued a Formation Alert for the disturbance at 0300 UTC.

 JTWC and JMA both initiated tropical depression warnings on the

developing LOW at 17/1200 UTC. The depression initially remained quasi-stationary in weak steering flow at a position about 475 nm east- northeast of northern Luzon or 425 nm south-southeast of Okinawa. The early development of this system was hindered somewhat as shearing induced by a TUTT cell located to the northwest kept the center exposed from the deep convection. Around 1200 UTC on 18 Sep there were signs that the TUTT was moving slowly westward and that the shear was beginning to lessen a bit, and by 1800 UTC the TUTT was far enough west of the depression that it was beginning to create diffluent flow over the system. This led to more favorable conditions for strengthening and JTWC upgraded the depression to Tropical Storm Bart at 19/0000 UTC. (JMA also upgraded the system to Tropical Storm 9918 at this time, and PAGASA as well began issuing warnings at 0000 UTC on Tropical Depression Oniang--which was upgraded to a tropical storm six hours later.) A Tropical Rainfall Measurement Mission (TRMM) pass indicated deep convection extending northwest through southwest around the vortex with two developing low-level cloud bands to the north.

 At the time that Bart was named it was located about 300 nm south

of Okinawa and moving generally northwestward. The storm was experiencing some northerly shear which kept the deeper convection pushed to the southern semicircle. A Special Sensor Microwave Imager (SSM/I) pass depicted a distinct low-level cloud band wrapping in toward the LLCC from the southwest. By 0000 UTC on 20 Sep this cloud band had developed around the eastern side of the vortex--indicative of further strengthening, and visible imagery revealed an eye 10 nm in diameter almost completely enclosed by an eyewall. With a developing upper-level anticyclone over Bart, JTWC upgraded the storm to a typhoon at 0000 UTC. With regard to Bart's motion--a subtropical ridge to the north had weakened and the typhoon came to a virtual standstill from around 19/1800 UTC through 21/0600 UTC in an area roughly 200 nm southwest of Okinawa. Bart did drift very slowly northward during this time and by 21/0000 UTC had intensified to 110 kts with a well- defined eye 10 nm in diameter. Bart's outflow was inhibited slightly to the northeast by a small TUTT, but this feature was filling by 1200 UTC, and by 1800 UTC Bart's estimated MSW had reached 130 kts, thereby making it the first super typhoon of the year. The diameter of the eye was 21 nm and Bart had excellent outflow in all quadrants.

 Super Typhoon Bart moved northeast on 21 Sep, and then turned north

on the 22nd on a track which carried the eye just to the west of Okinawa shortly after 1200 UTC. Kadena Air Force Base reported sustained winds (2-min avg) of 80 kts with a peak gust of 126 kts. The JTWC warning for 22/1200 UTC also mentioned that gusts to 140 kts had been experienced somewhere on the island, but the location wasn't specified. By 1800 UTC Bart had begun to weaken slightly due to some dry air advection, and the eye diameter was 8 nm and shrinking. At 1800 UTC Futemna (ID 47933) reported sustained winds (presumably 10-min avg) of 70 kts from the southeast with gusts to 90 kts, and Kadena (ID 47931) was reporting 45-kt winds with a peak gust of 64 kts. According to some information sent by Patrick Hoareau of Rennes, France, Naha recorded 474 mm of rain in the 24 hours ending at 0900 UTC on 23 Sep. (The September average rainfall for Naha is 170 mm.)

 By 23/0000 UTC Bart was undergoing an eyewall replacement cycle.

