MACHINE LEARNING-BASED ALGORITHM FOR SEA LEVEL PRESSURE RETRIEVAL FROM THE JOINT OBSERVATIONS OF 50-60 AND 118 GHZ SPACEBORNE MICROWAVE RADIOMETERS

Z. Zhang¹

¹National Space Science Center, Chinese Academy of Sciences

Accurate sea level pressure (SLP) data are critical for numerical weather forecasting, monitoring and forecasting of tropical cyclones (TCs), and research on climate change. In 2018, we proposed a novel method to retrieve SLP from spaceborne O2 band microwave radiometers. Retrieval algorithms were established for the 50-60 GHz and 118 GHz microwave radiometers, respectively. The O2 absorption characteristics of the 60 GHz absorption band and 118 GHz line are very different, and therefore retrievals that combine observations at both bands are expected to provide more information on SLP, potentially improving retrieval performance. The Microwave Temperature Sounder-2 (MWTS-2) and Microwave Humidity Sounder-2 (MWHS-2), onboard the Chinese FY-3 satellite (FY-3C and FY-3D), offer synchronous (both temporal and spatial) dual O2-band observations. This synchronization provides a valuable opportunity to enhance SLP retrievals through joint applications of 60 and 118 GHz observations.

This study develops a neural network-based algorithm to retrieve SLP fields from combined observations of brightness temperature (TB) in both frequency bands. Application studies were carried out using the joint observations from the MWTS-2 and MWHS-2 onboard the FY-3D satellite. Optimal frequency channels for the SLP retrieval were selected based on an information content analysis. SLP retrievals were compared with SLP data from the ERA-Interim reanalysis and the Global Data Assimilation System Final Analysis (GDAS/FNL). For low-to-medium wind conditions, the retrieval accuracies are 1.86 hPa under clear-sky conditions, 2.34 hPa under cloudy conditions and 3.20 hPa under rainy conditions. For TC conditions, the retrieval accuracies are 2.78 hPa for tropical depressions and storms, 4.23 hPa for hurricanes or typhoons, and 6.15 hPa for major hurricanes or severe typhoons. Furthermore, comparative experiments between the new algorithm and the previous algorithm that used only observations from a single oxygen band demonstrated that the new algorithm provides more comprehensive SLP information and achieves higher accuracy.