As real-time numerical forecasting at grid resolutions of less than 50 km becomes increasingly common, there is a growing need to develop new methods of mesoscale forecast verification. Traditional statistical verification scores often reward successful prediction of smooth, large-scale features, and do not accurately assess forecast skill with respect to smaller-scale cloud structure and moisture gradients.

Since May 2002, SSEC/CIMSS at the University of Wisconsin-Madison has been running the PSU/NCAR MM5 once daily over the Southern Great Plains DOE Atmospheric Radiation Measurement site. During this time, MM5 forecasts of boundary layer temperature and water vapor have been continuously validated against retrievals from five ground-based Atmospheric Emitted Radiance Interferometer (AERI) instruments. More recently, brightness temperatures derived from MM5’s radiation parameterization are being validated against GOES-08 imager band 4 (11 micron) brightness temperatures.

In this paper, we present the results of comparison between MM5 and the above-mentioned remote-sensing instruments. Using a number of case studies as examples, we demonstrate the utility of AERI and GOES-08 as tools for the verification of mesoscale boundary layer and cloud structure respectively. We then suggest a number of techniques that might be used in long-term verification of cloud and boundary layer structure on the mesoscale.