The overall objective of this project is to develop a method for extracting time derivative information from geostationary meteorological satellite imagery for the purpose of improving numerical weather prediction. In Phase I, we shall undertake to carry out a proof-of-concept study to demonstrate the feasibility of using pattern recognition techniques and a statistical cloud classification method to estimate time rates of change of large-scale meteorological fields from remote sensing data. For this purpose, we shall analyze visible and infrared geostationary satellite images. The cloud classification methodology will be based on Typical Shape Function analysis of parameter sets characterizing the cloud fields. An idealized numerical weather prediction model will be used to test the potential value of the concept for improving modeled physical processes using the differences between observed and computed time rates of change.Forecast verification studies, in conjunction with predictability theories, show that the early hours of numerical weather prediction are typically characterized by a rapid decay of skill. If observational estimates of time derivatives can be compared with numerical computations during this critical period, a potentially valuable new tool will have been made available for improving numerical weather prediction models.STATUS: Phase I Only
