1. IFPS Science Steering Team (ISST) -- a first status report to the S&T Committee -- S&T Committee VTC 28 May 2003

2. ISST Vision The IFPS Science Steering Team facilitates an efficient and effective process that allows the NWS to identify, collect, prioritize, and propose focused solutions and recommended courses of action to IFPS science issues. Recognized as a primary conduit between IFP operations and NWS Headquarters, this team of field experts serves to better ensure scientific and technological integrity in the digital forecast process.

3. ISST Membership Brad Colman, team leader, WFO Seattle, WA Peter Manousos, backup team leader, NCEP/HPC Dan Baumgardt, WFO La Crosse, WI Mark Jackson, WFO Salt Lake City, UT Larry Lee, WFO Greenville-Spartanburg, SC Andy Patrick, WFO Corpus Christi, TX Eric Stevens, WFO Fairbanks, AK Bill Ward, WFO Guam Ken Waters, Pacific Region Headquarters Kevin Schrab, facilitator, NWS Headquarters Office of Science and Technology

4. Brief review of our short history… SOO whitepaper Interest and support from Regional and National Headquarters Cliff Mass’ paper S&T committee forms ISST Western Region SOO/DOH workshop

5. Today’s briefing Immediate opportunities –MDL’s expanded MOS development –Verification efforts Early recommendations –Immediate and optimal use of SBN –AWIPS grid change procedure –Enhancements to the GFE Overarching challenges –Analysis of record –Incorporating probabilistic information

6. Objectives Approval of recommended courses of action Discuss and validate potential directions on overarching challenges Recognize science needs and urgency Reprioritize and reallocate resources, if necessary

7. S&T Committee Science Issues/Questions 1.What is the best mix of probabilistic and deterministic forecast information out to Day 7? 2.Are there scientific techniques to maximize forecast “value added”? Downscaling Statistical guidance Resolution of NDFD and model guidance 3.How do we assess forecast accuracy and skill? 4.What are the science issues of adding a third dimension to the NDFD? 5.What analysis and short-term forecasting tools are needed? 6.What are the best methods to ensure consistent surface elements?

8. Immediate Opportunities MDL currently expanding GFS MOS data (Science Issue 2) –By Fall ’03 GFS MOS sites will include METAR, Co-Op, Buoy, and CMAN sites (MaxT/MinT to 10 days) –Partially meets immediate need for enhanced statistical guidance –Could allow early proof-of-concept evaluation of gridded MOS –No formal plan in place for the distribution and use of these new data –We are talking with MDL to assess possible options Verification efforts (Science Issue 3) –MDL and WR starting prototype effort to verify NDFD MDL has grid verification software for NDFD Western Region has 2.5 km ADAS analysis –Existing Verification Working Group Could provide basis for an overseeing, permanent team or working group with broad representation to direct the IFPS verification system ISST wants to endorse this effort and explore ways to enhance its role and impact

9. Early ISST recommendations Selected from WR SOO/DOH meeting brainstorming sessions earlier this month Each topic had strong endorsement at the meeting These are the high impact – likely attainable in a short time with limited resources – steps They are not ranked or presented in priority

10. Immediate and Optimal Use of SBN Identified Need: –Improved grids for IFPS Smart Init process. Proposed Solution: –Use SBN ‘lulls’ to transmit additional high resolution NCEP model data to the NWS field offices. –Incorporates the idea of “data latency.” Recommended Course of Action: –Achieve solution by end CY03 (within existing resources) ‘Lulls’ Science Issues: 2 and 6

11. AWIPS Grid Change Procedure Identified Need: A change management process that plans for, programs, and delivers model improvements and best-available model grids to the field on the day they go operational. –Currently, the ability to add new model datasets are contingent to the AWIPS Build Process (2 to 6 months). –This process takes too long and needs to be streamlined. –A better method for the delivery of AWIPS grid updates is needed. Proposed Solution: This involves decisions and actions regarding generation, transmission, and WFO/RFC receipt (including processing, storage, and display). Recommended Course of Action: S&T committee needs to validate this as a need/requirement and then pass on to appropriate groups for implementation. Science Issues: 2, 4, and 6

12. Enhancements to the GFE Daily Forecast Critique Ingest and use of real-time (Nowcast) data in GFE “Smart Init” improvements and modifications (FSL feedback is that development work on these enhancements can begin as early as late July)

13. Daily Forecast Critique (DFC) Identified Need: Local verification and real-time gridded forecaster feedback within the GFE –Currently: Forecasters can only compare forecasts to observation points and not grids. Proposed Solution: –Daily Forecast Critique (DFC) represents the best verification foundation to build on at this time –Bring DFC within the GFE environment –Enhance to include gridded verification using gridded analysis Ability to incorporate verification grid into forecast grids –Flexible user framework to derive varying statistical feedback –Develop, test, and implement above solutions in RPP-like process Recommended Course of Action: –Advocate this as a high priority for FSL Science Issues: 2 and 3

14. GFE Ingest and Use of Real-time Data Identified Need: To provide gridded real-time information in GFE to update the short-term and incorporate into short-term forecast process (0- to12-h) –Not for pure display purposes Proposed Solution: –Ingest satellite and radar data into the GFE –Satellite Data Convert to GFE-compatible gridded form Modify GFE to ingest sub-hourly data Need for algorithm development –Radar Data Development work needed (possible ROC involvement) Recommended Course of Action: –Advocate this as a high priority for FSL Science Issues: 2 and 5

