The modernization of NWS operations involved the development of a new radar system, an automated surface observing system, a new series of geostationary satellites and products, and a communications and forecaster workstation system, the Advanced Weather Interactive Processing System (AWIPS). FSL has been participating for years in risk reduction activities to help the NWS meet its goals in the development and deployment of AWIPS.

New datasets were added in both of these builds. Aircraft reports can now be decoded and displayed; these messages report turbulence, icing, cloud layers, winds, temperature, and other parameters important to air navigation. In some cases, detailed temperature and humidity profiles are obtained from aircraft that are arriving or departing major airports. With FSL's implementation of a decoder for synoptic observations, many more reports of surface weather conditions are available, particularly from other countries. Atmospheric soundings retrieved by geostationary and polar orbiting satellites have been added as well. These provide much greater horizontal and temporal coverage than is obtained from radiosonde balloons. Numerical weather prediction grids for the Great Lakes, Atlantic, and Gulf Coast have been added in support of marine forecasting and navigation.

An important addition in Build 5.2.2 is the introduction of ensemble numerical weather prediction model grids. These data provide an objective method of dealing with the uncertainty inherent in weather prediction by generating multiple solutions based on different initial conditions. Algorithms to calculate the mean and standard deviations of these fields have been added to the volume browser menu, which is used to access gridded model data. To accommodate the volume of data anticipated with ensembles, a compression algorithm has been included for storing and retrieving these data.

New radar capabilities include detailed inventories, high-resolution storm relative velocity, and the rotational velocity (Vr) shear function that works in four-panel mode. Detailed inventories allow the user to differentiate between multiple request products such as cross sections or storm relative motions. The system will automatically load a loop of the particular cross section selected by the user. Now that high-resolution velocity data are available, a capability has been added to the AWIPS workstation to calculate storm relative velocity using user input for storm motion. This is an important tool for diagnosing severe weather. The Vr shear function is another tool used to diagnose severe convective storms. It can now function on four different radar tilts simultaneously so the user can measure how shear in a storm changes with height.

The volume browser has been enhanced with the addition of two new methods for displaying data. To view some high resolution vertical data, a new viewing option that plots a variable versus height was added. Users can now plot any variable (such as wind speed) versus height. The user can also zoom in on one axis to display the additional detail for high resolution data sets. Time-series plots were enhanced to allow multiple plots stacked vertically. This is a widely used method of displaying parameters in a display arrangement known as a meteogram. Typically this is done with METAR (meteorological aviation report) observations, but in AWIPS point forecast data from the various models can also be displayed in this manner.

During 2002, progress was made on the transition of AWIPS to the Linux operating system. The main task was to port certain decoders to Linux in an attempt to offload the HP data servers. The two decoders chosen represent the greatest volume of data sent over the Satellite Broadcast Network (SBN), the grib decoder, and the satellite decoder. These decoders will run on a new preprocessor Linux system where cpu and storage capabilities greatly exceed the existing HP servers. Another accomplishment involved porting of the Product Maker to the Linux workstation. This is a general purpose utility that allows the user to define computations to be performed on gridded or image data with the results displayed on AWIPS.

Collaboration with the Korean Meteorological Agency (KMA) continued. KMA has modified an AWIPS workstation to use the data generated by their ingest systems. These data include radar, surface and upper air observations, satellite imagery, and numerical weather prediction grids. The menu interface to use Korean characters was also modified. The Korean forecasters visiting FSL received two training sessions on how to use AWIPS in forecasting various weather phenomena.

During 2003, continued collaboration with the Korea Meteorological Administration and the Taiwan Central Weather Bureau will focus on nowcasting tools, particularly those that use radar data.

Enhancing the Graphical Forecast Editor – The GFE provides tools for the forecaster to view and edit grid fields that capture the essential information needed to generate a variety of forecast products. Productive interactions between field forecasters and developers prompted numerous enhancements of the GFE graphical user interface, GFE Inter-Site Coordination (ISC) capabilities for coordinating forecasts between NWS offices, and scripting and other capabilities. Smart tools were written to aid the forecaster in comparing differences between their forecasts and their adjacent site's forecasts. The ISC facility (Figure 72) has been enhanced so that the forecaster can now use tools that take into account data from neighboring forecast offices. These tools enhance the forecaster's ability to maintain spatial consistency between adjacent offices. An enhanced Temporal Editor now displays the range of values, giving forecasters more information about how the values vary over any region they specify. The new interactive query interface provides a much more flexible way to define areas of interest based on forecast data values. This will help forecasters create forecasts that are better coordinated among physically related weather elements. Three versions of the GFE were delivered to the NWS for inclusion in the new AWIPS Rapid Alpha Process (RAP), IFPS10, 11, and 12. Seven major and five minor revisions of the Rapid Prototype Process (RPP) software were provided to RPP field sites for focused testing on particular areas of the GFESuite, such as text product formatting.

Developing Field Customizable GFE Text Products – During 2002, the NWS asked the branch to develop a set of locally customizable text formatters of operational quality. The branch created the initial infrastructure and "first-look" versions of the 10 products required for the IFPS Initial Operating Capabilities. The products range from tabular summaries and coded tables to narratives, and cover the public, marine, and fire weather services in the NWS. Using the RPP software approach, the branch worked with 15 field offices and provided 7 software releases to these offices for the express purpose of arriving at an initial text formatting infrastructure, which was completed by spring 2003.

Developing the Daily Forecast Critique – Last summer, the EFT branch released the initial version of the Daily Forecast Critique (DFC, Figure 73), a data visualization package that, for the first time, allows forecasters to compare their digital forecasts to observations measured at the surface. DFC includes a data archiving facility that permits each site to archive surface observations as far back as several months, greatly extending the current AWIPS limitation of 36 hours. The system archives previous forecasts as well, both model-based and human generated, providing a rich data set with which forecasters can determine where their forecasts deviate from observed conditions. Forecasters view these point-based datasets with the DFC Viewer, which plots these data in time-series format. The flexible way in which the data can be plotted allow forecasters to investigate many attributes of forecasts such as direct comparison between various models, determining trends over subsequent model runs (dProg/dt), and examining error as a function of forecast lead time. Future work includes providing the capability to compare grids to grids rather than points, routinely generating statistics to determine model and official forecast bias, and tools to apply these statistics to the new forecast to remove these biases, thus improving the forecast.

Providing Forecaster Training – The branch held three hands-on workshops and reached virtually every NWS office. Approximately 140 NWS forecasters and IFPS focal points attended the workshops. The intensive 3-day workshops covered advanced concepts of GFESuite (such as writing smart tools), creating smart initialization algorithms for model processing, and writing text formatters. The format of these workshop sessions ranged from lecture to interactive hands-on sessions, where forecasters performed exercises to familiarize themselves with technical aspects of GFESuite. Open discussion sessions provided a chance for forecasters to share ideas, suggest improvements, and discuss particular forecast problems. Forecaster feedback about these workshops was very favorable.

Providing IFPS Field Support – The NWS maintains three IFPS-related listservers for the field to discuss issues, find solutions, and interact with the developers. The branch makes good use of the listservers to the benefit of the forecasters and focal points by providing quick turnaround to questions, tips on better use of the software, and both field troubleshooting as well as focal point consultations. During 2002, every forecast office spun up GFE operations.