Extended Abstract (44K)5.1
Initializing Gridded Forecasts from Numerical Models
Tom J. LeFebvre, NOAA/FSL, Boulder, CO; and M. Romberg and T. Hansen
The Graphical Forecast Editor Suite (GFESuite) comprises the grid editing component to the Interactive Forecast Preparation System (IFPS)which allows National Weather Service (NWS) forecasters to express weather forecasts in gridded, digital form. Since defining these gridded forecast from scratch is far too tedious and time consuming, the Forecast Systems Laboratory (FSL) has been developing a suite of algorithms, called ifplnit, that transforms local or synoptic-scale numerical model output into surface sensible weather element grids. Forecasters may copy these objectively-derived grids into their forecast as a starting point, before using other tools to revise and refine them, thus saving much time and effort.
Recently, FSL has developed a new framework in which these algorithms produce surface-based gridded forecasts employing Python, a very high-level scripting language. A performance-enhancing extension to this language called Numeric Python, allows the algorithms to operate on entire grids at once, sharply reducing the amount of code required to implement each algorithm. This combination not only gives NWS forecasters the additional capability to modify the algorithms locally, but also provides better performance so that gridded forecasts are generated quickly.
This paper describes how the GFESuite system generates these surface forecast grids at an NWS forecast office. In addition, we discuss the algorithms used to generate the approximately 18 forecast weather elements and how local forecsters can modify the algorithms to improve their accuracy.
Session 5, Interactive Forecast Preparation System—Operational Employment
Tuesday, 15 January 2002, 2:00 PM-5:15 PM
Next paper