The Tomorrow.io Microwave Sounder Constellation: Pre-launch Performance Assessment and Early On-Orbit Results

Tomorrow.io, in collaboration with MIT Lincoln Laboratory, has developed the Tomorrow.io Microwave Sounder (TMS), a 6U Cubesat for all-weather temperature and humidity sensing, precipitation monitoring and nowcasting, and tropical cyclone observation. TMS is the next generation of the TROPICS (Time-Resolved Observations of Precipitation structure and storm Intensity with a Constellation of Smallsats) sensor. The first 4 TMS space vehicles (SVs) will be launched in the second half of 2024, with an additional 10 to be launched in 2025 in a mixture of polar and mid-inclination orbits. This constellation, together with a distributed commercial ground station network, will provide unprecedented revisit (< 40 min median) and low-latency (< 15 min median) passive microwave observations.

Pre-launch performance characterization for each payload was performed in a thermal-vacuum chamber utilizing blackbody targets of known temperature. This process characterized the radiometer non-linearity, noise diode output, and overall measurement noise equivalent delta temperature (NEDT) for each of the 12 TMS channels. To assess the impact of expected noise on temperature and humidity profile retrievals, an algorithm for retrieving temperature and humidity profiles using Quantile Regression Neural Networks (QRNN) was developed. The training data for this method consists of a database of approximately 10 million simulated TMS observations sampled from the NASA GMAO DYAMOND (DYnamics of the Atmospheric general circulation Modelled On Non-hydrostatic Domains) Phase II 3 km, 181 vertical level global nature run. These observations were simulated using the measured TMS beam patterns, spectral response functions, polarization angles, noise, and calibration error derived from pre-launch calibration measurements. Based on these simulated retrievals, the expected accuracy of L2A-ATVP (Atmospheric Thermodynamic Vertical Profile) temperature and humidity product is expected to have root mean square error (rmse) < 1.6 K for temperature, and < 20% relative rmse for water vapor at 3-km vertical resolution for the first three TMS SVs. These pre-launch estimates will be compared with on-orbit results once they become available.