IMPACT ON HURRICANE TRACK AND INTENSITY FORECASTS OF GPS DROPWINDSONDE OBSERVATIONS FROM THE SYNOPTIC SURVEILLANCE MISSION INTO HURRICANE MITCH ON 25 OCTOBER, 1998.

Sim D. Aberson

Hurricane Research Division
Atlantic Oceanographic and Meteorological Laboratories
National Oceanic and Atmospheric Administration 4301 Rickenbacker Causeway
Miami, Florida


Click here for N49RF transmitted sonde data.
Click here for N49RF HSA sonde data.


1. Synoptic situation

Hurricane Mitch developed from an easterly wave that moved off the coast of Africa 10 October, but did not develop into a tropical depression until reaching the western Caribbean Sea on 22 October. The system moved slowly northward, strengthening rapidly into a hurricane by 25 October. Due to the potential threat to Southern Florida if the northward motion continued as suggested by the guidance, a synoptic surveillance mission was tasked for nominal time 25 October 1998 0000 UTC, with a follow-on mission the following day. At that time, Mitch was moving slowly westward and was located in the west-central Caribbean Sea midway between the northeastern coast of Honduras and Jamaica. Mitch was located to the south of the subtropical ridge, and the anticyclonic gyres surrounding it the previous day hav dissipated, all suggesting continued motion toward the west (Fig. 1).


2. General Assessment of dropwindsonde impact

A. GFDL model

Figure 2 shows the GFDL forecast tracks for Hurricane Mitch, and Table 1 shows the errors and impact of the synoptic surveillance mission. The results are mostly positive, with improvements at all times except 36 and 48 h. The dropwindsonde data initially moved the forecast track to the left, and also slowed the track, allowing for the improvements.

Mitch made landfall 84 h into the forecast on the north coast of Honduras east of La Ceiba, Atlantida. Neither the run with nor the run without the dropwindsonde data forecast landfall.

B. VICBAR

Figure 3 shows the VICBAR forecast tracks for Hurricane Mitch, and Table 2 shows the errors and impact of the synoptic surveillance mission. The dropwindsonde data degraded the forecast at all forecast times except for a slight improvement at 120 h. The dropwindsonde data pushed the storm fasther and to the west. Neither forecast was qualitatively good.

Mitch made landfall 84 h into the forecast on the north coast of Honduras east of La Ceiba, Atlantida. The VBAR forecast with the dropwindsonde data forecast landfall south of Cozumel on the Mexican coast of Yucatan, 348 km away. The forecast without the dropwindsonde data forecast landfall in Western Cuba near Galafre, Pinar del Rio, 728 km from the actual lanfall point.

C. GSM

Figure 4 shows the GSM forecast tracks for Hurricane Mitch, and Table 3 shows the errors and impact of the synoptic surveillance mission. The dropwindsonde data had a negative impact on the GSM forecast track at all forecast times. The dropwindsonde data pushed the forecast to the right and made the storm move slower than in the forecast without the additional data. This allowed the storm to get picked up in the weakness in the subtropical ridge more quickly than in the run without the dropwindsonde data.

Mitch made landfall 84 h into the forecast on the north coast of Honduras east of La Ceiba, Atlantida. The run with the dropwindsonde data made landfall near Jaguey Grande, Matanzas, Cuba, 816.3 km to the northeast. The run without the dropwindsonde data made landfall near Los Palicios, Pinar del Rio, 757.9 km to the northeast.

D. Intensity

Table 4 shows the GFDL intensity forecast errors and impact of the synoptic surveillance mission. The dropwindsonde data had a positive impact at all times except 12, 48, and 120 h. Tuleya and Lord (1998) also showed modest improvements to GFDL model intensity forecasts in the HRD synoptic flow cases.


3. Targeting

The atmosphere has long been recognized as a chaotic system (Lorenz 1963), e.g. very small perturbations to initial conditions result in increasingly large differences in the evolution of the atmosphere with time. Since the exact state of the atmosphere can never be measured, all analyses contain errors whose magnitudes can only be estimated. An indeterminate number of initial conditions consistent with the observational data can therefore be used in numerical weather prediction, and single model runs at any synoptic time only give one possible solution to the evolution of the atmosphere. Many operational forecast centers around the world, therefore, now employ ensemble forecasting as a means of quantifying the uncertainty in the evolution of the atmospheric system. Small perturbations from a best "control" state are calculated and added to and subtracted from this control to allow for different integrations starting from theoretically equally likely initial states. These perturbations are designed so as to mimic the fastest growing modes in the model and to create the largest envelope of possibilities in the forecast. Therefore, they generally correspond to locations where large analysis errors will most impact the forecast. Those features corresponding to perturbations which most impact the tracks of tropical cyclones must be found, properly sampled, and thoroughly tested, to prove the efficacy of targeting techniques.

