High Resolution Cloud Analysis and Forecast System

work, and potential uses of this system. Clouds have a significant impact on United States Air Force operations, and as such the analysis and forecasting of clouds is one of the highest priority missions of the Air Force Weather Agency (AFWA). Currently, AFWA uses a combination of three Defense Meteorological Satellite Program satellites and one NOAA Polar-orbiting Operational Environmental Satellite (POES) to generate a gridded automated cloud analysis with a grid-spacing of approximately 48 km. A major acquisition program, the Cloud Depiction and Forecast System II, will upgrade AFWA's hardware and software to use both polar-orbiting and geostationary satellites to produce an analysis based on algorithms developed under the Support to Environmental Requirements for Cloud Analysis and Archive (SERCAA) program. Unlike today's operational algorithms, the SERCAA algorithms perform cloud detection at the pixel-level using all channels and spectral information from the electro-optical imagers on the operational satellite systems. CDFS II will provide worldwide cloud analyses and forecasts with an approximate grid spacing of 24 km, or 4-times higher than the current products. However, given the need for higher resolution regional products, an effort was put forth to develop a capability to use fine-mode (0.5 km resolution) data from DMSP and Local Area Coverage (LAC, 1 km at nadir) data from POES to produce a regional analysis with a 6 km grid-spacing. Additionally, the Air Force Weather Agency modified their semi-Lagrangian cloud advection model (ADVCLD) to use these analyses to produce 6 km forecasts out to 30 hours. This is 64 times the spatial resolution of today's products!

The results of this effort proved the feasibility of generating the higher resolution regional products. Products generated from the prototype, which now runs in an automated mode 24 hours per day, are being made available to USAF customers for evaluation purposes. The system has been producing automated cloud analysis from 4 DMSP satellites since March 1999 and 2 NOAA satellites since June 1999. Additionally, this marked the first quantitative use of DMSP fine-mode data for automated cloud analysis, as well as the first worldwide implementation of the SERCAA cloud analysis algorithms. It also provided an opportunity to identify several opportunities for improvement to the SERCAA analysis algorithms as well as the unique problems created when using them at higher resolution than originally planned. The use of an advection model to generate mesoscale forecasts from synoptic scale model output also proved to be quite challenging, and an effort was put forth to use output from the AFWA operational version of the MM5 model as the input to the ADVCLD model. This poster will describe the algorithms used and will show some of our results. Additionally, we will address some of the challenges, futur