Tropical Cyclone Report
Hurricane Danielle
13 - 21 August 2004
Stacy R. Stewart
National Hurricane Center 19 November 2004
Danielle was a high-end category 2
hurricane that remained over the open waters of the far eastern
Atlantic Ocean without threatening land.
a. Synoptic History
The vigorous westward-moving tropical wave
that spawned Danielle moved off the west coast of Africa early on
12 August. While over land, the system already possessed several
characteristics associated with tropical cyclones - a well-defined
low-level wind field, bands of deep convection spiraling into the
center, and a well-established anticyclonic outflow pattern. After
the wave reached the warm Atlantic waters about 450 n mi southeast
of the Cape Verde Islands, more deep convection developed near the
center of circulation. Curved convective bands became better
defined and Dvorak satellite classifications were initiated at 1800
UTC that day. The wave moved west-northwestward at 12-14 kt,
becoming better organized, and it is estimated that a tropical
depression formed from it around 1200 UTC 13 August about 210 n mi
southeast of the southernmost Cape Verde Islands. The "best track"
chart of the tropical cyclone's path is given in Figure 1, with the
wind and pressure histories shown in
Figure 2 and Figure 3, respectively.
The best track positions and intensities are listed in Table 1.
Owing to the already well-organized
structure, low vertical shear, and very warm sea-surface
temperatures, deep convection continued to become better organized
and it is estimated that the cyclone became a tropical storm at
0000 UTC 14 August. Upon reaching an intensity of 45 kt just 12 h
later, the favorable environmental conditions enabled Danielle to
undergo a period of rapid intensification (> 30 kt/24 h),
with the cyclone becoming a hurricane at 0000 UTC 15 August about
295 n mi west-southwest of the southernmost Cape Verde Islands.
Rapid intensification continued until an intensity of 80 kt was
obtained around 1200 UTC that day. Afterwards, the intensification
trend leveled off to a slower than average rate, possibly due to
the eye and radius of maximum winds having decreased and
subsequently stabilized to a small diameter (for example, see
Figure 4).
Moving northwestward toward a weakness in
the subtropical ridge, Danielle reached its estimated maximum
intensity of 95 kt (category 2 on the Saffir-Simpson Hurricane
Scale) at 1800 UTC 16 August about 755 n mi west of the
northwesternmost Cape Verde Islands. Shortly after Danielle reached
its peak intensity, a large mid- to upper-level trough that had
eroded the subtropical ridge and enhanced the poleward outflow also
began to increase the southwesterly vertical shear across the
cyclone. The increasing shear brought about steady weakening over
the next 72 h as the hurricane moved northward through a large
break in the subtropical ridge. Danielle became a tropical storm by
1200 UTC 18 August and turned northeastward under the influence of
the moderate southwesterly mid-level flow ahead of the approaching
diffluent trough. The vertical shear continued to increase and
caused most of the deep convection to separate from the
circulation, and Danielle weakened to a tropical depression around
1800 UTC 20 August when the cyclone was located about 600 n mi
south-southwest of the westernmost Azores Islands. The now
vertically shallow cyclone continued to weaken while moving west
and west-northwestward around the southern periphery of a high
pressure system situated over the Azores Islands. Danielle
degenerated into a non-convective remnant low pressure system by
1800 UTC the next day. The remnant low moved slowly northwestward
and remained devoid of significant convection for the next 3 days.
It dissipated at 0000 UTC 25 August about 690 n mi west-southwest
of the westernmost Azores Islands.
b. Meteorological Statistics
Observations in Hurricane Danielle
(Figure 2 and Figure 3)
include satellite-based Dvorak technique intensity estimates
from the Tropical Analysis and Forecast Branch (TAFB), the
Satellite Analysis Branch (SAB) and the U. S. Air Force Weather
Agency (AFWA). Microwave imagery from NOAA polar-orbiting
satellites, the NASA Tropical Rainfall Measuring Mission (TRMM),
the NASA QuikSCAT program, and the Defense Meteorological Satellite
Program (DMSP) was also useful in tracking Hurricane Danielle.
