Tropics near Real Time Atmospheric Vertical Temperature and Water Vapor Profile Retrieval

We have developed and implemented a near-real-time retrieval algorithm to estimate temperature and water vapor vertical profiles from TROPICS [1] L1b brightness temperature observations. A neural network approach, with heritage in our past work [2] in the operational Atmospheric Infrared Sounder science products [3], was selected due to fast execution time, overall accuracy, and robustness to a wide variety of meteorological conditions. Here, we describe our methodology, present initial performance results on the TROPICS Pathfinder mission on test data sets, and describe our ongoing efforts to validate the algorithm.