The PIs will draw on newly available measurements to determine the limits of model validity by comparing predictions with a representative set of direct, large-scale maps of high-latitude convection. The database, furnished by the SuperDARN system of HF radars, has demonstrated that these measurements resolve the large-scale features of convection that the models predict. The PIs will select a number of SuperDARN archival events as a testbed for model validation, then generate the predictions of the models for the selected events and devise evaluation schemes for gauging their success. The research will produce a characterization of the predictive value of convection models, recommendations for model improvement, and methods of combining models and direct measurements to optimally map convection.