The peak intensity of 95 kt and minimum
pressure of 964 mb at 1800 UTC 16 August is based on a combination
of subjective Dvorak satellite intensity estimates and Objective
Dvorak T-numbers at that time. A 2322 UTC 16 August TRMM composite
overpass (Figure 4) indicated Danielle possessed a small eye, but
that the cyclone had likely just passed its peak intensity based on
the erosion and warming of the cloud tops in the eastern semicircle
since 1800 UTC. Degradation of the inner core convective cloud
pattern and overall structure continued after this time based on
subsequent conventional and microwave satellite data.
There were no reports of winds of tropical
storm force associated with Hurricane Danielle.
c. Casualty and Damage Statistics
There were no reports of damage or
casualties associated with Hurricane Danielle.
d. Forecast and Warning Critique
Average official (OFCL) track errors (with
the number of cases in parentheses) for Danielle were 36 (31), 65
(29), 103 (27), 148 (25), 232 (21), 332 (17), and 452 (13) n mi for
the 12, 24, 36, 48, 72, 96, and 120 h forecasts, respectively.
Through 48 h, the OFCL errors are near or less than the average
official track errors for the 10-yr period 1994-2003 of 44, 78,
112, 146, 217, 248, and 319 n mi, respectively (Table 2). However,
the NHC track errors at 72-120 h are much higher than average for
that time period[1]. The reason for the larger track
errors at the longer time periods was due to the unanticipated turn
to the northeast and an associated decrease in forward speed that
occurred on 18 August. Nearly all of the global and regional
models, including the Florida State University Superensemble (FSSE)
model, considerably outperformed the OFCL forecasts; the exception
was the U.S. Navy NOGAPS model. The poor performance of the NOGAPS
model, which kept Danielle moving in a general west to
west-northward direction for several days, contributed to a
significant degradation in the performance of the usually reliable
consensus models GUNS (GFDL-UKMET-NOGAPS) and GUNA
(GFDL-UKMET-NOGAPS-GFS).
Average official intensity errors were 6,
11 13, 16, 23, 24, and 23 kt for the 12, 24, 36, 48, 72, 96, and
120 h forecasts, respectively. Through 48 h, these errors were near
or slightly below average (average official intensity errors over
the 10-yr period 1994-2003 of 6, 11, 15, 17, 20, 18, and 19 kt,
respectively), and higher than average at 72, 96, and 120 h. The
largest errors occurred as a result of the period of rapid
intensification that was not reflected in the official forecasts.
Those 35-40 kt underforecasts were partly due to the SHIPS
intensity model forecasting a less-than-average rate of
intensification, especially during the first 24 h of the Danielle's
existence. The SHIPS model underforecasts appear to have been due
to its reliance on intensity persistence. Large overforecasts of
30-35 kt also occurred toward the middle of Danielle's lifetime
when significant weakening was not indicated in the official NHC
intensity forecasts after the hurricane had reached its maximum
intensity. While the SHIPS intensity model did capture the
weakening trend reasonably well after Danielle peaked, its forecast
winds were too strong.
[1]Errors given
for the 96 and 120 h periods are averages over the three-year
period 2001-3.
Table 1: Best track for Hurricane Danielle,
13-21 August 2004.
