1 northeast river forecast center national weather service implementation of the community...

26
1 Northeast River Forecast Center National Weather Service Implementation of the Community Hydrologic Prediction System David Vallee Hydrologist-in-Charge Northeast RFC NROW November 4-5, 2009

Upload: mia-miah-helin

Post on 14-Dec-2015

215 views

Category:

Documents


0 download

TRANSCRIPT

1

Northeast River Forecast CenterNortheast River Forecast Center

National Weather Service Implementation of the CommunityHydrologic Prediction System

David ValleeHydrologist-in-ChargeNortheast RFC

NROWNovember 4-5, 2009

National Weather Service Implementation of the CommunityHydrologic Prediction System

David ValleeHydrologist-in-ChargeNortheast RFC

NROWNovember 4-5, 2009

• First – a bit about NERFC

• The Community Hydrologic Prediction System

– Background and history

– Why are we doing this?

– How are we doing it?

• First – a bit about NERFC

• The Community Hydrologic Prediction System

– Background and history

– Why are we doing this?

– How are we doing it?

Outline

NERFC Service AreaMajor river basins

include Genesee, Hudson, Mohawk, Housatonic, Connecticut, Merrimack , Blackstone, Pawtuxet,

Kennebec ,Penobscot and Saint John

~180 forecast points

Staffing profile: Hydrologist-in-ChargeService Coordination Hydrologist and Development & Operations Hydrologist 4 Senior Hydrologists & 1 Senior Hydromet Analysis & Support Meteorologist (HAS)3 Hydrologists & 2 HAS1 Admin Support/Hydro Technician

Thirteen River Forecast Centers

River Forecast Center Responsibilities Calibrate and implement

hydrologic /hydraulic models Perform robust Data QC Provide 48-72 hour QPF and

Temperature forecasts River flow and stage

forecasts at 180 locations (220 basins)

Flash Flood Guidance Ensemble streamflow

predictions Ice Jams and Dam Breaks Water Supply forecasts Partner with NOAA Line

Offices to address issues relating to Hazard Resiliency, Water Resource Services, Ecosystem Health and Management, and Climate Change

Moderate flooding - Connecticut River at Portland, CT.

Met-member Based Short Range Ensemble River Forecast System

Unbiased GEFS and SREF model forcings (temp/precip)

Runs 2x-day for each model

Decision Support at the local level

http://www.erh.noaa.gov/nerfc/self_brief.shtml

Getting the high priority hydro-meteorological information at their fingertips!

Providing Remote or On-site Support!

Whittenton Dam Crisis

MontpelierIce Jam

NOAA Regional CollaborationHarmful Algae Bloom Research

http://www.oceanservice.noaa.gov/redtide

NARR 1000 mb monthly wind anomolies (May 2005 and June 2009)

Future Services: National Inundation Mapping

Program

x

11

AlgorithmServices

DataServices

ControlServices

DisplayServices

SecurityServices

Community Hydrologic Prediction SystemCHPS

Community Hydrologic Prediction SystemCHPS

NOAA’s National Weather Service (NWS) River Forecast Centers (RFCs) are key partners in the provision of water resource information.

The Community Hydrologic Prediction System (CHPS) will enable NOAA’s water research, development enterprise and operational service delivery infrastructure to be integrated and leveraged with other federal water agency activities, academia, and the private sector.

NOAA’s National Weather Service (NWS) River Forecast Centers (RFCs) are key partners in the provision of water resource information.

The Community Hydrologic Prediction System (CHPS) will enable NOAA’s water research, development enterprise and operational service delivery infrastructure to be integrated and leveraged with other federal water agency activities, academia, and the private sector.

                                  

            

So Why the Move to CHPS?To meet the growing demands of our customers

Floods, Drought, Water Supply, HABs,Power Generation, Fisheries Management,Ecosystem Health, Water Quality, Salt Water Intrusion, Dam/Levee Failures,Tropical Cyclone Impacts, Climate Change Impacts

Why CHPS?Why CHPS?

