The NOAA Operational Numerical Guidance System: Meeting the WoF and WRN Challenge
description
Transcript of The NOAA Operational Numerical Guidance System: Meeting the WoF and WRN Challenge
WoF Modeling Meeting 2/4/13 1
William. M. LapentaActing Director
Environmental Modeling Center
NOAA/NWS/NCEP
Geoff DiMego , John Derber , Yuejian Zhu , Hendrik Tolman , Vijay Tallapragada, Shrinivas Moorthi , Mike Ek , Mark Iredell , Suru Saha
NCEP
The NOAA Operational Numerical Guidance System:
Meeting the WoF and WRN Challenge
AMS Future of the Weather Enterprise 11/27/12
NOAA Center for Weather and Climate Prediction: “A Game Changer”
2
A.K.A.—the new building….• Four-story, 268,762 square foot
building in Riverdale, MD will house 800+ Federal employees, and contractors
• 5 NCEP Centers (NCO, EMC, HPC, OPC, CPC)
• NESDIS Center for Satellite Applications and Research (STAR)
• NESDIS Satellite Analysis Branch (SAB)• OAR Air Resources Laboratory
• Includes 465 seat auditorium & conference center, library, deli, fitness center and health unit
• Includes 40 spaces for visiting scientists
• Represents a “Game Changer” in our ability to do business
WoF Modeling Meeting 2/4/13
NOAA Operational Numerical Guidance Supports the Agency Mission
– Numerical Weather Prediction at NOAARelated to ability to meet service-based metrics (below)
– National Weather Service GPRA* Metrics
(* Government Performance & Results Act)Hurricane Track and Intensity Winter Storm WarningPrecipitation Threat Flood WarningMarine Wind Speed and Wave Height
– Customer Service ProviderOperational numerical guidance provides foundational tools
used by Government, public and private industry to Improve public safety, quality of life and make business decisions that drive US economic growth
3
Lead Time and
Accuracy!
WoF Modeling Meeting 2/4/13 4
DRAFT Storm Prediction Center Potential Products and Services
2015 2017 2022•Probabilistic Outlooks of Tornado, Severe Hail, and Severe Wind hazard types extended to Day 2
•Probabilistic Outlooks of Tornado, Severe Hail, and Severe Wind Hazard Types Extended to Day 3
•Probabilistic Outlooks of Tornado, Severe Hail, and Severe Wind hazard types extended to Day 7
•Day 1 Probabilistic Outlooks of Tornado, Severe Hail, and Severe Wind Hazards for 4-hr periods
•Probabilistic Outlooks of Tornado, Severe Hail, and Severe Wind Hazards for 4-hr Periods extended to Day 2
•Probabilistic Outlooks of Tornado, Severe Hail, and Severe Wind Outlooks for 4-hr Periods Extended to Day 3
•Probabilistic Short-Term Forecasts of Tornado and Significant Tornado Hazards issued every 2-4 hrs
•Hourly Updated, Short Term Probability Forecasts for Tornado, Severe Hail, Severe Wind, and Significant Severe Hazards to support transition to continuously evolving public severe weather watches
•Seasonal Severe Weather Outlooks (with CPC)
•Hourly Day 1 Thunderstorm Outlooks issued every 2 hours, including “dry” lightning conditions critical to wild fire initiation
•Week 2 and Monthly Severe Weather Outlooks (with CPC)
Potential Operational Products and Services
WoF Modeling Meeting 2/4/13
DRAFT Storm Prediction Center Desired Numerical Guidance Attributes
2015 2017 2022•North America Short-Range Mesoscale Ensemble: 12-15 km ~30 member multi-model/multi-physics/multi-IC ensemble with EnKF/hybrid DA, run every 6 hrs with forecasts to 84 hrs
•North America Short-Range Mesoscale Ensemble: 12-15 km ~40 member multi-model/multi-physics/multi-IC ensemble with EnKF/hybrid or newer-state-of-the-art DA, run every 6 hrs with forecasts to Day 4*
•Global Mesoscale Ensemble Forecast System: 10-12 km global domain 40-50 member ensemble with advanced state-of-the-science DA, run every 6 hrs with forecasts to Day 7-10*
•CONUS Storm Scale Ensemble Forecast (SSEF): 4 km CONUS ~10 member multi-model/multi-physics/multi-IC storm scale ensemble with EnKF/hybrid DA, issued every 6-hrs with forecasts to 48-60 hrs*
•CONUS Storm Scale Ensemble Forecast (SSEF): 3 km CONUS ~15 member multi-model/multi-physics/multi-IC storm scale ensemble with EnKF/hybrid or newer state-of-the-art DA, issued every 6-hrs with forecasts to 48-60 hrs*
•CONUS SSEF: 2 km CONUS 30-50 member multi-model/multi-physics/multi-IC storm scale ensemble with advanced state-of-the-science DA, issued every 6-hrs with forecasts to 48-60 hrs*
•Movable domain update SSEF: 2-3 km movable regional domain ~7 member storm scale ensemble with EnKF/hybrid run every 2-4 hrs with forecasts to 15-18 hrs*, focused on “severe weather of the day” areas
•Movable domain update SSEF: 2 km movable regional domain 10-15 member storm scale ensemble with EnKF/hybrid or newer state-of-the-art DA ,run every 1-2 hrs with forecasts to 18-24 hrs*, focused on “severe weather of the day” areas
•Movable domain update SSEF: 1 km movable regional domain 20-30 member storm scale ensemble with advanced state-of-the-science DA run every 1 hr with forecasts to 18-24 hrs*, focused on “severe weather of the day” areas
•Stormscale 3D Analysis: 4 km EnKF/hybrid DA, CONUS storm scale analysis updated every 1-hr.
