Transforming NOAA Water Resources Prediction

Transcription

1 Transforming NOAA Water Resources Prediction WestFAST May 18, 2017

2 T O O M U C H Presentation Outline Impetus for Change NOAA Water Initiative National Water Center (NWC) Status and Plans Partnerships New Prediction Capabilities: The National Water Model (NWM) Summary 2 P O O R Q U A L I T Y T O O L I T T L E

3 Impetus for Change Growing Water Challenges Multiple Threats Population growth and economic development are stressing water supplies and increasing vulnerability A changing climate is impacting water availability and quality, increasing uncertainty An aging water infrastructure is forcing critical, expensive decisions Socio-economic risks of floods and droughts are escalating 3

4 National Academy of Sciences Report, 2012 Weather Services for the Nation: Becoming Second to None Findings Recommendations A significant gap exists between the state of hydrologic science and current hydrologic operations The level of sophistication, representation of processes, and characterization of uncertainties in external research and operational communities outpace those used in NWS hydrology operations NWS Hydrologic Forecasters are extensively in the forecast loop Improve pathways for collaboration & accelerate R2O Establish a hydrologic prediction testbed as part of the National Water Center Implement a consistent framework for hydrologic prediction skill assessment Transition RFC forecasters to over the loop enabling a shift in focus to model and product development, forecast interpretation, and decision support Lack of skill in modern computational programming, construction and use of new Earth System Models, current hydrologic data assimilation methodologies, and preparation and interpretation of meaningful ensemble predictions Hydrologist staff require re-education and continual retraining to enable adoption of state-of-the-art prediction methodologies Instill evolutionary culture Add value to hydrologic forecasts through the use of more advanced models, data assimilation and employment of more sophisticated ensemble techniques 4

5 Integrated Water Resources Science and Services (IWRSS): Partners and Missions Collaborative Science-Based Solutions to Address Societal Needs Water Information: Collects and disseminates reliable, impartial, and timely information needed to understand the Nation's water resources to minimize loss of life and property from natural disasters US Army Corps of Engineers Water Management: Strengthens our Nation's security, energizes the economy, and reduces risks from disasters Water Prediction: Provide weather, water, and climate data, forecasts and warnings for the protection of life and property and enhancement of the national economy. FEMA Response and Mitigation: Supports our citizens and first responders to ensure that as a nation we work together to build, sustain, and improve our capability to prepare for, protect against respond to, recover from and mitigate all hazards IWRSS Partnership anticipated to expand over time IWRSS Partnership will expand over time 5

6 Initial Stakeholder Priorities Flooding Water Quality Water Availability Drought Climate Change Need integrated understanding of near- and long-term outlook and risks Provide consistent, high resolution ( street level ), integrated water analyses, predictions and data to address critical unmet information and service gaps Transform information into intelligence by linking hydrologic, infrastructural, economic, demographic, environmental, and political data Integrate Social Science to create Actionable Water Intelligence 6

7 NOAA Water Initiative Overarching Goal: Transform water information service delivery to better meet and support evolving societal needs User-Oriented, informed by Regional and National Conversations on Integrated Water Information Leverages the National Water Center to provide next-generation, science-based water information and decision support services. Calls for collaboration across federal agencies and with partners outside government Released in December

8 NOAA Water Initiative Key Objectives Improve Modeling & Prediction Enhance Water-Related Observations Accelerate Research & Development Strengthen Decision Support Tools Informed by social science Enhance Service Delivery

9 Integrated Water Prediction Setting the Stage for Transformation Centralized Water Forecasting Demonstration (2015) Enhanced Water Prediction Capability (2016) Integrated Water Prediction (2017 Proposed) National Water Model (NWM) Development and Demonstration Centralized Water Resources Data Services Water Resources Test and Evaluation Service Hyper-Resolution Modeling Real-Time Flood Forecast Inundation Mapping Enhance Impact- Based Water Resources Decision Support Services Stand up the National Water Center Operations Center Couple terrestrial freshwater and coastal estuary models for total water predictions in the coastal zone Increase high performance computing capacity 9

