1 / 34

NWS Hydrologic Forecasting AHPS Program February 21, 2013

National Weather Service. NWS Hydrologic Forecasting AHPS Program February 21, 2013. Ernie Wells Hydrologic Services Division NWS Office of Climate, Water and Weather Services. 1. Outline. AHPS Program Focus on Forecast Uncertainty Hydrologic Ensemble Forecasting Challenges.

lowri
Télécharger la présentation

NWS Hydrologic Forecasting AHPS Program February 21, 2013

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. National Weather Service NWS Hydrologic Forecasting AHPS Program February 21, 2013 Ernie Wells Hydrologic Services Division NWS Office of Climate, Water and Weather Services 1

  2. Outline • AHPS Program • Focus on Forecast Uncertainty • Hydrologic Ensemble Forecasting • Challenges

  3. Advanced Hydrologic Prediction Service (AHPS) • Provide enhanced water availability and flood warning information by leveraging NOAA’s infrastructure and expertise • Modernize services through infusion of new science and technology • - Flash-flood to seasonal freshwater forecasts • - Quantification of forecast certainty • - More accurate and timely forecasts and warnings • - Partnered flood-forecast area mapping • - Visually-oriented products • Provide consistent access to standardized graphics via web interface 3

  4. Accessing AHPS Information “click on” the water tab for current river conditions http://weather.gov/ 4

  5. Accessing AHPS Information “click on” the forecast location to access local hydrograph http://water.weather.gov/ 5

  6. Accessing AHPS Information “click on” tabs for probabilistic forecasts For over 2500 locations, NWS provide probabilistic river forecasts 6

  7. Uncertainty estimates needed across all time scales Years Seasons Months Weeks Days Hours Minutes Forecast Uncertainty Forecast Lead Time Benefits Protection of Life & Property Recreation State/Local Planning Hydropower Ecosystem Environment Flood Mitigation & Navigation Reservoir Control Agriculture Health Commerce

  8. Need for Uncertainty Estimates Confirmed • Consistent feedback from customers and research community indicated the need for this uncertainty information • 2006 NRC report • 2008 CFI survey • Aptima study • Multiple Internal NWS Service Assessments re-affirmed this need: • Red River Floods in 1997 and 2009 • Central U.S. Floods in 2008 • Nashville Flooding in 2010 • Missouri-Souris Flooding in 2011 • Recent Request for long-range low flow forecasts for Middle Mississippi

  9. Current Ensemble Capabilities (Long Range) • RFCs use the Ensemble Streamflow Prediction (ESP) component to produce long-term probabilistic forecasts for water supply applications and long range outlooks. • Limitations of existing operational approach • Addresses only the uncertainty in future atmospheric conditions using historical observations of temperature and precipitation • Produces primarily longer-term probabilistic forecasts

  10. Current (Seasonal) Ensemble Streamflow Prediction vs. HEFS

  11. Hydrologic Modeling Satellite Data Precipitation Estimates River Gage Data Reservoir Releases Diversion Radar Data Precipitation Forecasts Hydrologic Forecasting Inputs/Outputs Snow Soil Moisture States Deterministic / Probabilistic River Forecasts Temperature Forecasts

  12. Uncertainty in hydrologic forecast = + “Hydrologic Uncertainty” “Input Uncertainty” model initial conditions, model parameters, model structure, anthropogenic impacts (regulation, diversions, etc.) precipitation, temperature, potential evaporation, etc.

  13. HEFS System Architecture Verification products  Meteorological Ensemble forecast processor Verification system (EVS) Weather/climate forecasts Forcing input ensembles Post- processed “Raw” Input flow data Hydrologic Ensemble Processor (models) Product generation system flow ensembles Streamflow Hydrologic Ensemble Post-processor Initial conditions and model parameters (e.g. DA) No specific uncertainty modeling in HEFSv1 Ensemble products

  14. Mesh ensemble forcings from short, medium, and long range techniques. medium range wx models mesoscale wx models Ensemble Forecasting Challenge long range global circulation models downscaling downscaling downscaling time variable downscaling forecaster skill climate forecasts and indexes

  15. HPC/RFC forecasts Ensembles (days 1-5) Short-Range Calibrated short- to long-range forcing ensembles GFS/GEFS forecasts Ensembles (Day 1-14) Medium-Range Merging Ensembles (out to 8/9 months) CFSv1/CFSv2 forecasts Long-Range Ensembles (out to one year) Climatology Meteorological Ensemble Forecast Processor

  16. Ensemble Forecast Challenge • Accurately incorporate the impacts of reservoirs and diversions • Ensure meaningful uncertainty information is retained beyond water control structure • Need operating rules for reservoirs • Reservoir models only approximate the actual operator decisions

  17. Ensemble Forecast Challenge of a different kind • Provide uncertainty information in a form and context that is useful to our customers • Education and training • Context, validation and verification • Compatibility with decision support tools • Realizing the full utility of this information Internal NWS customers (WFOs) External partners and customers (Water Managers, USACE, BoR, EMs, local communities, public)

