Future geostationary and low earth orbit environmental observational systems will substantially increase the volume and timeliness of data provided to the user communities. A multitude of new and improved products will be available as a result of increases in temporal, spatial, and spectral coverage and resolution of future remote sensing systems. Combined with the evolution of numerical weather prediction (NWP) models and the resulting increase in model data, users will face unprecedented challenges to utilize all relevant data to make timely and appropriate decisions relating to severe weather events (thunderstorms, tornadoes, floods, etc.). Automation of the identification, synthesis, integration, and analysis of observations and numerical forecasts within decision support processes is critical to maximizing the societal benefits of future observational systems.

The Weather Information and Decision System (WxIDS) is a concept for implementing an automated decision support system that synthesizes observations and probabilistic forecasts with user operational experience and risk criteria designed to meet user requirements. There are four components to the WxIDS concept; the Local Data Manager manages weather information acquisition and storage; the Forecast Generation System identifies critical situations, combines NWP model predictions and observations into comprehensive probabilistic forecasts aimed at user requirements; the Decision Support System combines probabilistic forecasts with user operational experience and risk criteria for decision advice and recommendations; and the WxIDS User Interface provides interactive capabilities to users or their risk management decision system.

We present the results of a prototype of the WxIDS system focused on utilizing ensemble forecast probabilities, GOES satellite observations, and radar data to forecast and provide early warnings of severe convection. Bayesian methods coupled with the ensemble forecasts provide warnings of potential regions for severe convection 6, 12, and 24 hours prior to onset. These probabilistic forecasts identify areas with a high probability for severe convection to be monitored by GOES satellite observations for convective initiation. Radar data are used to verify both the ensemble forecasts and the GOES identified convective initiation cases (those that develop into severe convective storms). This prototype represents the initial development of the Local Data Manager and Forecast Generation System components of WxIDS.