Evaluation of cirrus cloud predictions from the MM5 and WRF/NAM weather models

Cirrus clouds are known to impact certain Air Force missions and systems. High altitude reconnaissance and laser propagation for communications and missile defense are activities that are currently or are expected to be affected by the presence of cirrus. Therefore, it is critical that accurate predictions of cirrus locations, altitudes, coverage and water content be available to mission planners 24-36 hours in advance. This project documents the ability of two operational mesoscale numerical weather prediction models, and a model application executed as a post-processor to one of them, to provide realistic forecasts of the cirrus characteristics required for Air Force mission support. For this project, cirrus clouds are defined as all ice clouds with tops exceeding 6 km in altitude. Approximately 30 days of 24- and 36-hour ice cloud forecasts from the Air Force Weather Agency operational version of the MM5 and from the National Centers for Environmental Prediction WRF/NAM mesoscale weather models were compared with GOES satellite imagery analyses of ice clouds in Fall 2006 and Winter and Spring 2007. In addition, cirrus forecasts from the AFWA Diagnostic Cloud Forecast algorithm, applied to MM5 forecasts, were included in the evaluation. Statistics from these comparisons include contingency tables of Yes/No cirrus predictions vs. Yes/No cirrus analyses by location and time, and assessment of areal coverage, altitude and water content. Preliminary results from an earlier comparison involving only observed cirrus events showed that the MM5 model was able to correctly predict the presence or absence of cirrus overhead for about 2/3 of the locations/times of comparisons. Statistics shown at the conference will be based on comparisons made without regard to observed cirrus conditions, allowing for adequate accounting of both missed forecasts and false alarms.