Tracking of Hydrometeors State in Severe Weather using LEO CubeSat Fleet

A constellation of ~30 Low-Earth Orbit (LEO) microwave sensors based upon CubeSat specifications has been identified as an achievable cost-competitive means to realize the Precipitation and All-weather Temperature and Humidity (PATH) mission goal, which requires observations of mesoscale severe weather, such as hurricane, with ~15-30 minutes temporal resolution and ~15-30 km spatial resolution. To this end a 3U CubeSat microwave spectrometer operating at both the 118.7503 GHz oxygen channels and the 183.31 GHz water vapor channels is being developed at CU to provide a spatial resolution of ~16 km from a 425 km altitude orbit. Meanwhile, efforts have been carried out to develop observing system simulation experiments (OSSEs) aiming at exploring the potential of hydrometeors state tracking in a severe weather event using the LEO fleet concept. A technique of microwave radiance assimilation based on iterative extended Kalman filtering and use of the Unified Microwave Radiative Transfer (UMRT) model with fast Jacobian calculation plays a key role in the OSSEs to facilitate tracking of the precipitation state variables. The technique will provide effective assimilation of LEO constellation data into a regional-scale Numerical Weather Prediction (NWP) model. An empirical state-dependent background error covariance model based on a rapidly computable covariance function separable in the vertical and horizontal coordinates, along with a baseline covariance lookup library derived from aircraft and satellite active radar observations, are critical components to the all-weather assimilation cycle. The background error covariance model will function within a scalable local 3DVar iterative assimilation step. A demonstration of the stabilized tracking system of hydrometric state using simulated microwave radiance data for Hurricane Sandy in the framework of LEO fleet concept will be presented, along with the prediction of assimilation process in real time based on computational complexity analysis of forward UMRT model and extended Kalman filtering. The degree to which the NWP model is able to lock on to simulated hurricane radiance data and maintain hydrometric locking status will be demonstrated as well.