The inner eyewall had dissipated and a 47 nm-wide partial eyewall was forming. The storm had dropped below super typhoon intensity at this time with JTWC estimating the MSW at 120 kts. (JMA's peak 10-min avg MSW for Bart was 90 kts from 22/1200 UTC through 23/0600 UTC.) After passing Okinawa, Bart began to accelerate to the north due to a mid- latitude trough approaching from the west. By 1800 UTC on the 23rd the typhoon was only 28 nm off the Kyushu coast and about 60 nm south-southwest of Nagasaki, still carrying 110-kt winds (1-min avg estimate from JTWC--JMA's 10-min avg estimate was 80 kts). Data from a TRMM pass showed a symmetric 40-nm wide eye with a significant low-cloud band northeast of the LLCC moving over Kyushu.

 At about the time that Bart reached southwestern Japan the trough

to the west increased in amplitude concurrently with an increase in the amplitude of a ridge to the northeast. The typhoon responded with a turn to a northeasterly track with a great acceleration in forward motion. Bart made landfall in west-central Kyushu near Omuta, then crossed the extreme western end of Honshu, and by 24/0000 UTC was emerging into the Sea of Japan west of Hiroshima. The storm was racing northeastward at 30 kts by this time, and the JTWC warning for 24/0000 UTC noted that deep convection had persisted even though the center of the cyclone had been moving over land. At 0000 UTC Iwakuni Marine Corps Air Station reported southeast winds of 52 kts with a peak gust of 76 kts.

 Typhoon Bart continued moving rapidly northeastward in the Sea of

Japan west of Honshu. By 24/1200 UTC most of the deep convection had been sheared to the northeast by strong upper-level southwesterlies. The storm by this time was west of the northern tip of Honshu and racing northeastward at 40 kts. JTWC issued its last warning at 1200 UTC, declaring Bart to be in the process of transitioning into an extratropical cyclone. JMA continue to issue warnings on the storm, now downgraded to a tropical storm, for another 18 hours as Bart continued northeastward, crossing western and northern Hokkaido. By 0600 UTC on 25 Sep the system had become extratropical in the Sea of Okhotsk east of northern Hokkaido.

On Okinawa two fatalities were attributed to Bart.   In Japan the

storm claimed 25 lives--12 of these caused by high waves which engulfed 60 homes at Shiranui on Kyushu. Rainfall over western Japan ranged from 200 to 300 mm which led to some flash flooding. A tornado spawned by the storm struck parts of Toyohashi with 210 students at a school injured by flying debris. Typhoon Bart also damaged the Itsukushima Shrine at Hiroshima which is included on the United Nations World Cultural Heritage list.

               Tropical Storm Cam  (TC-25W / TS 9919)
                          22 - 26 September
               --------------------------------------

 The first mention by JTWC of the disturbance from which Tropical

Storm Cam developed was in the daily STWO on 23 Sep which stated that an area of convection had developed over the South China Sea west of Luzon. An exposed LLCC was evident just southeast of some deep convection and there was a band of convection developing 80 nm south of the LLCC. The system was located in the monsoon trough and was given a development potential rating of Fair. However, JMA had begun treating the system as a tropical depression the previous day, so there must have been some evidence of a LLCC on 22 Sep. JTWC began writing warnings on TD-25W at 23/0600 UTC when the system was centered approximately 275 nm south-southwest of Hong Kong. Deep convection was beginning to build around the LLCC, and the depression tracked slightly east of due north in a reverse monsoon trough flow pattern.

 The depression was upgraded to Tropical Storm Cam at 0600 UTC on

24 Sep when the center was located roughly 175 nm south-southeast of Hong Kong. The system had become better organized with deep convection wrapping around the southern half of the LLCC. Cam continued moving generally on a northward course toward the southern coast of China. The storm reached a peak intensity of 40 kts at 24/1200 UTC as per JTWC warnings while JMA estimated the 10-min avg MSW to have reached 45 kts at 25/0600 UTC. On 25 Sep Cam was fairly well-organized with deep convection over the center but displaced slightly to the southwest, but by 1200 UTC the deep convection was beginning to weaken due to interaction with the land mass of southeastern China.