15. Enhancements to GFE Smart Initialization Identified Need: An improved Smart Init process that takes advantage of best-available grids to generate best possible base and derived forecast grids. Proposed Solution: –Provide full-resolution model data for initialization –Evaluate baseline and field-modified Smart Inits –GFE modifications necessary for additional GFS data IFPS 14 Patch, post OB2.1 (Fall ‘03) –Support formation of Smart Init Focus Team charged with incorporating field modifications to Smart Inits into baseline Close collaboration with COMET/NWP Team effort charged to assess Smart Tools and Smart Init Must include field (WFO) expert representation Test Smart Inits on native resolution data at NCEP Recommended Course of Action: –Advocate this as high priority for FSL Science Issues: 2 and 6

16. Overarching Challenges (for discussion and validation) An analysis of record –A real-time gridded analysis of the sensible weather elements that are forecast within IFPS: temperature, precipitation, wind, etc. Incorporation of probabilistic information –Probabilistic information is needed as guidance within IFPS. This is necessary to allow the NWS to identify opportunities that will allow us to deliver probabilistic forecasts and value-added information to a wide range of customers.

17. Analysis of Record Motivation: –Real-time seamless verification –Provides forecasters useful feedback –Allows forecasters to assess the initialization and performance of NWP models –Serves as input to the GFE for use in short- term forecasts –Contributes to the ongoing development of a gridded climatology Science Issues: 2, 3, 5, and 6

18. Considerations: –Be at the same resolution (both spatial and temporal) as the forecast grids –Incorporate data from all sources: RAWS, COOP, satellite, radar –Be as independent from the NWP models as possible Potential directions: –A collaborative effort will likely be needed between the NWS, ERL, and universities. –Opportunities for outsourcing should be explored –External peer-review process will be beneficial –A long-term effort is required. Work should begin now, as it will likely be some time before results are available.

19. Motivation: –NRC Workshop summary 2003 Exemplifies the need to communicate uncertainty to the end user –NWS Vision 2005 Calls for NWS to provide weather, water, and climate forecasts in probabilistic terms by 2005 –State of meteorological science demands such a direction High level objectives: –Intelligently convey uncertainty in a practical manner to forecasters and end users –Facilitate the concise visualization and assimilation of ensemble output into the forecast process Incorporate Probabilistic Information Science Issues: 1, 2, and 3

20. Considerations: –End users would benefit from the ability to produce a wide variety of probabilistic output formats –Forecasters need access to proper ensemble output –Required training Forecaster training on the proper use of ensemble output End user training to properly utilize available output formats –Need to tap expertise in and outside the NWS Potential Directions: –Deliver ensemble gridded fields to WFOs –Develop forecaster training –Explore potential use of probabilistic density functions (PDFs) derived from ensemble output –Work with NWS partners to identify high-priority probabilistic forecasts

21. Summary of Decisions Approval of recommended courses of action: –MDL’s expanded MOS development –Verification efforts –Immediate and optimal use of SBN lulls –AWIPS grid change procedure –Enhancements to the GFE Validate priority and potential directions on overarching challenges: –Analysis of record –Incorporating probabilistic information

22. Final Comments First baby steps We look forward to an active dialogue with the S&T Committee Recognize the need and urgency

23. Background information follows

24. Core SOO Whitepaper Proposals Develop a National real-time, gridded verification system Provide full-resolution NCEP model grids Objectively produce bias-corrected model grids for WFO use Implement methods to objectively downscale forecast grids Incorporate climatology grids into the GFE process Deliver short and medium-range ensemble grids Modify the GFE software to ingest real-time data Optimize ways to tap forecaster expertise

25. Details Regarding Recommended Course of Action for the Immediate and Optimal Use of the SBN: –Analyze transmission timing scenarios (July; SEC - Tuell). –Identify and prioritize new NCEP data sets for review and concurrence. (June; ISST, SSDs and SOOs) –Complete necessary notifications, modifications, and begin new SBN transmissions (August; OCCWS and OST) –Incorporate required software update in OB 2.1 (August; SEC) –Provide a patch to IFPS 14 for immediate use of new data sets (August; FSL/SEC). –Incorporate modifications in IFPS 15 as baseline IFPS SmartInit (late ‘03). Immediate and Optimal Use of SBN

26. Highest Priority NCEP Data for SBN: Target: IOC (9/03) Transmitted on SBN with zero latency –GFS additional levels at 80km Sfc, BL (0-30,30-60,60-90,90-120,120-150,150- 180) 1000-500 x 25 mb, 500-100 x 50 mb (Z,T u, v, RH) thru 168 hrs –Eta 12 surface and BL thru 84 hrs Transmit on SBN with minimal latency –Eta 80km/40km data 60-84 hours at 3 hourly increments. –Eta 12 pressure levels thru 84 hrs. 1000-500 x 25 mb, 500-100 x 50 mb MREF 84-240 hrs (same fields as currently available 0-84) (use WAN?). –GFS BUFR sounding data Immediate and Optimal Use of SBN

27. Detailed Benefits Associated with Streamlined AWIPS Grid Change Procedure: - Allows WFOs to acquire updates to AWIPS grid tables on monthly basis. - WFOs can have access to higher resolution model more quickly. - A delivery mechanism can be developed that is easily manageable. - WFOs can acquire this information and update AWIPS workstations at their own convenience. - Grid updates do not greatly perturb the AWIPS baseline greatly nor affect the AWIPS Build cycle (Bullock, FSL). AWIPS Grid Change Procedure