• NWSRFS was a great architecture originally developed for use on mainframe computers

– lacks modern modularity– difficult to add new models and

techniques– inhibits collaboration and research

to operations

• CHPS will allow:

– greater ease in implementing new models

– greater collaboration with agency partners, universities, international community

• NWSRFS was a great architecture originally developed for use on mainframe computers

– lacks modern modularity– difficult to add new models and

techniques– inhibits collaboration and research

to operations

• CHPS will allow:

– greater ease in implementing new models

– greater collaboration with agency partners, universities, international community

13

1970s 1980s 1990s 2000s 2003 today

• The CHPS Acceleration Team – or CAT pursuit of CHPS

– ABRFC, NCRFC, NWRFC volunteered to work w/ OHD. CNRFC later drafted.

– Criteria established through evaluation of NWSRFS strengths and weaknesses (Apex facilitation, available from HSEB)

– Evaluate Delft-FEWS as a CHPS candidate

– FEWS = Flood Early Warning System

• The CHPS Acceleration Team – or CAT pursuit of CHPS

– ABRFC, NCRFC, NWRFC volunteered to work w/ OHD. CNRFC later drafted.

– Criteria established through evaluation of NWSRFS strengths and weaknesses (Apex facilitation, available from HSEB)

– Evaluate Delft-FEWS as a CHPS candidate

– FEWS = Flood Early Warning System

CAT River Forecast Centers CAT River Forecast Centers

Beta Sites(2003) 2006 2007 2008 2009 2010 2011

Exploration phase in 2003

FEWS Architecture

15

(2003) 2006 2007 2008 2009 2010 2011

What is FEWS?– JAVA based system and repository for hydrologic and

hydraulic modeling

– Client/server architecture

– Mature system developed by Delft Hydraulics|WL (now Deltares), a Dutch foundation, since the 1990s

– Deterministic and ensemble capabilities

– Employs Postgres database, system synchronization• Backup and stand-by

– Inputs are grids!

What is FEWS?– JAVA based system and repository for hydrologic and

hydraulic modeling

– Client/server architecture

– Mature system developed by Delft Hydraulics|WL (now Deltares), a Dutch foundation, since the 1990s

– Deterministic and ensemble capabilities

– Employs Postgres database, system synchronization• Backup and stand-by

– Inputs are grids!

FEWS Users

16

CAT Recommendation

• Proceed with implementation of Delft-FEWS as the CHPS software infrastructure.

• Target operational use at all RFCs within 3 years.

• Resource implications– Re-direct OHD resources toward CHPS/FEWS– Terminate “dead-end” NWSRFS enhancements– Align/re-evaluate HSMB-oriented “Research to Operations”

(RTO) projects

• Proceed with implementation of Delft-FEWS as the CHPS software infrastructure.

• Target operational use at all RFCs within 3 years.

• Resource implications– Re-direct OHD resources toward CHPS/FEWS– Terminate “dead-end” NWSRFS enhancements– Align/re-evaluate HSMB-oriented “Research to Operations”

(RTO) projects

17

Approved by Office Director – Gary Carter – Jan 17, 2008Approved by Office Director – Gary Carter – Jan 17, 2008

(2003) 2006 2007 2008 2009 2010 2011

Implementation Strategy

• Port models that require calibration

– Hydrologic/Hydraulic Models– Soil Moisture Models– Snow Models– Reservoir ModelsBASEFLOW, SARROUTE, CONS_USE, LAG/K, LAY-COEF, TATUM,TIDEREV, MUSKROUT, RES-J, RSNWELEV, SNOW-17, CHANLOSS,STAGE-Q, SSARRESV, STAGEREV, UNIT-HG, RES-SNGL, SAC-SMA

• Create adapters for other NWSRFS routines we wish to maintain

• Create adapters for several new models (e.g. HEC-RAS)

• Port models that require calibration

– Hydrologic/Hydraulic Models– Soil Moisture Models– Snow Models– Reservoir ModelsBASEFLOW, SARROUTE, CONS_USE, LAG/K, LAY-COEF, TATUM,TIDEREV, MUSKROUT, RES-J, RSNWELEV, SNOW-17, CHANLOSS,STAGE-Q, SSARRESV, STAGEREV, UNIT-HG, RES-SNGL, SAC-SMA

