Statistical Analysis for Performance of Detection and Imaging of Dynamic Targets Using the Geostationary Interferometric Microwave Sounder

Abstract

Passive microwave remote sensing at 50–56 GHz is used for retrieving atmospheric temperature profiles from radiometric observations. The profiles can be used to investigate internal structure of tropical cyclones (TCs). Recently, a number of geostationary Earth orbiting (GEO) microwave instruments at 50–56 GHz have been proposed with the potential for continuous observation of fast-changing weather from GEO orbit. Since microwave application in GEO is a relatively new research area and microwave observation data from GEO are still unavailable, this study is intended to simulate brightness temperatures (BT) for GEO observation using weather forecast and radiative transfer model. A sequence of BT maps for full-earth disk containing typical observation target TCs are fed to GIMS simulation system and statistical BT variation pattern for TCs is analyzed based on a study of all TCs that occurred in Western Pacific Region in 2013. This is important because such simulation can serve as a preresearch to determine image distortion for dynamic target observation of GIMS and results can guide time-related parameter configuration for GEO microwave instrument. Results show that observing targets with rapid BT changes such as TCs with 30-min imaging period will increase root-mean-square error (RMSE) of retrieved image by about 0.003–0.016 K and 0.05–0.17 K for full-earth disk and TC area, respectively, at 50–56-GHz observation frequencies. Within 50–56-GHz range, higher frequencies tend to have smaller retrieved image RMSE than lower frequencies do. Statistical BT change for higher grade of TC is about 0.25–3 K greater than lower grade of TC in 6 h. The distinction among different grades of TC and observation frequencies indicates time-related parameters such as imaging period can be properly adjusted according to operational frequency channels as well as grade of the TC being monitored.