Debby struck the islands of the northeast Caribbean as a 65-knot hurricane,
but the impact was not very significant. It appeared that Debby was going
to be a significant threat to south Florida, but the system dissipated
rather unexpectedly.
a. Synoptic History
A strong tropical wave moved off the west coast of Africa on 16 August,
accompanied by winds to near 50 kt at the 650 mb level over Dakar. The
following day, a broad area of low pressure associated with this wave was
noted in the vicinity of 10N 30W, but there was insufficient curvature in
the associated bands of deep convection for a Dvorak classification of the
system. A low-level circulation center was identified by the Tropical
Analysis and Forecast Branch (TAFB) at 0000 UTC 18 August, but the system
was still considered "too weak to classify". An initial
classification was done by TAFB at 1145 UTC on the 18th. By that
time the curvature of the convective bands had increased, however the system
was still broad and poorly organized. As the disturbance continued westward
around 15 kt, it gradually became better organized. By 1800 UTC 19 August
the cloud pattern had become consolidated around a well-defined center and
it is estimated that the fourth tropical depression of the season developed,
about 900 n mi east of the Windward Islands
(Table 1, Figure 1).
Vertical shear was weak over the area, with anticyclonic outflow developing
aloft. In this environment, the cyclone strengthened and became Tropical
Storm Debby around 0600 UTC 20 August. A pronounced mid-level ridge to the
north of the cyclone maintained a west-northwestward motion. By midday on
the 20th, microwave imagery data began to suggest some southwesterly
shearing over Debby, with the low-level center displaced a bit to the
southwest of the mid- to upper-level center. Nonetheless the storm
strengthened further and is estimated to have become a hurricane by 0600 UTC
21 August. By this time the infrared imagery also showed the signature of a
sheared system. Even though Debby was not well organized it strengthened
even further, and reconnaissance data indicated that its maximum winds
increased to 75 knots, which was the peak intensity, a little later on the
21st.
Radiosonde data from the Lesser Antilles indicated that the southwesterly
shearing may initially have been due to strong lower-tropospheric easterly
flow, with the low-level center "outrunning" the mid- to upper-level center.
By late on 21 August, the strengthening ceased. Debby was a 65-knot
hurricane when its center moved across the extreme northern Leeward Islands
from 0600 to 1200 UTC on the 22nd. Continuing west-northwestward,
Debby's center moved over the British Virgin Islands around 1500 UTC on the
22nd, and passed about 30 n mi off the northeast coast of Puerto
Rico a few hours later. By early on the 23rd, southwesterly
shearing over Debby increased (dropsonde data from a two-aircraft synoptic
flow mission revealed that upper-tropospheric flow was now largely
responsible). Although not well organized, the system maintained minimal
hurricane strength until around 1200 UTC on the 23rd when the
cloud pattern became even more disorganized-looking.
The weakening storm turned toward the west and moved along the northern
coast of Hispaniola. Although the mountainous land mass of that island may
have played some role in the weakening by restricting inflow from the south
and disrupting the southern part of the cyclone's circulation, it appears
that vertical shear was the main cause for weakening. Around midday on the
23rd, a distinct low-cloud circulation center was evident just to
the north of Hispaniola, displaced well to the west of the main area of deep
convection. The cyclone continued westward, entering the Windward Passage
around 0000 UTC 24 August. It dissipated near the south coast of eastern Cuba
on the morning of the 24th. Debby's remnant, a strong tropical
wave, continued to track westward, spreading locally heavy showers and gusty
winds over Cuba, the Straits of Florida, and southern Florida over the next
couple of days.
b. Meteorological Statistics
Table 1 lists the best track positions and intensities of Debby at
six-hourly intervals. Figure 1 is a display of this track.
Figure 2 and Figure 3
depict the curves of maximum one-minute average "surface" (10 m above ground
level) wind speed and minimum central sea-level pressure, respectively, as a
function of time. Also plotted are the observations on which the curves are
based. These consist of measurements by reconnaissance aircraft, Global
Positioning System (GPS) dropsondes, Dvorak-technique estimates using
satellite imagery from the Tropical Analysis and Forecast Branch (TAFB), the
Satellite Analysis Branch (SAB), and the U.S. Air Force Weather
Agency (AFWA).
