LAPS WORKSHOP SUMMARY Oct , ESRL, Boulder, CO

LAPS SUMMARY User base users –NWS – WFOs via AWIPS2 –Other US agencies – DHS, FAA, DoD, NASA –Private sector – Renewable Energy, weather forecasting –Internationally – Weather forecast agencies in Europe and Asia Primary application areas –Situational awareness Convective initiation, aviation, winter weather –Warn-On-Forecasting Severe weather, fire weather, hydrometeorology Unique combination of features –Consistency with observations –Fine scales –Very rapid update –High portability –Ease of use

WORKSHOP RECOMMENDATIONS Continue development and support of LAPS –To serve enterprise needs Maintain & further enhance unique features of LAPS –Fine scales, very rapid updates, high portability, ease of use

WORKSHOP LOGISTICS Organized by Forecast Applications Branch (FAB) of GSD Held Oct 2010 at GSD/ESRL/OAR/NOAA in Boulder 35 participants, including 6 via telecon –OAR – GSD, NSSL –NWS – Southern & Central Regions, Training Center –NESDIS –Academia – NCAR, CIRA –International – Finland, Spain, China –Private Sector – Vaisala, WDT, Precision Winds, Radiometrics 30 presentations, including 5 posters 2 sessions of Working Group discussions on recommendations

AREAS OF DISCUSSION Role of LAPS in NOAA & weather enterprise User requests Scientific opportunities / needs Data Software Collaboration

ROLE OF LAPS IN NOAA & ENTERPRISE Purpose – NWP system for –Situational awareness Nowcasting Verification –Warn-On-Forecasting For severe weather, convective initiation, aviation, fire weather, hydrometeorology, etc Critical qualities –Consistency with observations –Fine scales to resolve severe weather phenomena –Very frequent update –Portability –Ease of use Users – 150+ users –NWS WFOs – AWIPS2 Critical for WFO control of analysis (incl QC) & work flow –Other US agencies, private sector – 10+ groups –International applications – 6+ countries / weather services

USER REQUESTS Quality of analyses & forecasts –Number of problems highlighted Noise in analysis, spin-up in forecast, need for better QC, etc Must set priorities –Some solutions offered by partners Portability –AWIPS2 & other platforms Efficiency –Must run in timely manner Ease of use –Must remain easy to set up –User interface must remain simple

SCIENCE OPPORTUNITIES / NEEDS 3/4Dvar development – Space-Time Multiscale Analysis System (STMAS) –Better use of observational data –Cross-variable background error covariances –Physical balances consistent with observations Errors in observations –QC improvements –Observational and representativeness errors Choice of background field –Forecast from previous LAPS analysis (cycled DA) Desired but problems with errors on fine scales due to –Chaotic error growth and boundary effects Role of dynamical downscaling Initialization of NWP forecasts –Analysis must be consistent with NWP models used Consistency with eg microphysical and land surface model compenents Estimation of uncertainty in analysis and forecasts –Ensemble approach –Coupled with 3/4Dvar Ensemble-based covariances

DATA NEEDS “Static” data - Use standard procedures for access – –Fine resolution terrain, land surface, etc –E.g., Topograbber Observational data – Include new types Dual polarization etc radar, lidar, satellite, etc Background forecast –Test new approaches Cycled forecasts from LAPS analysis ECMWF 16 km global forecast

SOFTWARE REQUIREMENTS LAPS / WRF Portal –Develop user friendly GUI for Set-up, data access, QC, running & monitoring LAPS & ensuing WRF forecast –Many elements exist WRF Portal, LAPS domain wizard, NNEW, etc –Great potential benefits – Incident meteorology Assimilate fire weather related observations in field Efficiency –Emphasis on parallelization Conform with coding etc standards –How to interface with GIS Documentation Protocol for external contributions to LAPS –Korea, Finland, Spain, Taiwan, private sector, etc

LAPS & THE COMMUNITY Recommend DTC include LAPS in analysis and forecast evaluation / verification work Recommend HWT test use of LAPS for –Situational awareness / nowcasting –Warn-On-Forecasting Training –NWS developing training package for Set-up, running of LAPS –Need mandate / funding for package on use of LAPS Products, examples, good practices, etc Consistency of LAPS with other NWP systems –Use/integrate community supported tools/formats (bufr, geogen, GIS, etc) –Transfer DA methods tested successfully on fine scales into NCEP operations

FOLLOW-UP Do we need regular LAPS User Workshops? –Biannual? What to change, include, etc? Publicize outcome of workshop & LAPS? –A BAMS meeting summary? OR –Several papers (special issue) for a journal (BAMS)? Workshop summary LAPS overview NWS survey Description how LAPS used at WFOs Contributed papers by other users?

