SR SSD 2000-08

Technical Attachment

Interactive Forecast Preparation System
Rapid Prototype Project
GFESuite Training Summary
November 15 - 18, 1999

The Rapid Prototype Project (RPP) is a new activity for the Interactive Forecast Preparation System (IFPS) being conducted at one site per region and at NCEP. There are three primary areas of focus for the non-operational evaluation of the IFPS software components at these sites: making recommendations for software modifications to the model interpretation and grid editing tools; locally prototyping modernized products; and initiating the forecast process change which accompanies the use of IFPS. This evaluation is being done with IFPS software components on a personal computer (PC) running the Linux operating system, which is installed at RPP sites on the Advanced Weather Interactive Processing System (AWIPS) Local Area Network (LAN).

IFPS software is being developed jointly by the Techniques Development Laboratory (TDL), and the Forecast Systems Laboratory (FSL) for official product generation at NWS offices. TDL has overall responsibility for the development and integration of IFPS, including techniques for MOS initialization, interactive model interpretation, matrix editing, watch/warning/advisory preparation, and the formatting of text products and voice-ready broadcast scripts from a digital database. FSL's development for IFPS is focused on the capabilities required for initializing and editing of gridded data sets.

The first training session for the RPP site managers and focal points was held from November 15 - 18, 1999 at the FSL in Boulder. The goal of this workshop was to provide the participants information on the RPP and training on the capabilities and use of the Graphical Forecast Editor Suite (GFESuite), the first IFPS software component to be evaluated in the RPP. GFESuite provides a graphical forecast editor which includes routines for model initialization of sensible weather elements, and the capability to generate graphical products.

Seven official RPP sites have been established: Charleston, West Virginia; Tulsa, Oklahoma; Boulder, Colorado; Boise, Idaho; Honolulu, Hawaii, the Alaska Region Headquarters, and the Hydrometeorological Prediction Center of NCEP. In addition, there are presently six non-official sites participating to a limited extent in the RPP: Salt Lake City, Utah; Portland, Oregon; Tucson, Arizona; Missoula, Montana; Grand Rapids, Michigan, and Pleasant Hill, Missouri. Managers from the official sites, and the focal points from both the official and non-official RPP sites, the TDL developer for the model interpretation tools, and two instructors from the NWS Training Center participated in this training.

Three consecutive training sessions were conducted, beginning with a session for the RPP site managers. These managers were provided an overview of the project goals, the roles, responsibilities, and deliverables for all participants, and the constraints, resources and tools which will be available to the sites for this project. Ward Seguin, chief of the systems engineering office of the AWIPS Program Office (APO), opened the meeting with an overview of the vision of the RPP. Mike Tomlinson, of the Office of Meteorology (OM) discussed the policy issues relating to the prototyping of modernized products. A Rapid Prototype Coordination Team (RPCT) comprised of the official RPP site managers, Ward Seguin (APO), Dave Ruth (TDL), Mark Mathewson (FSL), and Jamie Kousky (OM), has been established to manage these project activities. The RPCT selected John Jannuzzi, the meteorologist in charge at Boise, Idaho to be the team leader. The following training session was held for both the managers and focal points jointly, where an overview of the RPP plans and the GFESuite capabilities was presented by the FSL staff and John Jannuzzi.

A two and a half day session was devoted to hands on training for the focal points, at which each participant was able to utilize the PC which was ordered specifically for use at their office. This allowed each participant to train on a PC for which the software was configured to display the grids for each participants' respective forecast area of responsibility.

This training provided the focal points instructions for use of the spatial editor, temporal editor, edit tools, writing smart tools, consistency checks, and graphical product generation. Smart Tools allow the user the capability of rapidly editing multiple gridded fields, for generation of tailored gridded fields from which graphical products such as forecast snow amount may be generated. Smart Tools are present in the GFESuite, and the users are provided the capability to locally write and apply additional smart tools using the python scripting language. Tracy Hansen of FSL provided lessons on python and writing smart tools using examples for a convective weather forecast, and a winter storm quantitative snow forecast. GFESuite can be utilized for generation of data in several formats such as text tables, ASCII grids, GRIB, and GIF images for display on the Internet.

An overview of the initialization algorithms used to obtain gridded fields of sensible weather elements was presented by Stuart Wier of FSL. These routines are run to provide weather element values at IFPS grids for any map projection for data which is available as a part of the GFESuite or acquired from the Satellite Broadcast Network (SBN). The highest possible grid resolution for the GFESuite on the RPP PCS is presently set to 10 km.

To assist with the evaluation of the IFPS software components, a RPP listserver and a database for bug reports and software modification requests have been set up. The RPP listserver is primarily for focal point questions and exchange of information. The database has an interactive website access which provides storage of and access to the software requests and software bug reports submitted by the users and developers.

At the focal point training session, Andy Edman of WRH queried the focal points to determine which aspects of GFESuite appear to be least user friendly, and from these discussions, a list of items for future development was proposed. This list was compiled by FSL, and these items have been added to the software modification request database where a level of effort will be assigned by the developers.

Each month, the RPCT will participate in a teleconference at which national prioritization will be made for these items which have been submitted to the software modification request database for future development by FSL and TDL.

The next RPP training session planned for the IFPS software components will be for the model interpretation tools in IFPS, and is tentatively scheduled for February or March of 2000 in Silver Spring, Maryland.

For additional information on the GFESuite training provided by FSL, see:

For additional information on RPP and the GFESuite, see:

For additional information on IFPS, see:

Interactive Forecast Preparation System
Rapid Prototype Project
GFESuite Training Participants
November 15 - 18, 1999

Ward Seguin, APO
Mike Tomlinson, OM
Jamie Kousky, OM
John Jannuzzi, BOI
Steve Amburn, TUL
Steve Nelson, TUL
Dan Luna, RLX
Tom Mazza, RLX
Bob Glancy, BOU
Todd Dankers, BOU
John Dragomir, ARH
Paul Jendrowski, HNL
Ed Danaher, HPC
Steve Listemma, HPC
Jeff Manion, EAX
Gerry Byrne, NWSTC
Bob Hamilton, NWSTC
Tim Boyer, TDL
Dave Elson, PQR
Chris Gibson, SLC
Mike Johnson, MSO
Jeff Davis, TWC
Andy Edman, WRH
Carl Bullock, FSL
Woody Roberts, FSL
Mark Mathewson, FSL
Tom LeFebvre, FSL
Stuart Wier, FSL
Tracy Hansen, FSL