Cloud Forecast Simulation Model.
Abstract:
The cloud forecast simulation model generates synthetic worldwide 3- hour total cloud cover forecast fields at 50-nautical mile resolution. The synthetic forecasts are generated stochastically, based on input verifying observed total cloud cover fields, in such a manner that the agreement between the synthetic cloud forecast field and its verifying observed field is no better or worse than the agreement between actual cloud prognoses and their verifying observations. Moreover, a sawtooth wave submodel is used to insure the synthetic cloud forecast fields have the same spatial correlation as actual cloud prognoses have. Thus, the cloud forecast simulation model generates synthetic total cloud cover forecast fields that have the same skill and spatial correlation as the operational forecast product has. The model is used to generate meteorological input to system planning and optimization simulations and system design studies. The sawtooth wave submodel could also be used to generate synthetic two-dimensional observed weather fields as well as cloud forecasts.