This quarter, the graphics depicting parcel trajectories from TDL's trajectory model available on AFOS have been converted to "Redbook" graphics and implemented on the AWIPS SBN. The 1998 SCAN field test at the Washington DC/Baltimore Forecast Office was begun; a number of new applications are being tested. LAMP implementation into AWIPS continued and was made more robust in a fully operational setting; preparation for implementation of the QPF portion of LAMP continued. An on-site survey was conducted of the Mississippi Gulf Coast SLOSH basin to map physical features not evident on quadrangle maps, and a SLOSH display program is being prepared that will run either or on AWIPS or on PCs, the latter for use by emergency managers. The first version of IFP-- the combined ICWF and AFPS--was assembled; it will be tested in the Norman WFO. Requests for a new MOS system based on the AVN and ETA models were coordinated with the Office of Meteorology.

OBJECTIVE WEATHER PREDICTION PROJECT (P. Dallavalle)

Short-Range Weather Forecasting Task (P. Dallavalle): A seminar entitled "A Continuous-in-space MOS System for Maximum/Minimum Temperature Forecasts" by Rebecca L. Allen of the Pennsylvania State University was presented in May. This seminar described work done at the Cooperative Institute at Penn State to develop a methodology to produce MOS forecasts of the max/min temperature at any point in the contiguous U.S. regardless of the availability of historical observations. Work continues on developing similar techniques for producing forecasts of dew point, wind speed, clouds, and precipitation probabilities.

Development and testing of the MOS-2000 software system on the Cray and the local HP workstations is continuing. Efforts this quarter were focused on writing software to compute derived model forecast variables (moisture divergence, dew point, wet bulb temperature, stability indices, etc.), to prepare predictand observations (for example, the estimated daytime maximum tempera- ture), and to put observed surface hourly observations and satellite cloud reports into MOS-2000 format. In addition, an extensive number of updates are being made to the TDL dictionary defining the hourly surface observation sites. This dictionary is essential to the MOS-2000 system and will also be provided to the AWIPS Implementation Office for use in AWIPS.

In preparation for the development of the next generation of MOS guidance, the Office of Meteorology (OM) queried the NWS field forecasters about requirements for MOS guidance. Preliminary results indicate that the field needs complete packages of MOS guidance based on the Eta and Aviation (AVN) models for forecast projections out to 72 hours, and a complete package of MOS guidance based on the Medium-Range Forecast (MRF) model for projections out to 7 days. Guidance is required for as many sites as warranted by available observations.

Efforts to make Y2K-compliant all of the TDL software required to produce the centralized guidance continued. The TDL software library used by TDL and NCEP programmers is now Y2K compliant. During the next quarter, the operational codes will be modified as necessary. The software will also be converted to Fortran90 syntax in preparation for transition of the operational codes to NCEP's new Class VIII computer.

Medium-Range Weather Forecasting Task (M. Erickson): We have completed supplementing our new archive of the MRF model with data available from NCEP for the period of January 1992 through March 1997. These data were available every fifth day for projections of 12 to 192 hours after 0000 UTC. The complete archive of MRF data through the current date will be used to develop the new MRF-based MOS guidance.

On June 15, NCEP implemented an updated global spectral model. This new version of the model runs at a T170 horizontal resolution with 42 vertical layers and is used to produce the AVN and MRF runs. Several changes were made to the physics, analysis, and data assimilation used in this model, and we are monitoring the characteristics of the new model for impact on the current MOS as well as on future MOS developments.

National Verification Processing Task (V. Dagostaro): In support of the AFOS-Era Verification (AEV) program, we continued our effort to convert the data processing system to run on UNIX-based platforms. Software to write the AEV data in the standard MOS- 2000 packed format was developed and tested, and data for November 1996 through March 1998 were converted to the new format. A procedure developed last quarter to automatically execute the weekly AEV data collection software has also been successfully tested and is ready for implementation on the HP workstation. To comply with Y2K standards, we've modified the daily AEV data collection software which runs in NCEP's operational job stream. At the same time, software modifications were made to facilitate the daily monitoring of the operational job. Testing of the software is underway. Finally, development of software to quality control the AEV data is continuing.

Other efforts during this quarter were oriented toward the implementation of the current public and aviation weather verification system in the modernized NWS. Assistance was provided in the development of the algorithm required to verify the maximum/minimum temperature. In addition, coordination with the new San Joaquin Valley forecast office was necessary to ensure that the current VERIFY software was installed properly at the site.

