JP1.42
Defining uncertainty in hurricane maximum surface wind estimation
Stephanie Ann Mullins, University of Louisiana - Monroe, Monroe, LA; and P. G. Black, C. S. Velden, M. D. Powell, E. W. Uhlhorn, T. L. Olander, A. Burton, and J. L. Beven
Over the past 50 years, the technology used to estimate the maximum surface wind speed in a hurricane has progressed as satellite and airborne remote sensing technology has advanced. Aircraft instrumentation has improved, ground-based and airborne Doppler radar technology has advanced, and various satellite techniques have advanced. The airborne Stepped Frequency Microwave Radiometer (SFMR) is a new technology that has matured over the past 20 years and now provides a new tool for hurricane forecasters to use in the Atlantic to estimate peak surface winds. The first real time transmission of SFMR data to Atlantic forecasters occurred during Hurricane Dennis in 1999 and in 2005 they began using this technology operationally from NOAA aircraft in 2005. Operational reconnaissance WC-130J aircraft from the Air Force Reserve Command (AFRC) 53rd Weather Reconnaissance Squadron will began reporting SFMR surface winds during the 2007 season, in addition to a continuation of this service from the NOAA-AOC WP-3D aircraft, with the full fleet of 10 AFRC aircraft equipped in 2008. With the new technology has come the realization of the uncertainty in this crucial measurement as various measurement technologies often produce conflicting estimates. By measuring microwave emissions from the sea surface, the SFMR retrieves a surface wind speed estimate as well as a simultaneous estimate of rainfall. Reliable and consistent calibration of the SFMR system has been maintained since 1998 and a nine year data base created through 2006 following development of a new wind algorithm in 2005. Calibration of SFMR winds using GPS dropsonde estimates of the 10-m sustained surface wind has produced consistent estimates of maximum surface winds over this period, providing a consistent hurricane maximum surface wind data base for the first time.
It is the objective of this study to reduce the overall uncertainty in the hurricane maximum wind estimate by comparing these new SFMR surface wind measurements with archived maximum wind values in the NHC Best Track data base and as well as with maximum winds estimated from the satellite-based Dvorak Technique (DT), the .9 method and the new SFMR-based method of estimating maximum surface winds from flight level maximum winds. The wind speeds from the SFMR on 53 flight patterns flown into 17 tropical cyclones, consisting of 178 radial flight legs, from 1998 to 2006, are used as a basis for estimating the range of uncertainty amongst the various types of surface wind speed estimation. Surface winds from these various sustained surface wind speed estimates are evaluated to determine how closely they match the surface wind speeds reported by the SFMR. The pressure-wind relationship recently fit to Best Track data is used to compute maximum winds from observed minimum pressures which were then compared with the interpolated Best Track and Dvorak (DT) maximum winds and the SFMR and Powell et al maximum winds.
The statistical distribution of the data used in this study suggest that Best Track maximum wind estimates overestimate SFMR maximum surface winds by 6 m/s on average for tropical storm through Category 4 tropical cyclones, representing a 12-15 % overestimate, or one Saffir-Simpson storm category on average. The data distributions for Category 5 hurricanes show only small differences. The Powell et al. method estimates SFMR maximum surface winds from flight level maximum winds to within ± 3 m/s, the same random error for SFMR winds relative to GPS dropsonde wind measurements. The maximum surface wind estimates from Atlantic pressure-wind relations all overestimate SFMR maximum winds by 6 m/s on average, similar to the Best Track estimates to which they were fit. The issue of under-sampling maximum wind values due to flight tracks missing the true wind maximum is evaluated using two years of processed airborne Tail Doppler radar data, with early indications suggesting that this may only account for 2-3 percent of the observed differences. The present study suggests the possibility of reducing the uncertainty in hurricane maximum surface wind estimates to the level of scatter in the SFMR vs GPS sonde comparisons by tuning future Best Track and satellite DT maximum wind estimates to those produced by SFMR measurements.
Uploaded Presentation File(s):Joint Poster Session 1, Tropical Cyclones and Probability/Statistics Posters
Monday, 21 January 2008, 2:30 PM-4:00 PM, Exhibit Hall B
Previous paper Next paper