An OSCAT Simultaneous Wind/Rain Geophysical Model Function

Conventional near-surface vector wind (VW) estimates from microwave scatterometers degrade in quality in the presence of rain because the standard wind-only (WO) geophysical model function (GMF) does not model rain effects. The WO GMF maps the noise-free normalized radar cross-section ($\sigma ^\circ$) to the near-surface VW but omits other geophysical phenomena, such as rain rate ($R$). Previous studies have developed simultaneous wind/rain (SWR) GMFs that map $\sigma ^\circ$ to the near-surface VW and $R$ to account for rain effects [1]. This article develops an SWR GMF for the Ku-band OceanSat-2 Scatterometer (OSCAT) using a $\sigma ^\circ$, VW, and $R$ triple collocated database. OSCAT $\sigma ^\circ$ measurements are collocated with Tropical Rain Measuring Mission (TRMM) near-surface $R$, and European Centre for Medium-Range Weather Forecasts (ECMWFs) numerical weather prediction near-surface VW product creating the OSCAT, TRMM, and ECMWF Database (OTED). Four novel approaches for creating an SWR GMF are discussed and the most accurate GMF is found to be the path integrated attenuation method. The estimate accuracy is analyzed using the OTED, which reveals that the SWR wind speed estimate accuracy is between 0.25 and 0.7 m/s more accurate (depending on the rain rate) on average compared to WO speed estimates. However, wind direction estimates are not improved by SWR with WO outperforming the SWR estimates by between 1$^\circ$ and 10$^\circ$ (depending on the $\sigma ^\circ$ spatial resolution).