Severe Weather Prediction Task (R. Reap): A new suite of MOS thunderstorm and severe local storm probability forecast equations is being developed from Eta model predictors and archived thunderstorm and severe local storm predictands. As a first step, cloud-to-ground lightning reports and severe weather events from the Storm Prediction Center's log have been archived on tape in MOS-2000 format for the period of April 1, 1994, to October 31, 1997. The 4-year sample contains over 90 million lightning flashes and 70,000 severe storm reports. Initial runs were made with the new interpolation/calculation program (U201) from the MOS-2000 system to test the generation of datasets required for the regression analysis. We expect to begin development and testing of the forecast equations as soon as we generate the appropriate relative frequency values.

After coordination with NCEP, OSO, and OM, software to produce NGM-based parcel trajectory forecasts in "redbook" graphical format was implemented. The products containing forecasts of parcel trajectories at the surface, 850, and 700 mb are now available to AWIPS over the Satellite Broadcast Network (SBN).

LOCAL TECHNIQUES DEVELOPMENT PROJECT (R. Reap)

0-3 Hour QPF and Severe Weather Task (D. Kitzmiller): We are continuing real-time tests of the 0-1 hour rainfall forecast and severe local storm nowcast algorithms in AWIPS. Based on early results, the rainfall forecast algorithm was modified to utilize the mean convective storm cell velocity from the Storm Track Algorithm during convective events. This extrapolation velocity yields forecasts that are consistent with the orientation of storm tracks. The rainfall algorithm's internal lag-correlation reflectivity pattern matching method is still used to estimate the echo velocity in stratiform events. A validation study of the rainfall algorithm, using data from the 1996 warm seasons, has been completed and a manuscript on the results is in preparation.

We are currently testing a prototype version of the 0-3 h QPF algorithm, which utilizes radar, satellite, and NGM input, and comparing its output with national mosaics containing Stage II rainfall estimates.

In cooperation with OM, we are examining the WSR-88D archives from recent convective storm events in the southern United States to assess the current siting of the Jackson, Mississippi, radar and to assess the performance of storm algorithms during specific severe weather outbreaks.

Discussions are also underway with OSO on implementation details for the Radar Coded Message (RCM) mosaic and AUTOROB procedures algorithms.

Thunderstorm Identification and Forecasting Task (S. Smith): After delivering the SCAN 1.0 code to PRC for integration into AWIPS 4.1, the SCAN DDT team fixed a number of problems through the Discrepancy Report (DR) mechanism. The SCAN DDT team began designing and prototyping the functionality of the National Severe Storms Laboratory's (NSSL's) Warning Decision Support System for SCAN 2.0.

An auto-verification of the automated thunderstorm detection algorithm was developed for the NHDW system. The program checks real-time METAR reports of thunder against output from the site storm threat algorithm. Large amounts of verification statistics can now be generated in a short period of time for the Sterling and Wakefield, Virginia, radars. The program could be implemented at all AWIPS sites if desired. Work was begun to convert the site thunderstorm threat to a gridded product which will be a part of SCAN 2.0.

The 1998 SCAN Field Test at the Washington D.C./Baltimore forecast office in Sterling, Virginia, began on June 1. New applications being tested this season include NSSL's implementation of the Areal Mean Basin Estimated Rainfall (AMBER), a GOES thunderstorm anvil tracker, and the MIGFA gust front detection algorithm. In addition, the National Center for Atmospheric Research is testing its complete Thunderstorm Autonowcast system. The test is scheduled to conclude on September 30.

Local AWIPS MOS Program Task (J. Ghirardelli): Site-specific equations, thresholds, weight files, and location files were created for 16 additional AWIPS sites, namely: BOI, CLE, ILN, FFC, HGX, IND, JKL, LMK, MLB, PAH, AMA, MFL, MSO, SEW, TFX, and Southern Region Headquarters. These files were delivered to PRC via DAT tape and have been put under the AWIPS configuration management system.

We completed our evaluation on the timing of the cron jobs to determine if LAMP could be run any earlier. As a result of this investigation, specific changes were implemented in AWIPS 4.1 causing LAMP to run at 25 minutes after the hour, which is 20 minutes earlier than previous versions. In addition, all LAMP source codes on the GDP were analyzed for year 2000 compliance. Codes dealing with the BUFR MOS and the NGM grid files are being modified to bring them into compliance. The remaining code appears to be year 2000 compliant. The LAMP pre-processing C code on the NHDW was also tested for compliance and no problems were found other than those regarding the BUFR MOS and NGM grid files which were already known. We also made changes to the LAMP code for the ICWF group in order to facilitate decoding the MRF BUFR message directly in order to move away from decoding the MRF bulletins from AFOS.

