Progress toward global operational utilization of the Tomorrow Microwave Sounder with the Joint Effort for Data Assimilation

Passive microwave (MW) sounders in Low Earth Orbit (LEO) garner some of the largest observational impacts on daily weather forecasts from operational centers. The Tomorrow.io Microwave Sounder (TMS) is the next generation of NASA’s Time-Resolved Observations of Precipitation structure and storm Intensity with a Constellation of Smallsats (TROPICS). Driven by cost, latency, coverage, and the aging public fleet of MW sounders, the TMS constellation of 18 satellites will provide global revisit rates on the order of 30 minutes.

In the first part of this study, we conduct several global Observing System Simulation Experiments (OSSEs) through 45 days of continuous cycling. The expected TMS impacts are compared to those from the Microwave Humidity Sounder (MHS) and the Advanced Microwave Temperature Sounder (ATMS). Forecasts and Data Assimilation (DA) use the c768 mesh of the Unified Forecast System (UFS) version of the Finite-Volume Cubed-Sphere (FV3) atmospheric model, and the Joint Effort for Data assimilation Integration (JEDI). We simulate observations with JEDI’s Unified Forward Operator (UFO) and the Community Radiative Transfer Model (CRTM). Nature run observations are simulated from the FV3-GEOS atmospheric fields provided from DYnamics of the Atmospheric general circulation Modeled On Non-hydrostatic Domains (DYAMOND) during January-March 2020. We apply 3D variational data assimilation (3DVar) using a c192 static background error covariance (BEC) implemented in the Background error on an Unstructured Mesh Package (BUMP). We use modern all-sky assimilation techniques for TMS, and plan to publicly share all of the quality control, observation error configurations, and CRTM coefficient lookup tables with the community.

In this idealized framework, two TMS instruments yield up to 3% forecast improvement out to 96 hours for water vapor and wind fields versus a configuration closer to NOAA’s operational observing network. Temperature impacts are mostly neutral. The largest improvements are at tropical latitudes, where LEO MW data is sparsest in the control experiment. In other terms, the forecast impact of a single MHS and ATMS instrument is equivalent to assimilating roughly two and four TMS satellites, respectively.

Using the same infrastructure, we further explore this space by assimilating real data from multiple TROPICS instruments in an Observation System Experiment (OSE) during 2024. There we employ 6-h UFS forecasts from the 30-member Global Ensemble Forecast System (GEFS) in hybrid 3D ensemble variational (3DEnVar) DA. Hybrid 3DEnVar ought to better-constrain clouds and precipitation with the all-sky TROPICS radiances, and will give a closer estimate of the potential operational impact of TMS.