Debby's peak intensity of 75 kt on 21 August is based on 85% of the maximum
1500-ft flight level wind speed, 88 kt, measured by aerial reconnaissance.
Interestingly, the minimum central pressure was about 1004 mb at that time,
well above the values typically associated with even minimal hurricane
intensity. A central pressure of 991 mb was measured about 12 hours
after the occurrence of Debby's estimated peak winds.
Wind gusts to 48 kt were reported at Antigua late on 21 August during the
passage of a band of convection in advance of Debby. Maximum sustained
winds there were only 25 kt. At St. Barthelemy, a gust to 66 kt was
reported at 0830 UTC on the 22nd (that station was in the eye at 0915 UTC),
and at St. Maarten, sustained winds of 33 kt with gusts to 52 kt were
reported at 0900 UTC, with a minimum pressure of 998 mb at 1000 UTC. There
were no measurements of sustained tropical storm force winds from the U.S.
Virgin Islands or Puerto Rico, however, wind gusts of 39 and 37
kt were reported from St. Thomas and St. Croix, respectively, on 22 August.
Rainfall totals were as high as 12.63 in at Rio Piedras and 10.28 in at Rio
de la Plata, Puerto Rico. These heavy rains occurred mainly after the
center of Debby had moved northwest of the island.
c. Casualty and Damage Statistics
A man died in Puerto Rico when he fell off the roof of his home while trying
to remove a satellite antenna. This was an indirect fatality.
There was moderate damage to the roofs of a few structures in Barbuda.
There, and in several of the other islands of the extreme northern Leewards,
some damage to fruit trees, utility poles and lines took place. Overall,
the damage did not appear to be severe, but there are no monetary estimates
available.
In Puerto Rico, the main impact came from heavy rainfall. There were
reports of mud slides and damaged or collapsed bridges. Over 400 homes were
reportedly "affected" by flood waters. Five homes suffered moderate to
severe structural damage. Total damage estimate is 0.5 million dollars.
There was also some damage due to storm surge and wave action along the
northern coast of the Dominican Republic.
Debby's rains were beneficial over eastern Cuba, since that region had been
suffering from a severe drought.
d. Forecast and Warning Critique
Table 2 lists the average track prediction errors for Debby,
for a selection of the numerical models and for the official forecast, as
well as the long-term averages of the latter. The average official track
forecast errors were lower than the most recent ten-year averages. The
average official forecast errors for Debby are generally of comparable
magnitude to, or below, the corresponding average model errors. It should
be added, however, that several of the official and model forecasts having a
northward bias were not verified because of Debby's unpredicted dissipation.
The numerical track predictions for 0000 UTC 23 August are of particular
interest, since GPS dropsonde data from a two-aircraft synoptic flow
surveillance mission appeared to have a significant impact on some of the
models. In comparison to earlier forecasts, there was a significant
southward shift in the model tracks. This was particularly noticeable in
the U.K. Met. Office global model (UKM), whose 72-h forecast shifted from
near or just east of Florida, to near western Cuba. Dropsonde measurements
of the steering flow around the above time showed that there was a strong
mid-level anticyclone to the north of Debby, and that a trough was bypassing
the hurricane. It should be added that later sensitivity runs of the UKM
from the above initial time, with and without the dropsonde input, confirmed
that the track forecast improvements were attributable to these data.
The Debby case reminds us, all too well, of our limitations in forecasting
tropical cyclone intensity change. In spite of the sheared nature of the
system, the GFDL model repeatedly predicted that Debby would reach southern
Florida area with a central pressure corresponding to a major hurricane.
Moreover, the best available intensity guidance, SHIPS, consistently
forecasted Debby to intensify (even when it was apparent that the tropical
cyclone was dissipating). Given the objective guidance, and the possibility
that the vertical shear could diminish, the official forecasts mostly
over-predicted the wind speeds. The average bias in the official intensity
forecasts ranged from +5.8 knots at 24 h to +28.0 knots at 72 h. These are
similar to the biases shown by SHIPS.
Table 3 summarizes the various watches and warnings
issued in connection with Debby. The hurricane warning for the northern
Leeward Islands was issued 15 to 18 h prior to the arrival of the center
in that area.