Date/Time (UTC) | Position | Pressure (mb) | Wind Speed (kt) | Stage |
Lat. (°N) | Lon. (°W) |
13 / 1200 | 12.3 | 21.8 | 1009 | 30 | tropical depression |
13 / 1800 | 12.4 | 23.0 | 1009 | 30 | " |
14 / 0000 | 12.6 | 24.2 | 1009 | 35 | tropical storm |
14 / 0600 | 12.9 | 25.5 | 1004 | 40 | " |
14 / 1200 | 13.2 | 26.8 | 1004 | 45 | " |
14 / 1800 | 13.5 | 28.1 | 994 | 55 | " |
15 / 0000 | 13.8 | 29.3 | 987 | 65 | hurricane |
15 / 0600 | 14.1 | 30.8 | 981 | 75 | " |
15 / 1200 | 14.7 | 32.1 | 978 | 80 | " |
15 / 1800 | 15.2 | 33.5 | 975 | 85 | " |
16 / 0000 | 16.0 | 34.8 | 970 | 90 | " |
16 / 0600 | 16.8 | 36.0 | 970 | 90 | " |
16 / 1200 | 17.7 | 37.2 | 970 | 90 | " |
16 / 1800 | 19.0 | 38.2 | 964 | 95 | " |
17 / 0000 | 20.3 | 38.9 | 965 | 95 | " |
17 / 0600 | 21.7 | 39.6 | 970 | 90 | " |
17 / 1200 | 23.3 | 40.0 | 970 | 90 | " |
17 / 1800 | 24.6 | 40.3 | 974 | 85 | " |
18 / 0000 | 25.9 | 40.6 | 981 | 75 | " |
18 / 0600 | 27.3 | 40.3 | 985 | 65 | " |
18 / 1200 | 28.1 | 39.8 | 994 | 55 | tropical storm |
18 / 1800 | 28.9 | 38.9 | 1005 | 45 | " |
19 / 0000 | 29.3 | 37.8 | 1007 | 40 | " |
19 / 0600 | 29.7 | 37.7 | 1007 | 40 | " |
19 / 1200 | 29.9 | 37.7 | 1007 | 35 | " |
19 / 1800 | 29.9 | 37.2 | 1007 | 35 | " |
20 / 0000 | 29.8 | 36.8 | 1009 | 35 | " |
20 / 0600 | 30.2 | 37.0 | 1010 | 35 | " |
20 / 1200 | 30.5 | 37.2 | 1011 | 35 | " |
20 / 1800 | 30.9 | 37.6 | 1012 | 30 | tropical depression |
21 / 0000 | 30.9 | 38.0 | 1012 | 30 | " |
21 / 0600 | 30.6 | 38.6 | 1013 | 30 | " |
21 / 1200 | 30.7 | 38.9 | 1014 | 25 | " |
21 / 1800 | 30.5 | 39.2 | 1014 | 25 | remnant low |
22 / 0000 | 30.3 | 39.8 | 1015 | 25 | " |
22 / 0600 | 30.2 | 40.3 | 1015 | 25 | " |
22 / 1200 | 30.3 | 40.8 | 1015 | 25 | " |
22 / 1800 | 30.4 | 41.2 | 1016 | 25 | " |
23 / 0000 | 30.8 | 42.0 | 1016 | 25 | " |
23 / 0600 | 31.7 | 42.5 | 1017 | 25 | " |
23 / 1200 | 32.3 | 43.3 | 1017 | 25 | " |
23 / 1800 | 33.1 | 43.8 | 1017 | 25 | " |
24 / 0000 | 33.9 | 44.2 | 1018 | 20 | " |
24 / 0600 | 34.7 | 44.6 | 1018 | 20 | " |
24 / 1200 | 35.5 | 45.0 | 1019 | 20 | " |
24 / 1800 | 36.4 | 45.2 | 1019 | 20 | " |
25 / 0000 | | | | | dissipated |
16 / 1800 | 19.0 | 38.2 | 964 | 95 | minimum pressure |
Table 2: Preliminary forecast evaluation (heterogeneous sample) for
Hurricane Danielle, 13-21 August 2004. Forecast errors (n mi) are followed
by the number of forecasts in parentheses. Errors smaller than the NHC
official Forecast are shown in bold-face type. Verification includes the
depression stage.
