2020, 40(3): 364-375.
doi: 10.11728/cjss2020.03.364
Abstract:
Sea level pressure is an important meteorological factor and plays a key role in Numerical Weather Prediction (NWP), tropical cyclone forecasting and solar activity studies. However, until recently, sea level pressure data have mainly been provided by in-situ measurements. It is of great significance to obtain sea level pressure with a high spatial and temporal resolution by means of remote sensing. In this study, we investigated the retrieval of sea level pressure over mid-to-low latitude (40°S-40°N) regions using the observations from the Microwave Humidity and Temperature Sounder (MWHTS) onboard the Fengyun-3C (FY-3C) satellite. Sea level pressure sounding is achieved by MWHTS 118.75GHz channels due to their ability to measure the total columnar oxygen absorption. The sensitivity of the MWHTS 118.75GHz channels to surface pressure was analyzed using the radiative transfer equation. Compared with the channels far into the oxygen absorption band, the channels lie on the wing of the band are more sensitive to the change of surface pressure. A statistical retrieval algorithm based on the Back-Propagation (BP) neural networks was established. In-situ buoy measurements and reanalysis data were used to assess the retrieval performance. Results showed that the proposed retrieval algorithm can estimate sea level pressure over mid-to-low latitude areas (40°S-40°N) with the accuracy of 2.0, 3.0, and 3.5hPa for clear-sky, cloudy and rainy conditions, respectively. In addition, several tropical cyclone retrieval experiments showed that the proposed method was useful in the early identification of tropical depression.