• Create adapters for other NWSRFS routines we wish to maintain

• Create adapters for several new models (e.g. HEC-RAS)18

(2003) 2006 2007 2008 2009 2010 2011

19

NWSRFSHydrologic

Models

Ensemble ofPrecipitationTemperatureTime Series

Current Conditions:SoilSnowReservoir LevelsStreamflow

HistoricalData

Forecast

• FEWS will provide for existing level of ensemble operations, products, and services

• Combination of porting existing software meshed with existing FEWS capabilities

• FEWS will provide for existing level of ensemble operations, products, and services

• Combination of porting existing software meshed with existing FEWS capabilities

Ensemble Streamflow Prediction Services

Models in FEWS

20

Model Type Supplier/Owner Country

ISIS Hydrodynamics HR/Halcrow UK

PDM Rainfall-Runoff CEH UK

TCM Rainfall-Runoff CEH UK

KW Routing (kinematic wave) CEH UK

PACK Snow Melt CEH UK

ARMA Error Correction CEH UK

PRTF Event Based RR PlanB UK

TRITON Surge propagation/Overtopping PlanB UK

STF Transfer functions EA UK

DODO Routing (layered Muskingum) EA UK

MCRM Rainfall-Runoff EA UK

Modflow96/VKD 3D groundwater Deltares/Adam Taylor Netherlands/UK

Mike11 Hydrodynamics DHI Denmark

NAM Rainfall-Runoff DHI Denmark

LISFLOOD Distributed Rainfall-Runoff JRC Italy

TOPKAPI Rainfall-Runoff Univ. of Bologna Italy

HBV Rainfall-Runoff (inc snowmelt) SHMI Sweden

Vflo Distributed Rainfall-Runoff Vieux & Assiciates USA

SWMM Urban Rainfall-Runoff USGS USA

HEC-RAS Hydrodynamics USACE USA

Snow17 Snow Melt NWS USA

SACSM Rainfall-Runoff NWS USA

Unit-H Unit-Hydrograph NWS USA

PRMS Rainfall-Runoff Univ. of Karlsruhre Germany

SynHP Hydrodynamics BfG Germany

SOBEK Hydrodynamics, Water Quality, RR Deltares Netherlands

SOBEK-2d Linked 1d/2d inundation modelling Deltares Netherlands

Sacramento Rainfall-Runoff Deltares Netherlands

RIBASIM Water distribution + Reservoir Deltares Netherlands

REW Distributed Rainfall-Runoff Deltares Netherlands

DELFT3D 2/3D Hydrodynamics/ Water quality Deltares Netherlands

TWAM 2D Hydrodynamics PlanB UK

Implementation Strategy

• Two tiers of deployment 4 CAT RFCs: CN, NW, AB, NE

– Initial hardware delivered Oct ‘08 (Stand-by System)

– Initial migration software and training occurred in February ‘09

– Second suite of operational hardware delivery fall 2009(On-line System)

– Full Baseline Operating Capability BOC-1 arrived in October 2009.

• Two tiers of deployment 4 CAT RFCs: CN, NW, AB, NE

– Initial hardware delivered Oct ‘08 (Stand-by System)

– Initial migration software and training occurred in February ‘09

– Second suite of operational hardware delivery fall 2009(On-line System)

– Full Baseline Operating Capability BOC-1 arrived in October 2009.

21

(2003) 2006 2007 2008 2009 2010 2011

Parallel operations are about to commence at NERFC!!

Implementation Strategy

• Two tiers of deployment

Remaining 9 RFCs (BOC-2)– Initial hardware delivery in October 2009– Migration/Systems training – Fall 2009– Migration begins January 2010– Parallel operations by October 2010– Goal is to “Retire” NWSRFS by January 2011– Entire system needs to be baselined into

AWIPS II configuration at the RFCs

• Two tiers of deployment

Remaining 9 RFCs (BOC-2)– Initial hardware delivery in October 2009– Migration/Systems training – Fall 2009– Migration begins January 2010– Parallel operations by October 2010– Goal is to “Retire” NWSRFS by January 2011– Entire system needs to be baselined into

AWIPS II configuration at the RFCs

22RFCs not required to drop NWSRFS until “fully ready”

