Forecasting IR Satellite Imagery for Adaptive Sensor Tasking

Accurate forecasts of infrared satellite imagery are needed to optimize tasking of valuable and limited space borne surveillance assets. Atmospheric constituents such as clouds, aerosols, and water vapor can negatively impact the ability of IR surveillance systems to monitor activities on the ground. Due to their chaotic nature, atmospheric impacts are particularly severe for applications that require ground monitoring over extended time periods. In our Phase I effort we will develop a prototype 0-72 hour IR imagery forecast system with particular emphasis on characterizing the impacts of clouds. We will build on our expertise and existing capabilities in a number of relevant areas including: analysis and short-term cloud forecasting from Meteorological Satellite (MetSat) imagery; use of mesoscale Numerical Weather Prediction (NWP) models to forecast atmospheric state and diagnose cloud physical properties; application of Mie and Modified Anomalous Diffraction Approximation (MADA) theory to estimate cloud radiative properties from forecasted physical properties; calculation of in-band Top Of Atmosphere (TOA) radiance values from knowledge of the surface characteristics and forecasts of the atmospheric state and cloud properties; and visualizations of forecasted scene elements derived from the TOA radiance estimates. A final demonstration of the prototype system will be conducts using live data inputs.