Year: 2025
Type: article
Abstract: In recent years, high-resolution sea surface wind speed estimates in tropical cyclones (TCs) have been provided by synthetic aperture radar (SAR) retrievals. Since SAR provides only instantaneous wind speeds, an appropriate spatial averaging is needed to estimate time-mean wind speeds relevant for operational use. Currently, operational centers use 3 × 3 km 2 averaged products to estimate 1-min mean wind speeds, assuming a representative wind speed of 50 m s −1 . This study presents and evaluates two additional methods: a dynamic-box averaging and annular-sector averaging. The dynamic-box averaging adjusts the averaging box size based on estimated winds, while the annular-sector averaging samples over a dynamic wind speed–based annular-sector area taking advantage of the high degree of axisymmetry in TCs while minimizing contamination from more distant radii. Comparisons with in situ GPS dropsonde observations show that SAR retrievals agree better with the dropsonde-based 1-min mean winds rather than instantaneous winds at a 10-m altitude. All three averaging methods yield consistent 1-min mean wind estimates with small root-mean-square errors (∼4 m s −1 ) and high cross correlations (∼0.96). However, positive biases (∼10%) emerge at high wind speeds. The study also examines the statistical validity of all three methods for estimating 10-min mean wind speeds and finds that the annular-sector method performs best. The relationship between the maximum wind speeds of 1- and 10-min mean winds is also discussed. These findings demonstrate that, with appropriate averaging, SAR-derived winds can provide reliable and operationally useful TC diagnostics. Significance Statement This study introduces and evaluates methods for spatial averaging of synthetic aperture radar (SAR)-derived wind retrievals in tropical cyclones (TCs), which provide high-resolution sea surface wind estimations. Three averaging techniques to estimate the time-mean winds—static-box (3 km × 3 km), dynamic-box, and annular-sector averaging methods—are introduced and evaluated. The estimated 1-min mean winds generally match in situ GPS dropsonde measurements closely for all three averaging methods. However, notable positive biases emerge at high wind speeds. The findings highlight the annular-sector averaging method as the most statistically valid approach for estimating 10-min mean wind speeds, underscoring the potential of SAR winds to offer reliable and valuable information for TC diagnosis, contributing to improved understanding and forecasting of storm dynamics.
Source: Weather and Forecasting
Institutions: Cooperative Institute for Research in the Atmosphere, Colorado State University, NOAA Center for Satellite Applications and Research, Yokohama National University, Japan Meteorological Agency, Meteorological Research Institute, Centre National de la Recherche Scientifique, Ifremer, Laboratory for Ocean Physics and Satellite Remote Sensing, Institut de Recherche pour le Développement
Language: English
Subfield: Oceanography
Field: Earth and Planetary Sciences
Domain: Physical Sciences
SDG: Climate action
Open Access status: green
OpenAlex ID: W4415517911