Forecast Technique | Period (hours) |
12 | 24 | 36 | 48 | 72 | 96 | 120 |
CLP5 | 47 (31) | 104 (29) | 177 (27) | 260 (25) | 424 (21) | 547 (17) | 667 (13) |
GFNI | 50 (26) | 84 (24) | 107 (22) | 138 (19) | 260 (13) | 334 (10) | 433 (7) |
GFDI | 41 (30) | 78 (28) | 108 (26) | 130 (24) | 182 (20) | 244 (16) | 241 (12) |
GFDL | 43 (31) | 74 (29) | 107 (27) | 129 (25) | 162 (21) | 231 (17) | 236 (13) |
GFDN | 56 (26) | 98 (24) | 120 (22) | 132 (19) | 210 (13) | 295 (10) | 393 (8) |
GFSI | 59 (30) | 115 (28) | 172 (26) | 223 (24) | 291 (20) | 350 (12) | 495 (7) |
GFSO | 62 (31) | 115 (29) | 168 (27) | 226 (25) | 282 (20) | 348 (12) | 483 (7) |
AEMI | 50 (29) | 88 (27) | 129 (25) | 165 (23) | 200 (20) | 274 (16) | 270 (11) |
NGPI | 47 (29) | 75 (27) | 111 (25) | 165 (23) | 308 (19) | 486 (15) | 619 (11) |
NGPS | 56 (29) | 88 (27) | 111 (25) | 150 (23) | 257 (19) | 434 (15) | 551 (12) |
UKMI | 43 (29) | 75 (27) | 109 (25) | 140 (23) | 193 (20) | 209 (14) | 229 (8) |
UKM | 64 (16) | 103 (15) | 121 (14) | 148 (13) | 194 (11) | 204 (8) | 186 (5) |
A98E | 43 (31) | 89 (29) | 140 (27) | 192 (25) | 330 (21) | 452 (17) | 533 (13) |
A9UK | 37 (15) | 81 (14) | 136 (13) | 192 (12) | 257 (10) | | |
BAMD | 63 (31) | 122 (29) | 170 (27) | 206 (25) | 257 (21) | 342 (17) | 359 (13) |
BAMM | 42 (31) | 82 (29) | 119 (27) | 145 (25) | 194 (21) | 242 (17) | 287 (13) |
BAMS | 54 (31) | 105 (29) | 150 (27) | 190 (25) | 274 (21) | 350 (17) | 394 (13) |
CONU | 39 (30) | 67 (28) | 95 (26) | 124 (24) | 201 (20) | 287 (16) | 355 (12) |
GUNA | 39 (28) | 69 (26) | 100 (24) | 128 (22) | 195 (19) | 290 (10) | 295 (3) |
FSSE | 40 (27) | 74 (26) | 104 (24) | 137 (22) | 197 (18) | 281 (10) | 301 (3) |
OFCL | 36 (31) | 65 (29) | 103 (27) | 148 (25) | 232 (21) | 332 (17) | 452 (13) |
NHC Official (1994-2003 mean) | 44 (3172) | 78 (2894) | 112 (2636) | 146 (2368) | 217 (1929) | 248 (421) | 319 (341) |
|
Figure 1:
Best track positions for Hurricane Danielle, 13-21 August
2004.
Figure 2:
Selected wind observations and best track maximum sustained
surface wind speed curve for Hurricane Danielle, 13-21 August 2004.
Figure 3:
Selected pressure observations and best track minimum
central pressure curve for Hurricane Danielle, 13-21 August 2004.
Figure 4:
2322 UTC 16 August 2004 NASA TRMM microwave overpass (lower
panels) showing the small but distinct eye of relatively compact Hurricane
Danielle shortly after its peak intensity of 95 kt and minimum pressure of
964 mb. In the infrared images (upper panels), increasing vertical shear was
already becoming evident at this time as noted by the elongation of the
cirrus outflow pattern to the northeast. A weakening trend developed within
12 h after the time of this data. (image courtesy of the U.S. Navy Fleet
Numerical Meteorology and Oceanography Center, Monterey, CA).
|