10 National Water Center Initial Operating Capacity: May 26, 2015 A catalyst to transform NOAA s water prediction program Mission: Nationally Integrated Water Prediction Center of excellence for water resources science and prediction Interagency and Academia Collaboration Operations Center for water resources common operating picture and decision support services Proving ground to accelerate research to operations Earth system modeling and geointelligence 10

11 National Water Center Annual Innovators Program Partnership between NWS and the academic community via an Interagency Agreement with the National Science Foundation and CUAHSI with two fundamental goals: Provide a Framework for Collaboration Target Emerging Technologies Year one included a competitive Summer Institute for 44 graduate students from 19 Universities at the National Water Center, June 1 to July 17, 2015 Demonstrated ability to simultaneously model the entire continental United States river network at high spatial resolution, in near real-time for 2.7 million stream reaches Year two included a competitive Summer Institute for 34 graduate students from 21 Universities at the National Water Center, June 6 to July 21, 2016 Demonstrated the ability to generate flood inundation maps utilizing NWM output Engaged social scientists and stakeholders from the Fire, Police and Emergency Management Communities to explore ways to best communicate water information

12 National Water Model Version v1.1 implemented in May 2017 Spatially continuous estimates of major water cycle components (snowpack, soil moisture, channel flow, major reservoir inflows, flood inundation) Operational forecast streamflow guidance for currently underserved locations: 3,600 forecast points -> 2.7 million (NHDPlus river reaches) Implement an Earth system modeling architecture with National Center for Atmospheric Research (NCAR) that permits rapid model evolution of new data, science and technology (i.e. WRF-Hydro) Ongoing Water Resource Evaluation Service (WRES) and Data Service (WRDS) to compliment implementation efforts Current NWS River Forecast Points overlaid with NWM Stream Reaches

13 1. NWM Forcing Engine (1 km grid) National Water Model Initial Operating Capability: Model Chain NWM uses NCAR supported community WRF-Hydro system NWM: WRF-Hydro: 2. NoahMP LSM (1 km grid) 2-way coupling 3. Terrain Routing Module (250 m grid) 4. NHDPlus Catchment Aggregation (2.7M unique catchments and river reaches) 5. Channel & Reservoir Routing Modules Forecast Products

14 National Water Model v1.1 Analysis and Forecast Cycling Configurations Cycling Forecast Forcing Outputs Model output available at:

15 Movie (slide show to view)

16 NWM V1.1 Forecast Assessment: Midwest Flooding April-May 2017 Illustrating the challenges of hydrologic forecasting NWM Forecast for Illinois River Near Tahlequah, OK on 4/27/17 Three Days Before Major Flooding Strong agreement Three of four simulations capture peak magnitude Event Summary Indication of flooding 9 days in advance Large forecast variability as precipitation forcing evolves Good though somewhat low-biased streamflow forecasts at this location Mixed performance across region Quality of forcing impacts NWM forecast 16

17 Beyond Streamflow Additional NWM Forecast Guidance Midwest Flooding April-May 2017 NWM Ponded Water Field (mm) at 23Z on April 29 th, 2017 Real-time NWM Analysis Short-Range NWM Forecast (18-hour lead time) Local Storm Report Flash Flood Warning Local Storm Report Flash Flood Warning NWM outputs several other water resource fields which can offer insight into potential for flooding Streamflow anomaly, near-surface soil saturation, depth to saturation, and ponded water Ponded water correlates well with placement of flash flood warnings and local storm reports Good performance of 18-hour forecast 17

18 Beyond Streamflow Additional NWM Hydrologic Guidance NWM Analysis Output at 23Z on April 29th, 2017 Depth to Saturation Hourly Precipitation Surface Layer Soil Moisture Saturation Ponded Water Flash Flood Watch Flash Flood Warning

19 Preventing Information Overload Prototyping Innovative Displays of NWM Information End users cannot click through 2.7 million stream reaches they need data packaged in a way that draws their attention to critical areas NWM Short Range Six-Hour Time Lagged Ensemble Forecast (probability of exceeding high flow threshold, based on 12 members) Areas of high probability align with flash flood watches and warnings Valid at 23Z on April 29 th, 2017 High Flow Threshold ln(flow) - ln(annual flow) > 4

