The National Weather Service Graphical Forecast Editor (GFE) in AWIPS acts a bridge between data models and forecast products. It provides the capability for meteorologists to edit grid-based data, produce a wide variety graphical forecasts and then output grid-based text products. As part of the Rapid Prototyping Project, the Tampa Bay Area NWS Office has participated in development of the GFE with NOAA Forecast Systems Laboratories. The system is enhanced by flexible Python scripting permitting extensive localization.

Forecasters must perform considerable grid manipulation to provide meaningful and accurate public, marine, and fire weather forecasts in one to six hour increments out to seven days. Methodology is very important to provide forecasters a clear concise routine with sound scientific basis for producing products. In this paper, we describe such a methodology showing the flow from model data to forecast products and how much of the basic editing is accomplished with GFE Smart Tools that use key fields to derive other fields. For example, from the Weather field, we can derive estimates of Sky Cover, QPF, POP, and Lightning Activity Level (LAL).

The software's scripting and grid manipulation capability extends beyond this by providing methods to derive new forecast parameters. For example, a wind regime-based lightning climatology is introduced into the GFE as a grid that forecasters may use as a basis for forecasting weather fields. The scripting capability and access to model data also permits derivation of simple models to produce new fields or enhance existing ones.