 The system began to weaken quickly as it approached the coast, and

JTWC issued the last warning at 1800 UTC when Cam's center was just off the southern Chinese coast about 65 nm east-southeast of Hong Kong. JTWC had been consistently forecasting the storm to turn to the west due to the strengthening of a mid- and low-level ridge to the north. This did not happen while JTWC was issuing warnings, but JMA continued to issue warnings on the storm for another 24 hours, keeping Cam as a tropical storm through 26/0000 UTC. The JMA track shows Cam making landfall east of Hong Kong and then turning due westward, passing north of Hong Kong and over Canton. The final JMA position at 1800 UTC on 26 Sep placed the dissipating center inland about 240 km west- northwest of Hong Kong or about 130 km west of Canton. The author has not learned of any fatalities as a result of Tropical Storm Cam, but there was a report of 23 injuries caused by the storm in the Hong Kong area.

                   ADDENDA to EARLIER SUMMARIES
                   ----------------------------

 John Wallace of San Antonio, Texas, sent to me some additional

information on the effects of Typhoons Olga and Sam. A special thanks to John for passing along this information.

 Olga - Olga was responsible for 79 deaths with 26 missing in Korea,
        and for one death in Japan.   Severe flooding in South Korea
        was reported and at one point 940,000 households were
        without electrical power.

 Sam  - Sam was the wettest tropical cyclone to affect Hong Kong
        since records were first kept in 1884.  During the period
        22-25 Aug, 609 mm of rain were recorded in the former Crown
        Colony.  Two persons were killed and 206 injured in the
        airliner crash which occurred there during the typhoon's
        passage.

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

Activity for September: No tropical cyclones


SOUTH INDIAN OCEAN (SIO) - South Indian Ocean West of Longitude 90E

Activity for September: No tropical cyclones


AUSTRALIAN REGION (AUG) - From Longitude 90E Eastward to Longitude 160E

Activity for September: No tropical cyclones


SOUTHWEST PACIFIC (SWP) - South Pacific Ocean East of Longitude 160E

Activity for September: 1 tropical depression

             Southwest Pacific Activity for September
             ----------------------------------------

 There were no tropical cyclones in the Southwest Pacific during the

month, but there was an out-of-season development which was carried by the Fiji TCWC as a tropical depression. The first warning on the system at 05/0600 UTC located the center about 325 nm south-southeast of Port Vila in Vanuatu, or roughly 500 nm southwest of Fiji. The depression drifted very slowly generally in a south-southeastward direction for a day or so, and then weakened--the last warning being issued at 1200 UTC on 6 Sep. According to Alipate Waqaicelua, the Chief Forecaster at the Nadi TCWC, this system's development was related to a weak easterly wave passing through Vanuatu which was enhanced by a sharp upper-level trough off the east coast of Australia. The associated convection did not survive long under westerly and northwesterly shear, and the depression was likely more of a hybrid- type system rather than a true tropical depression.


                          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 included in the July, 1998 summary. I will include this glossary from time to time, primarily in "lean" months without a lot of tropical cyclone activity to cover. But if anyone missed receiving it and wishes to obtain a copy, send me an e-mail privately and I'll forward 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
       September as an example:   sep99.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,
  1. 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: sep99.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://australiansevereweather.com/cyclones/>
http://www.geocities.com/capecanaveral/6825>
http://www.hcane.com/tropcl.html>
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 new website

the complete Annual Tropical Cyclone Report for 1997 (1996-1997 season for the Southern Hemisphere). Also, tracks only for the 1998 tropical cyclones are currently available.

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

 Also, TPC/NHC has available on its webpage nice "technicolor"

tracking charts for the 1998 Atlantic and Eastern North Pacific tropical cyclones; also, preliminary storm reports for all the 1998 Atlantic and Eastern North Pacific cyclones are now available.

 The URL is:  http://www.nhc.noaa.gov/1998.html>

Prepared by: Gary Padgett E-mail: [email protected] Phone: 334-222-5327 (nights & weekends) / 850-882-2594 (weekdays)