20 Multivariate Verification: Focus on Precipitation and Streamflow HRRR QPF Bias at mm/h threshold Precipitation-Streamflow Bias Comparison Precipitation forcing is critical to NWM forecasts Active focus on precipitation forecast assessment and impact on streamflow Special emphasis on mountainous terrain in Western US 20

21 Focus on the West: NWM Version 1.1 Snow Modeling Enhancements NoahMP multi-layer snow model was updated to modify surface roughness length formulation. Outcome: Improved characterization of peak snowpack and snow ablation

22 Focus on the West: NWM Version 1.1 Snow Modeling Enhancements V1.1 implemented in May Mean absolute error, correlation and bias of SWE improved versus V1.0 Validation using SNOTEL, SNODAS, NASA Lidar, MODIS/MODSCAG, VIIIRS snow cover Work underway to understand role of forcing versus snow model physics on model skill Month-Long CONUS-Wide Assessment V1.0 Change in NWM SWE MAE from V1.0 to V day NWM forecasts initialized on 12/24/16 (blue=improvement) Snotel Snowpack improved in critical water supply areas V1.1 SNODAS used as baseline observation Snotel

23 NWM: Experimental Seasonal Water Supply Forecasts Began in WY2016 for Upper Rio Grande Compact basins Custom NWM WRF-Hydro configuration Colorado-NCAR-OWP collaboration No water management Utilized classic ESP -based methodology 2004-present period NLDAS2 meteorological forcings (with some downscaling) Posterior bias correction applied to accumulated Apr 1-Sep 30 flows Results comparable to those of WGRFC and NRCS methods WY2017 for all of Colorado Mountain Headwater locations New domain Utilizing NOAA-NWM physics configuration and calibrations Still using NLDAS2 meteorological forcings The WRF-Hydro forecast takes into account a much wider range of variables than the older NRCS forecasting technique, and is providing us with more accurate forecasts as we face a changing landscape caused by climate change, beetle kill, and wildfires in our watersheds. -Craig Cotton, CDWR District Engineer for the Upper Rio Grande WY 2016 WY 2017 WY 2017

24 Flood Inundation Mapping Forecast discharge with National Water Model 2. Convert discharge to depth using rating curve 3. Convert depth to inundation using HAND (relative elevation of land surface cell above cell in NHDPlus stream to which it flows) 3 Source: Yan Liu and Hu Hao, University of Illinois at Urbana-Champaign David R. Maidment University of Texas at Austin

25 Summary NOAA s Water Services are Evolving We are building a foundation for change but have a long way to go Stakeholder input will continue to inform future development activities, and the delivery and evolution of new services New Organization, Cornerstone Facility and Philosophy Office of Water Prediction/National Water Center Collaborative, cross-noaa, interagency, academic partnerships Implementing State-of-the-Art Technical Approach Impact-based decision support services underpinned by geo-intelligence NWM establishes foundation for sustained improvement in water prediction and first ever nationally consistent operational hydrologic forecasting capability Focus moving forward will be on expanding evaluation capabilities, working with end users to improve NWM s base capabilities and establish new connections NWM V1.1 implemented in May, will be followed by V1.2 in Fall of 2017 Longer-term efforts will focus on improved regulation, domain expansions, coupling to coastal models, and hyper-resolution, low flow/drought and water quality modeling 25

26 Backup Slides 26

27 NWM: Experimental Seasonal Water Supply Forecasts Example plots of ESP water supply forecast traces (left) and background model simulation performance (right) Background multi-year performance used as a statistical bias correction for current forecasts Total seasonal water supply forecasts (volume in thousands of acre-feet) bias corrected and calculated as median and mean of these ensembles

28 Future Plans: Upgrading to NWM V1.2 Version 1.0 implemented August 2016, V1.1 implemented on May 8 th Version 1.2 upgrades in progress for Fall 2017 release Expanded calibration of soil, vegetation, and groundwater parameters to 1,100 basins across CONUS Updates to NHDPlus geofabric to fix topology errors and extension to USdraining Mexico basins Inclusion of additional lakes and reservoirs Improved harmonization between DEM and NHDPlus channel network Improved stream flow data assimilation using time-varying bias-correction Capability for dynamic parameter updates 28