Watches and warnings are not always necessary to trigger preparedness
actions. Based on the official 3-day forecasts which brought Debby near
south Florida, an evacuation of non-residents from the Florida Keys was
ordered, resulting in a significant loss of tourism revenue for Monroe
County.
Table 1.
Best track, Hurricane Debby, 19-24 August 2000.
Date/Time (UTC) | Position |
Pressure (mb) |
Wind Speed (kt) | Stage |
Lat. (°N) | Lon. (°W) |
19 / 1800 | 12.0 | 44.5 | 1010 | 30 | tropical depression |
20 / 0000 | 12.6 | 45.3 | 1010 | 30 | " |
20 / 0600 | 13.3 | 46.8 | 1009 | 35 | tropical storm |
20 / 1200 | 14.0 | 48.8 | 1008 | 40 | " |
20 / 1800 | 14.7 | 50.6 | 1007 | 45 | " |
21 / 0000 | 15.1 | 52.1 | 1006 | 55 | " |
21 / 0600 | 15.4 | 54.0 | 1005 | 65 | hurricane |
21 / 1200 | 15.7 | 56.3 | 1004 | 75 | " |
21 / 1800 | 16.1 | 58.5 | 1004 | 75 | " |
22 / 0000 | 16.8 | 60.1 | 995 | 70 | " |
22 / 0600 | 17.5 | 61.7 | 993 | 65 | " |
22 / 1200 | 18.1 | 63.5 | 994 | 65 | " |
22 / 1800 | 18.8 | 65.4 | 995 | 65 | " |
23 / 0000 | 19.2 | 66.7 | 995 | 65 | " |
23 / 0600 | 19.5 | 68.1 | 995 | 65 | " |
23 / 1200 | 19.8 | 69.7 | 1005 | 60 | tropical storm |
23 / 1800 | 20.0 | 71.5 | 1009 | 50 | " |
24 / 0000 | 19.9 | 73.3 | 1010 | 40 | " |
24 / 0600 | 19.6 | 75.1 | 1011 | 35 | " |
24 / 1200 | 19.5 | 77.0 | 1011 | 30 | tropical depression |
24 / 1800 | | dissipated |
21 / 1200 | 15.7 | 56.3 | 1004 | 75 | maximum intensity |
22 / 0300 | 17.1 | 60.9 | 991 | 70 | minimum pressure |
| Landfalls |
22 / 0600 | 17.5 | 61.7 | 993 | 65 | landfall near Barbuda |
22 / 0915 | 17.9 | 62.8 | 993 | 65 | landfall near St. Barthelemy |
22 / 1500 | 18.5 | 64.4 | 994 | 65 | landfall near Virgin Gorda |
Table 2.
Preliminary forecast evaluation of Hurricane Debby, heterogeneous sample.
(Errors in nautical miles for tropical storm and hurricane stages with
number of forecasts in parenthesis).
Forecast Technique |
Period (hours) |
12 | 24 | 36 | 48 | 72 |
AVNI | 54 (15) | 82 (13) | 106 (11) | 85 (9) | 37 (5) |
CLIP | 48 (15) | 91 (13) | 138 (11) | 172 (9) | 378 (5) |
GFDI | 45 (15) | 74 (13) | 100 (11) | 106 (9) | 145 (5) |
NGPI | 43 (15) | 80 (13) | 112 (11) | 138 (9) | 149 (5) |
UKMI | 37 (14) | 75 (12) | 96 (10) | 80 (8) | 87 (5) |
| | | | | |
NHC OFFICIAL | 43 (15) | 75 (13) | 90 (11) | 85 (9) | 132 (5) |
NHC OFFICIAL 1990-1999 10-year average | 46 (2057) | 85 (1842) | 122 (1650) | 158 (1471) | 235 (1164) |
Table 3. Watch and warning summary, Hurricane Debby, August 2000.
