Study looking at the impact of dropwindsonde data obtained at different distances from the tropical cyclone center released online in Weather and Forecasting – Hurricane Research Division

An experimental version of the Hurricane Weather Research and Forecasting model (HWRF) is used to conduct the most comprehensive assessment to date of the impact of dropsonde data obtained at different distances from the center of a tropical cyclone (TC) on forecasts. The main finding is that dropsonde data within 250 km of the TC center are most important for intensity forecasts as well as forecasts of where tropical-storm-force- (34-kt) and storm-force- (50-kt) wind will occur. Particularly notable is the impact of dropsonde data on TC maximum wind-speed forecasts, as reducing sampling in the TC or its environment would degrade those forecasts.

Hurricane Hunter aircraft (Fig. 1, left and middle) are flown into and around tropical cyclones (TCs) to collect data so that hurricane specialists can understand what is happening and for use in computer forecast models so that they have the best start to make the best forecast, what we call data assimilation. The Global Positioning Satellite dropsonde (Fig.1, right), is a single-use instrument launched out of these aircraft to record pressure, temperature, humidity, and wind velocity as they fall to the Earth’s surface. 

Fig 1. (left) A photograph of one U.S. Air Force reconnaissance aircraft, (middle) a photograph of two NOAA reconnaissance aircraft, and (right) a cartoon of a dropsonde (courtesy NCAR Earth Observing Laboratory).

Recently, the most comprehensive assessment to date of the overall impact of dropsonde data on TC forecasts was conducted. It assessed the impact of dropsonde data on track, intensity, and the maximum distance of 34-, 50-, and 64-kt wind from the TC center. It found that improving how observational data are used by forecast models would likely enhance dropsonde-data impacts. In particular, it found that dropsonde data can improve TC forecasts if they are assimilated with sufficiently advanced techniques. Particularly notable is the impact of dropsonde data on forecasts of how far from the center hurricane-, storm-, and tropical-storm-force winds extend, since improving those forecasts leads to more effective TC-hazard forecasts.

Fig. 2. Plan-view experimental-design schematics of (a,d) the two sets of computer forecasts analyzed in the previous work and (b–c, e–f) the four sets of computer forecasts that denied dropsonde observations within ranges corresponding with natural breakpoints in reconnaissance sampling. Light gray shading indicates where dropsonde observations were assimilated. The corresponding radii of the gray shading are given in the top right of each subplot.

This study expands on that work to comprehensively assess the varying impacts of dropsonde data at different distances away from the TC center on forecasts during active North Atlantic basin periods over the 2017–2020 hurricane seasons. We compare the relative impact of four sets of forecasts that used dropsonde observations within ranges corresponding with natural breakpoints in reconnaissance sampling (Fig. 2b,c,e,f) to the two sets of forecasts analyzed in the previous work (Fig. 2a,d). As the only difference between these experiments was whether dropsonde data were used to start the model, differences in forecasts highlight their impact. Since TCs typically weaken when their centers are over land, this study also looks only at the set of forecasts in which neither the real nor forecast TC was over land (i.e., over-water forecasts). 

Fig. 3 – A summary graphic of the relative impact of three sets of forecasts (NO-IC – blue, did not assimilate dropsonde data within75 km of the TC center; NO-VOR – purple, did not assimilate dropsondes within250 km of the TC center; NO-ENV – yellow, did not assimilate dropsondes further than 250 km from the TC center) relative to the set of forecasts where all dropsonde data were assimilated (ALL-DROP) for both the entire and the over-water-only sample. Results are shown for five variables: (column 1) track – TRK; two measures of TC intensity: (column 2) maximum sustained wind speed at 10-m altitude – VMAX and (column 3) minimum sea level pressure – PMIN; and two significant surface wind-speed radii reported by NHC: (column 4) 34-kt wind-speed radii – R34, and (column 5) 50-kt wind-speed radii – R50. Results are also broken down by initial classification (i.e., tropical storm – TS; category 1-2 hurricane – H12; category 3-5 hurricane – H345). Shaded boxes indicate where results at short lead times (i.e., 6-60 h) showed improvement (purple) or degradation (blue) relative to the set of forecasts all available dropsondes were assimilated.

■ Important Conclusions:

• Removing dropsonde data anywhere, particularly from within the vortex itself, substantially degraded forecasts of maximum sustained surface wind speed (shown by the blue and dark blue in Fig. 3b,g).
• In many instances, removing in-vortex dropsondes also degraded forecasts of the extent of the tropical-storm-force (34-kt) wind and storm-force (50-kt) wind (shown by the blue and dark blue in Fig. 3d,i,e,j). As such, in-vortex dropsondes contribute to a majority of the overall impacts of the dropsonde observing system.
• Track forecasts of weak TCs benefited the most from environmental sampling (shown by the purple in Fig. 3a,f), while track forecasts of strong TCs benefited the most from in-vortex sampling (shown by the purple in Fig. 3a,f).
• Only sampling the inner 75 km of a TC (i.e., the inner core) should be avoided (not shown). This result supports a change made to the U.S. Air Force Reserve’s sampling strategy in 2018 whereby dropsondes were released outside of the inner core at the ends of flight legs.

For more information, contact aoml.communications@noaa.gov. The paper can be found online at https://journals.ametsoc.org/view/journals/wefo/aop/WAF-D-23-0055.1/WAF-D-23-0055.1.xml. This research was carried out in part under the auspices of the Cooperative Institute for Marine and Atmospheric Studies (CIMAS), a Cooperative Institute of the University of Miami and the National Oceanic and Atmospheric Administration, cooperative agreement NA20OAR4320472 while the lead author (Sarah Ditchek) was supported by the FY18 Hurricane
Supplemental (NOAA Award ID NA19OAR0220188). The scientific results and conclusions, as well as any views or opinions expressed herein, are those of the author(s) and do not necessarily reflect those of OAR or the Department of Commerce