  18. What is needed for partners? • Proving the skill/value in these forecasts • Verification Information • Event specific • Communicating effectively (understandable,formats,etc) • Commitment to overcoming hurdles (policy,legislative mandates, bureaucracy, process, education, etc) • Silver Jackets, MB forecasters group, and others part of solution? • Local knowledge • Closer to specific issues/hurdles

  19. Back-up Slides

  20. HEFS Service Level Objectives • Produce ensemble streamflow forecasts: • Seamlessly span lead times from one hour to one year • Calibrated (unbiased, accurate spread) • Spatially and temporally consistent (linkable) • Effectively capture the information from current NWS weather to climate forecast systems • Consistent with retrospective forecasts • Verified • Deliver a wide range of products

  21. Basin A Basin B • Ensure forecast ensembles maintain spatial and temporal relationships across many scales Ensemble Forecasting Challenge rainy + cold clear + warm snowing cloudy + hot Irrational outcomes • Similarly, ensure consistency between precipitation and temperature is preserved in the forecast ensembles.

  22. Ensemble Forecasting Challenge • Maintain coherence between deterministic and ensemble forecasts

  23. New York Dept. of Environmental Protection (NYDEP) Project • Water Management for part of NYC water supply system • Will optimize a decision support system based on retrospective simulation using past forecasts • Avoidance of building expensive water filtration system • Better management to limit turbidity violations • Requirements include: • High Priority • Daily time-step ensemble streamflow forecasts with two week lead time • Forecast updates daily • Hindcasts for retrospective period of several decades • Strong (but not perfect) consistency in methods between real-time forecasts and retrospective hindcasts • Lower Priority • Forecast lead times out to one year • More frequent forecasts (3-hourly) during flooding • Additional forecast variables associated with streamflow ensembles to use for water quality prediction, etc.

  24. NYCDEP OST for Operations Support RFC Inflow Ensembles

  25. MEFP Methodology Goal: Produce reliable ensemble forcings that capture the skill and quantify the uncertainty in the source forecasts. Key Idea: Condition the joint distribution of single-valued forecasts and the corresponding observations using the forecast. Use forecasts from multiple modelsto cover short- to long-range. Model the joint probability distribution between the single-valued forecast and the corresponding observation from historical records. Sample the conditional probability distribution of the joint distribution given the single-valued forecast. Rank ensembles based on the magnitude of the correlation coefficients between forecast and observation for the time scales and associated forecast sources. Generate blended ensembles (using Schaake Shuffle) iteratively for all time scales from low correlation to high correlation. 26

  26. Development and Implementation Plan • Testing and Implementation at 5 test RFCs • Development Release 1 – Completed • Workshop: April 2012 • Beta Test: May - Aug 2012 • Development Release 2 • Workshop: Sep 2012 • Beta Test: mid-late Sep 2012 – early Jan. 2013 • Development Release 3 • Workshop: late April 2013 • Beta Test: late April 2013 – late July 2013 • Final HEFSv1 • Beta Test: early Oct. ’13 – early Nov. ‘13 • Operational Readiness Review – early Nov. ’13 • Operational Implementation at 2-5 RFCs by Dec 31 2013 • Operational Implementation at remaining RFCs –2014

  27. NWS Mission and Goals NOAA NWS Mission “NOAA’s NWS provides weather, hydrologic, and climate forecasts and warnings for the United States, its territories, adjacent waters and ocean areas, for the protection of life and property and the enhancement of the national economy.” NOAA Weather Ready Nation Objectives Reduced loss of life, property, and disruption from high-impact events. Improved freshwater resource management Improved transportation efficiency and safety Healthy people and communities due to improved air and water quality services A more productive and efficient economy through environmental information relevant to key sectors of the U.S. economy services. 28

  28. Record Flooding WY 2011 March 6, 2011 http://water.weather.gov 29

  29. Advanced Hydrologic Prediction Service (AHPS) $60 million/10 year program (completion year of 2015) Over 3,500 forecast locations with new Web-based services AHPS 60% complete $766 million estimated annual recurring benefit (National Hydrologic Warning Council study) Expanding AHPS Coverage 30

  30. Met-model ensemble forecast system (MMEFS) • OHRFC, SERFC, NERFC, MARFC • Ensemble Preprocessor (EPP3) based approach (led by OHD) • CNRFC, CBRFC, NWRFC Short range Ensemble Forecasting underway at most RFCs • “HPC QPF” Approach • NCRFC, LMRFC, MBRFC, ABRFC

  31. Accessing AHPS Information Long range outlook Probability of non-exceedance 32

  32. Push for Enhanced Ensemble Forecasting • Well defined need for providing uncertainty estimates • Limitations of current NWS ensemble forecasting • HEFS science development maturing • Prototype ensemble forecasting underway at RFCs • CHPS implementation completed in 2011 • NYCDEP requirement for hydrologic ensembles

  33. Reservoir Releases Diversion Observed precipitation Forecast Temperature Forecast Precipitation CHPS Model Observed flow Model ExecutionQuality of forecast depends on inputs Model states Data availability and Future uncertainty

More Related