We recently cooperated with the Products Development Branch and Synoptic Scale Branch in testing a new BUFR MOS message with two additional probabilistic QPF categories for purposes of Risk Reduction at the Pittsburgh and Charleston WFOs. We continue to respond to user inquiries concerning the LAMP Real-time Products web page. Specifically, we were contacted by Taunton, LaCrosse, and Newport. In response to these inquiries, we have improved the method for informing users of scheduled and known outages. We also implemented a plan for quicker dissemination of the LAMP products to the Internet. Updated forecast bulletins are now available up to twenty minutes earlier. We also improved the "professional look" of the web page and made it more informative concerning product update times.

With respect to LAMP implementation within AWIPS, we discovered that the 03Z and 15Z LAMP runs were failing due to a control file problem. A Discrepancy Report was written and the problem was fixed for Build 4.1. We also worked with the Products Development Branch on writing a Discrepancy Report and implementing a fix for the new METAR coding of Ice Pellets as PL. Informal documentation was also developed for the process of checking that LAMP is running correctly at the WFO sites. We routinely monitor LAMP performance at the WFOs, which recently led to the discovery that old BUFR messages were not purging correctly. A detailed description of how to purge these old files was delivered to the Network Control Facility. We are now working on a Discrepancy Report to correct this problem in the source code.

Heavy Precipitation Forecasting Task (J. Charba): An effort was undertaken to diagnose and correct errors that were discovered by NCEP's Climate Prediction Center (CPC) in hourly precipitation data supplied to them for the period 1948-62. This effort has been completed and CPC is currently using the corrected data. Acquisition of this additional dataset expands TDL's coop hourly precipitation database, which now covers the 49-year period from 1948 to 1996. In future applications, the dataset will be used for updating and expanding other datasets containing climatological analyses of precipitation which are currently used by the LAMP QPF system.

Efforts are continuing to maintain and upgrade the national LAMP QPF system that is run experimentally on TDL's workstations and to develop an analogous system for implementation on AWIPS workstations at WFOs. Recent upgrades to the national system include updated MOS QPF station list files, correction of errors in the decoding of hourly ASOS precipitation reports, upgrading of an obsolete surface station dictionary with TDL's current dictionary, compression of the topography and precipitation climatology databases by converting them into packed form, and improvements to the content and quality of the QPF product presentation in TDL's Home Page. Progress toward development of the local LAMP QPF system has included the creation of various files comprised of MOS QPF stations and LAMP QPF forecast points to be used for each of the ten "local" areas in which the 48 states is divided, a partitioning of the packed topographical and climatological databases into files for each of these local areas, and the preparation of UNIX scripts that perform the various data-type injests, pre-processing, and grid analyses for the local areas.

MARINE ENVIRONMENTAL PREDICTION (W. Shaffer)

Hurricane Storm Surge Forecasting Task (W. Shaffer): We conducted a survey of the Mississippi Gulf Coast SLOSH basin, examining physical features that were not evident from quadrangle maps. One of the significant features in this basin is the railroad that follows the coastline from New Orleans into Mississippi. The railroad embankment is often the highest barrier to the flow of water--we are working with NHC to get elevation measurements for the railroad. We met with all three of the county emergency managers along the Mississippi Gulf Coast, showing the new grid coverage and explaining updates in storm surge forecasting. Each was interested in the animated simulation of Hurricane Camille and variations that we produced. This Category 5 hurricane made landfall over St. Louis Bay and caused high water marks of 25 feet. We ran hypothetical variations of Camille 1/2 degree to the east and to the west.

Two 1-day training workshops were presented in combination with this survey--one was held in New Orleans for the Southeast Louisiana Hurricane Evacuation Task Force and another was held in Lake Charles for the Southwest task force. W. Shaffer also presented seminars at both the Slidell and Lake Charles, Louisiana NWS offices. At the request of (and with Corps funding), similar workshops were held for emergency managers in St. Thomas, St. Croix, and San Juan.

The SLOSH display program is being readied for inclusion into Build 4.2 of AWIPS. One new feature we have included is an animated display of past hurricanes. The software has been demonstrated to several groups, both within and outside NWS, and one beta tester is evaluating the Windows version of the software.

Extratropical Storm Surge Forecasting Task (W. Shaffer): We continue to run our West Coast version of the extratropical storm surge model twice daily, in an experimental mode. Because of the narrow shelf on the West Coast, surges (as we expected) are relatively small, even for the more intense Pacific storms. Large waves are the primary coastal threat along the West Coast.

Coastal Wave Forecasting Task (C.-S. Wu): The parametric ocean wave model is modified for forecasting waves generated by small and large radii of maximum wind. We combined the ocean model with a shallow water refraction-diffraction model. We also ran the breaking wave model for various beach slopes and incoming wave heights ranging from 8 to 12 meters to estimate typical hurricane wave set-up.