Date/Time (UTC) |
Action |
Location |
20/2100 | Hurricane watch issued | St. Maarten, Saba, and St. Eustatius |
20/2100 | Tropical storm watch issued | Antigua, Barbuda, and Anguilla |
21/0300 | Hurricane watch issued | U.S. Virgin Islands |
21/0300 | Tropical storm watch issued | British Virgin Islands, St. Martin, St. Barthelemy, Nevis, St. Kitts, Montserrat, and Guadeloupe and its surrounding islands |
21/0900 | Hurricane watch issued | Puerto Rico |
21/1500 | Hurricane warning issued | Guadeloupe northward and northwestward through the British and U.S. Virgin Islands |
21/1500 | Tropical storm warning and hurricane watch issued | Dominica |
21/2100 | Hurricane warning issued | Puerto Rico and its adjacent islands |
21/2300 | Tropical storm watch issued | Dominican Republic |
22/0300 | Hurricane watch issued | Haiti north of Port au Prince |
22/0300 | Tropical storm warning issued | Dominican Republic from Punta Palenque to Cabrera |
22/0400 | Tropical storm watch issued | Southeastern Bahamas and the Turks and Caicos Islands |
22/0900 | Hurricane warning issued | Northern coast of the Dominican Republic |
22/0900 | Tropical storm watch changed to hurricane watch | Southeastern Bahamas and the Turks and Caicos Islands |
22/1200 | Hurricane warning discontinued | Guadeloupe, Antigua, Barbuda, Nevis, St. Kitts, and Montserrat |
22/1200 | Tropical storm warning discontinued | Dominica |
22/1500 | Hurricane warning issued | Southeastern Bahamas and the Turks and Caicos Islands |
22/1500 | Hurricane watch issued | Central Bahamas |
22/1500 | Hurricane warning discontinued | St. Maarten, Saba, and St. Eustatius |
22/2100 | Hurricane watch issued | Northern coast of Cuba for the provinces of Holguin and Las Tunas |
22/2100 | Tropical storm warning issued | Haiti north of Port au Prince |
22/2100 | Hurricane warning changed to tropical storm warning | U.S. Virgin Islands and Puerto Rico |
22/2100 | Hurricane warning discontinued | All islands east of the U.S. Virgin Islands |
23/0300 | Tropical storm warning discontinued | U.S. Virgin Islands and Puerto Rico |
23/0600 | Hurricane warning issued | Central Bahamas |
23/0600 | Hurricane watch issued | Northwestern Bahamas |
23/1500 | Hurricane warning issued | Northern coast of Cuba for the provinces of Guantanamo, Holguin, and Las Tunas |
23/1500 | Hurricane watch issued | Northern coast of Cuba for the provinces of Camaguay and Ciego de Avila and southern coast of Cuba for the provinces of Santiago de Cuba and Granma |
23/1800 | Hurricane watch discontinued | Haiti north of Port au Prince |
23/2100 | Hurricane warning changed to tropical storm warning | Central and Southeastern Bahamas and the Turks and Caicos Islands, northern coast of Cuba for the provinces of Guantanamo, Holguin, and Las Tunas |
23/2100 | Hurricane watch changed to tropical storm watch | Northern coast of Cuba for the provinces of Camaguay and Ciego de Avila and southern coast of Cuba for the provinces of Santiago de Cuba and Granma |
23/2100 | Hurricane warning discontinued | Northern coast of the Dominican Republic |
23/2100 | Hurricane watch discontinued | Northwestern Bahamas |
24/0300 | Tropical storm warning discontinued | Haiti, Turks and Caicos Islands |
24/1000 | Tropical storm warning discontinued | Central and southeastern Bahamas |
24/1500 | All tropical storm watches and warnings discontinued | Cuba |
Figure 1.
Best track for Hurricane Debby, 19-24 August 2000.
Figure 2.
Best track maximum sustained surface wind speed curve for Hurricane Debby,
19-24 August 2000, and the observations on which the best track curve is
based. Aircraft observations have been adjusted for elevation using 90%,
80%, and 85% reduction factors for observations from 700 mb, 850 mb, and
1500 ft, respectively. Dropsonde observations include actual 10 m winds
(sfc), as well as surface estimates derived from the mean wind over the
lowest 150 m of the wind sounding (LLM ) , and from the sounding boundary
layer mean (MBL).
Figure 3.
Best track minimum central pressure curve and central pressure observations
or estimates for Hurricane Debby, 19-24 August 2000.