Figure 5 shows the variance of the size of the perturbations in the National Centers for Environmental Prediction (NCEP) global model ensemble forecasting system (Toth and Kalnay 1993). The largest maxima, over the Great Plains and east and southeast of Bermuda, are associated with shortwave troughs rotating through the midlatitude westerlies. The other maximum northeast of the Leward Islands is associated with an upper-level area of low pressure. Only small masima are close to Mitch. The two maxima north of Puerto Rico and Haiti represent the axis of the eastern end of the subtropical ridge. The maximum just north of Panama is associated with the weakening ridge line south of Mitch, and the weaker maximum in the northwestern Gulf of Mexico represents Mitch itself. The dropwindsonde data did not fully sample any of these maxima.

Figure 6 shows the difference in the 850 - 200 hPa averaged winds between the runs in which all the dropwindsonde data and none of the dropwindsonde data are included. The differences nearly encircle Mitch except to the east of the storm center. The northernmost dropwindsondes have only slight impact on the initial conditions, and these are all located in or near minima in the ensemble spread. The maximum difference is located just to the west of the hurricane, and the maximum is well outside the locations of the dropwindsondes data. In fact, the east-to-west line of dropwindsondes to the south of Mitch is represented by a minimum of impact, surrounded by maxima not represented by the data. A secon maximum is located in the Yucatan Straights. Since the response may not be physically meaningful, and since it is located in a region of moderate ensemble spread, these differences are expected to grow, and negatively impact the forecast.

Figure 7 shows that, by 24 h into the forecast, the largest difference between the forecasts with and without the dropwindsonde data has amplified and surrounded the hurricane, whereas the second maximum, initally located in an area of minimal ensemble spread, has moved over the Yucatan Peninsula and decayed. Since this initial difference may be at least partially spurious, this led to the degradation of the forecast. These results seem to confirm that the largest (smallest) ensemble perturbations correspond to amplifying (decaying) modes in the model.


4. Conclusion

The dropwindsonde data obtained during the synoptic surveillance mission for Hurricane Mitch at nominal time 25 October 1998 0000 UTC has provided mainly negative results. The MRF ensemble forecasting system suggests, and the model runs confirm, that data around the hurricane itself and in the weakness in the subtropical ridge north of the hurricane have the greatest positive impact on the forecast. However, none of these regions were fully sampled, and that led to possible spurious spread of the influence of the dropwindsonde data into unstable regions which greatly negatively impacted the forecasts.


Tables

Table 1
Track forecast errors for the no dropwindsonde GFDL control (GFNO), the all dropwindsonde run (GFAL), and the run with only targeted observations (GFTG), and the percent improvement of the latter two over the control.
Forecast
time (h)
GFNO
Error (km)
GFAL Error (km)
(% Improvement)
1278.78.( 0%)
24119.113.( 5%)
36123.149.(-21%)
48149.172.(-15%)
72194.192.( 1%)
84275.219.( 20%)
96363.269.( 26%)
120769.615.( 20%)

Table 2
Track forecast errors for the no dropwindsonde VBAR control (VBNO), the all dropwindsonde run (VBAL), and the run with only targeted observations (VBTG), and the percent improvement of the latter two over the control.
Forecast
time (h)
VBNO
Error (km)
VBAL Error (km)
(% Improvement)
1263.86.( -37%)
2415.183.(-1120%)
36113.291.( -158%)
48261.452.( -73%)
72444.667.( -50%)
84573.744.( -30%)
96726.819.( -13%)
108832.903.( -9%)
120998.978.( 2%)
Landfall728.348.( 52%)

Table 3
Track forecast errors for the no dropwindsonde GSM control (GSNO), the all dropwindsonde run (GSAL), and the run with only targeted observations (GSTG), and the percent improvement of the latter two over the control.
Forecast
time (h)
GSNO
Error (km)
GSAL Error (km)
(% Improvement)
1263.64.( -2%)
2415.54.(-260%)
36113.161.( -42%)
48261.316.( -21%)
72444.467.( -5%)
84573.612.( -7%)
96726.805.( -11%)
108843.987.( -17%)
120998.1183.( -19%)
Landfall758.816.( -8%)

Table 4
Intensity forecast errors for the no dropwindsonde GFDL control (GFNO), the all dropwindsonde run (GFAL), and the run with only targeted observations (GFTG), and the percent improvement of the latter two over the control.
Forecast
time (h)
GFNO
Error (kn)
GFAL Error (kn)
(% Improvement)
12-55-56(-2%)
24-72-70( 3%)
36-63-60( 5%)
48-45-46(-2%)
72-19-18( 5%)
844 4( 0%)
961613(19%)
1204345(-4%)


Return to 1998 assessment page
Return to HRD home page