Validating Tropical Cyclone Winds From Microwave Radiometers

Abstract:

Our presentation will discuss the work that has been done to validate the new Remote Sensing Systems (RSS) Tropical Cyclone Winds (TC-Winds) products. These products were created using newly developed algorithms that allow for microwave retrievals of high winds in tropical cyclones (TCs), including in areas of heavy rain.

The first such product is the Soil Moisture Active Passive (SMAP) global daily winds, released in 2017. Due to the low frequency of the observed signal, the SMAP retrievals are mostly unaffected by rain, and they retain signal sensitivity for hurricane winds ranging from Category 1-5. Recently, the SMAP winds have been used to calibrate a new algorithm for the Windsat and AMSR-2 radiometers, which are based on higher frequency channels that are affected by rain. By using a combination of channels, the rain effect can be greatly reduced. These new products are specifically developed for TCs and are identified as TC-winds to distinguish them from the standard global wind product. The TC-winds are now produced for SMAP, Windsat, and AMSR-2 in near-real time and freely distributed to the public.

We will present results of comparisons between TC-winds observed by these three sensors and winds from the Hurricane Weather Research and Forecasting Model (HWRF), which provides analysis and forecasts of TCs around the globe. Comparisons of satellite data to other ground truth in TCs is challenging due to the different spatial scales of the two. For our purposes, the finest HWRF grid at each analysis time is resampled from its original resolution (~1.5km) to a resolution of 40km, consistent with the satellite products. We employ several different resampling methods to the HWRF data and examine the differences between them. Once HWRF has been resampled for a given storm, we look for all SMAP, Windsat, and AMSR-2 passes over that storm. The model is then interpolated in time and space to the sensor pass and the two are compared.

We examine how the sensors and HWRF compare in regards to storm size and storm organization. We also look at what scaling factors describe the resampling from high-resolution to that of typical satellite observations. Similar studies are performed comparing the microwave TC-winds to the high-resolution winds observed by the Synthetic Aperture Radar (SAR) onboard platforms such as Radarsat-2 and Sentinel.