ON THE DEVELOPMENT OF "APPLICABLE" MODELS FOR FLOOD FORECASTING