LOCAL ANALYSIS & PREDICTION SYSTEM (LAPS): A LOOK INTO THE FUTURE Zoltan Toth Global Systems Division, ESRL/OAR/NOAA Acknowledgements: Steve Koch, John McGinley, members of Forecast Applications Branch LAPS Workshop, Oct ,Boulder, CO

SUMMARY OF LAPS ACCOMPLISHMENTS First comprehensive cloud initialization scheme –Achieved in early 1990s Very efficient, very rapid update, fine scale –5-15 mins cycles, 1 km resolution Highly portable analysis system –Pre-WRF achievement –Runs on variety of platforms LAPS or its adaptations operational at –15+ agencies Including 6 weather services worldwide Plethora of scientific achievements –Large number of publications worldwide

Cloud / Reflectivity / Precip Type (1km analysis) DIA Obstructions to visibility along approach paths

2hr HOT FcstAnalysis2hr NO-HOT Fcst 850 mb Analyzed and Simulated Reflectivity 16 June OAR/ESRL/GSD/Forecast Applications Branch Mature Squall Line Animation Initialized with LAPSInitialized with NAM

LAPS TEAM Data assimilation –Steve Albers, Yuanfu Xie Satellite and other observations –Dan Birkenheuer, Seth Gutman, Kirk Holub, Tomoko Koyama Physical processes –Paul Schultz Ensemble forecasting –Isidora Jankov Evaluation and verification –Ed Tollerud, Ed Szoke Software engineering –Linda Wharton, Paul McCaslin Technical support –Adrienne Rose, Joanne Krumel, Stanislav Stoichev Former colleagues –John McGinley, Huiling Yuan, Brad Beechler, Brent Shaw, etc Long list of collaborators, visitors, etc

SCIENTIFIC SOUNDNESS Solid foundation for traditional LAPS analysis Weak overall structure –Not 3- or 4-Dvar arrangement Room for significant improvement –New 3Dvar scheme is being developed Traditional “hot-start” rebuilt using variational principles –Space-Time Multi-scale Analysis System (STMAS) More information extracted from data More balanced initial fields Already used by US & international agencies!

USE IN OPERATIONS Quality –The good and the bad – need honest feedback Ease of use –AWIPS and elsewhere Choice of domain and execution – Information Technology (IT) issue –Local / Regional NWS – Local Weather Forecast Offices (WFOs) and Eastern Region –Central Korean Meteorological Agency Central Weather Bureau (Taiwan) Finnish Meteorological Institute Link with other tools –MADIS, NNEW, WRF, etc

LAPS IN NOAA & WEATHER ENTERPRISE Same methodology pioneered for fine scale local applications –Can be tested & transitioned to CONUS, national, or global scales –Scientifically, choice of execution (IT issue) does not matter Potential use in NCEP operations (Rapid Refresh, etc) –Ideally, same/similar algorithms should be considered for various scales –Methods tested successfully on fine scale may find their way into next generation national / global systems What is needed for successful technology transfer to NCEP Central Operations? –Local execution may require specific information technology solutions –Fine scale solutions may warrant deviations from GSI approach on specific issues –Build/retain consistency with system operational at NCEP (GSI) in all other areas –Very challenging task

NEW DIRECTIONS Data assimilation –3- and 4-Dvar Collaboration with WRF and DTC DA activities –Global LAPS Encouraging results re tropical and warm season convective systems Ensemble forecasting –Major expansion into probabilistic forecasting –Coupled Data assimilation / ensemble forecasting system Ensemble-based covariances for 3-4-Dvar Finer resolution applications –Convective initiation Warn-On-Forecast –Fire weather & Renewable Energy 100s or 10s of meters resolution Field deployment –Support incident meteorology –Real time assimilation of field observations Statistical post-processing –To remove systematic errors from ensemble

WORKSHOP EXPECTATIONS Provide advice and feedback to facilitate Adjustments / corrections in approach Making system more useful for customers New initiatives to better serve community Thank you all for coming!

BACKGROUND