(2003) 2006 2007 2008 2009 2010 2011

Ultimate Outcome:

• CHPS will include the same science algorithms initially – but tremendous potential to expand

• Improved Efficiency for Forecaster

– Greatly enhanced visualization, gridded forcings, and the ability to display time series of many more variables

• Improved Tools for “What If” Scenarios

• Improved Capabilities for New Science Integration

– Through the use of Adaptors; HEC-RAS, RESSIM

– Availability to use models in FEWS from other partners

• CHPS will include the same science algorithms initially – but tremendous potential to expand

• Improved Efficiency for Forecaster

– Greatly enhanced visualization, gridded forcings, and the ability to display time series of many more variables

• Improved Tools for “What If” Scenarios

• Improved Capabilities for New Science Integration

– Through the use of Adaptors; HEC-RAS, RESSIM

– Availability to use models in FEWS from other partners

24

Duty (On-line) and Stand-by system

– All data is synchronized between system via the Master Controller

– NWS testing synchronization to paired offices – similar to WFOs

Duty (On-line) and Stand-by system

– All data is synchronized between system via the Master Controller

– NWS testing synchronization to paired offices – similar to WFOs

Operator ClientOperator Client

Client

Client

Client

Database ServerDatabase Server

Master Controller Server

Master Controller Server

Forecasting Shell Server

Forecasting Shell Server

Master ControllerJBoss

Tomcat

Operator Client

Forecasting Shell

Central Database

(PostgreSQL)

Forecasting Shell

Master ControllerJboss

TomcatForecasting Shell

Central Database

(PostgreSQL)

Forecasting Shell

Etc Etc

ForecastingArchive

(PostgreSQL)

Duty System

Stand-by System

Architecture

Operator ClientOperator Client

ClientClient

Client

Database ServerDatabase Server

Master Controller Server

Master Controller Server

Forecasting Shell Server

Forecasting Shell Server

Master ControllerJBoss

Tomcat

Operator Client

Forecasting Shell

Central Database

(PostgreSQL)

Forecasting Shell

Master ControllerJboss

TomcatForecasting Shell

Central Database

(PostgreSQL)

Forecasting Shell

Etc Etc

ForecastingArchive

(PostgreSQL)

Operator ClientOperator Client

ClientClient

Client

Database ServerDatabase Server

Master Controller Server

Master Controller Server

Forecasting Shell Server

Forecasting Shell Server

Master ControllerJBoss

Tomcat

Operator Client

Forecasting Shell

Central Database

(PostgreSQL)

Forecasting Shell

Master ControllerJboss

TomcatForecasting Shell

Central Database

(PostgreSQL)

Forecasting Shell

Etc Etc

ForecastingArchive

(PostgreSQL)

Operator ClientOperator Client

ClientClient

Client

Database ServerDatabase Server

Master Controller Server

Master Controller Server

Forecasting Shell Server

Forecasting Shell Server

Master ControllerJBoss

Tomcat

Operator Client

Forecasting Shell

Central Database

(PostgreSQL)

Forecasting Shell

Master ControllerJboss

TomcatForecasting Shell

Central Database

(PostgreSQL)

Forecasting Shell

Etc Etc

ForecastingArchive

(PostgreSQL)

RFC NWSHQ NOHRSC Operator ClientOperator Client

ClientClient

Client

Database ServerDatabase Server

Master Controller Server

Master Controller Server

Forecasting Shell Server

Forecasting Shell Server

Master ControllerJBoss

Tomcat

Operator Client

Forecasting Shell

Central Database

(PostgreSQL)

Forecasting Shell

Master ControllerJboss

TomcatForecasting Shell

Central Database

(PostgreSQL)

Forecasting Shell

Etc Etc

ForecastingArchive

(PostgreSQL)

WFO??

25

Display Concepts

26

Northeast River Forecast CenterNortheast River Forecast Center

National Weather Service Implementation of the CommunityHydrologic Prediction System

David ValleeHydrologist-in-ChargeNortheast RFC

NROW XINovember 4-5, 2009

National Weather Service Implementation of the CommunityHydrologic Prediction System

David ValleeHydrologist-in-ChargeNortheast RFC

NROW XINovember 4-5, 2009