We obtained measurements of wave set-up during January's major extratropical storm--a gage that was jointly funded through NOAA, the University of Delaware, and the Corps of Engineers and located on the Corps' pier at Duck, North Carolina. The Corps summarized the preliminary water level data, putting it on an internet page. This inshore gage was calibrated using the NOS tide gage at the end of pier during low wind periods. The water level difference between the two gages is the wave setup. These are the first field measurements of wave-induced setup during extratropical storms along the East Coast.

LOCAL PRODUCTS DEVELOPMENT PROJECT (D. Ruth)

WFO Application Development and Support Task (D. Ruth): Design, Development, and Testing teams are currently focusing their efforts on completing, testing, and documenting hydrometeorological (HM) applications for AWIPS 4.2. These include TAF monitoring, forecast verification, and NOAA Weather Radio formatters for the Hourly Weather Roundup, Daily Climate Report, Service Area Forecast, as well as Watch, Warning, Advisories, and Statements. The useability of these applications as implemented on AWIPS will be tested at the Pittsburgh Weather Forecast Office beginning this summer.

IFP Implementation and Enhancement Task (R. Meiggs): As the AWIPS deployment schedule accelerates, TDL continues to provide initial site set-up for WFOs as a part of the Factory Acceptance Testing (FAT). PRC installs IFPS with this set-up during the FAT and conducts minimal testing of the software. This testing allows TDL to identify and fix problems associated with particular sites before deployment.

The latest IFPS has been installed on the AWIPS 4.1 SwIT platform at PRC. This enables testing of installation scripts and integration with the rest of AWIPS. Discrepancy Reports are being addressed by TDL developers. A Software Integration Working Group (SIWG) has been created to iron out installation issues that include the replacement/upgrading of site customized files, installation at OCONUS sites, and COTS upgrades.

The NWR Service Area Forecast (SAF) continued to be tested at the forecast office in Charleston, West Virginia. As a result, the SAF program has been enhanced to handle forecasts that cover the "entire radio listening area" and to provide specific county lists for each broadcast segment. Other improvements under development include the insertion of Watch/Warning/Advisory headlines and temperature/POP tables within the SAF. A follow-up design review was held in June to describe the latest changes being made to AWIPS Watch/Warning/Advisory software. The AWIPS 4.2 software is being developed to support NWR broadcast (including products from adjacent offices), intersite coordination and monitoring, segmented products (new format), VTEC, and watch decentralization. Short-fused warnings entered via Warngen are captured for further processing within a digital database. Long-fused products will be entered via the ICWF WWA interface.

Significant progress was made in entering and tracking field problem reports, requests for enhancements, and questions. The effort required the creation of a series of database tables, into which most field reports have now been entered. This information will be regularly updated and posted on the TDL web page. In addition, preliminary work has been completed on a plan which describes the process of reporting bugs, enhancements, and/or questions to TDL.

IFP Product Development and Evolution Task (M. Peroutka): The first version of the Interactive Forecast Preparation System (IFPS) was assembled. The IFPS Forecasters Working Group (IFWG) met in Silver Spring from April 14-17 to review the system. Since the meeting, IFPS development has focused on implementing the techniques needed to control the flow of data between gridded and matrix representations, adding areal coverage phrases to the precipitation phrase generation routines, and configuring the various components to support prototype testing at the Norman, Oklahoma WFO.

An enhanced set of capabilities for weather representation were designed for IFPS in addition to the areal coverage phrases mentioned above. Attribute information will be added to weather grids, new values for obstructions to vision will be available, and "wildcard" phrases will allow local WFOs the opportunity to configure and generate site-defined phrases.

Development continued to support the Probabilistic Quantitative Precipitation Forecast (PQPF) Risk Reduction in Pittsburgh, Pennsylvania and Charleston, West Virginia. A new program was developed which creates guidance grids of PQPF elements for initialization. TDL developers traveled to the University of Virginia to discuss enhancements to the PQPF software. This version will be delivered to TDL in early July and installed at the two WFOs.

IFP Interpretation and Editing Task (D. Ruth): The interface that enables forecasters to direct digital and worded public zone forecasts to NWR towers was enhanced according to suggestions from the forecast office in Charleston.

The capability to overlay original forecast guidance fields and model parameters as contours has been added to model interpretation with IFPS slider bars. A new graphical user interface is being developed to facilitate local site maintenance of ICWF control constants. These constants are now incorporated in ICWF database tables.

Field Applications Assistance Task (R. Beasley): Support for AFOS this quarter was limited to updating applications for Y2K compliance and making modifications which support the new METAR abbreviation for ice pellets.