Assessing the Capabilities of Smallsats Sensors Using an AI-Based OSSE/NowCasting Approach: Case of EON-MW, MicroMAS-2, TROPICS, TEMPEST-D, and CIRAS

Over the past several years, Machine Learning (ML) and Artificial Intelligence (AI), have become effective techniques for solving a wide variety of numerical problems, the main reason being that AI algorithms can perform assessments at very high spatial and temporal resolutions. This study describes an AI-based approach to assess the capabilities of several microwave (MW) and infrared (IR) sounders onboard SmallSats. Satellites observations for each instrument investigated were simulated with the Community Radiative Transfer Model (CRTM) using the NASA GEOS-5 Nature Run (G5NR) as input. Google’s TensorFlow™ ML framework was used to build and train a Deep Neural Network model to retrieve temperature and water profiles, cloud parameters, and surface temperature and spectral emissivity (used for inferring surface and cryospheric properties) from the simulated, all-sky observations. Product uncertainties were assessed against the perfectly known G5NR data. Results from the processing of Micro-sized Microwave Atmospheric Satellite-2 (MicroMAS-2), Time-Resolved Observations of Precipitation structure and storm Intensity with a Constellation of Smallsats (TROPICS), Earth Observing Nanosatellite-Microwave (EON-MW), Temporal Experiment for Storms and Tropical Systems Technology Demonstration (TEMPEST–D), and CubeSat Infrared Atmospheric Sounder (CIRAS) will be shown.