•Stormscale 3D Analysis: 2 km EnKF/hybrid or newer state-of-the-art DA, CONUS storm scale analysis updated every 15 min
•Stormscale 3D Analysis: 2 km EnKF/hybrid or newer state-of-the-art DA, CONUS storm scale analysis updated every 5 min
•Appropriate RAP (HRRR), Global deterministic / ensemble, and CFS capacities
•Appropriate RAP (HRRR), Global deterministic / ensemble, and CFS capacities
•Stretch Goal - Warn-on-Forecast Prototype Ensemble: Develop <1 km multi-WFO domain 10-20 member nested SSEF, advanced state-of-the science DA, run as needed every 5-10 min with forecasts to 1-2 hrs*
*Model and data assimilation (DA) complexity will be impacted strongly by High Performance Computing capabilities.
Desired Attributes of NOAA Operational Numerical Guidance System
5
• Inside every HRRRE member• Movable domain update SSEF: • 1 km movable regional domain • 20-30 member storm scale ensemble • advanced state-of-the-science DA • run every 1 hr with forecasts to 18-24 hr• focused on “severe weather of the day”
areas
AFWA NWP 11 Sept 2012
Plans for a National Mesoscale Ensemble System
• Convergence of NAM, RAP, HRRR and SREF between 2016 and 2018
• System must meet requirements of today and future
– Provide NextGen Enroute AND terminal guidance – Provide PROBABILITY guidance with full Probability
Density Function specified– Address Warn-on-Forecast as resolutions evolve
towards ~1 km
• Core elements– Common data assimilation system– Code Infrastructure and management system
• Potential Attributes of the North American Rapid Refresh Ensemble (NARRE)
– Hourly updated with forecasts to 24 hours– Control data assimilation cycles with 3 hour pre-
forecast period– 84 hr forecasts are extensions of the 00z, 06z, 12z, &
18z runs6
WoF Modeling Meeting 2/4/13
7
Air Quality
WRF NMM/ARWWorkstation WRF
WRF: ARW, NMMNMMBGFS, Canadian Global Model
Regional NAMWRF NMMB
North American Ensemble Forecast System
Hurricane GFDLHWRF
GlobalForecastSystem
Dispersion
ARL/HYSPLIT
Forecast
Severe Weather
Rapid Refreshfor Aviation
Climate ForecastSystem
Short-RangeEnsemble Forecast
NOAA’s OperationalNumerical Guidance Suite
GFS MOM4NOAH Sea Ice
NOAH Land Surface Model
Coupled
Global DataAssimilation
OceansHYCOM
WaveWatch III
NAM/CMAQ
7
Regi
onal
DA
Satellites + Radar99.9%
~2B Obs/Day
NOS – OFS• Great Lakes• Northern
Gulf of Mex• Columbia R.
Bays• Chesapeake• Tampa • Delaware
SpaceWeather
ENLIL
Regi
onal
DA
Sea Nettle
Forecast
7
WoF Modeling Meeting 2/4/13
Num
ber o
f Nod
es
High Water Mark 2010
24-h Snapshot 20 August 2012
Time of Day (UTC)
00 06 12 18 00
Numerical Guidance Suite Execution on the Operational NOAA Supercomputer
8
CFS NAM GFSGEFS SREF
WoF Modeling Meeting 2/4/13
Operational Compute Resource Consumption (%) by Component
Perc
ent o
f Tot
al C
PU
Numerical Guidance System Component
• Statistics accumulated over a 24-h period
• Total 64.9% of production machine
• NAM, CFS and SREF compose 29.7% of production
9
WoF Modeling Meeting 2/4/13
Global Data AssimilationSystem Upgrade
• Hybrid system– Most of the impact comes from this
change– Uses ensemble forecasts to help
define background error
• NPP (ATMS) assimilated– Quick use of data 7 months after
launch
• Use of GPSRO Bending Angle rather than refractivity– Allows use of more data (especially
higher in atmos.)– Small positive impacts
• Satellite radiance monitoring code– Allows quicker awareness of problems (run
every cycle)– Monitoring software can automatically detect
many problems
• Partnership between research and operations– (NASA/GMAO, NOAA/ESRL, Univ OK, and
NOAA/NCEP)
• Consolidation across systems– Unify operational data assimilation system
for global, regional and hurricane applications
– Cost effective—O&M– Configuration management 10
Implemented 22 May 2012
WoF Modeling Meeting 2/4/13 11
NCEP Closing the International GapJune, July, August 500hPa Geopotential RMSE
NCEP achieved significant improvement in 2012 for day 3 and beyond
NCEP is now similar to UKMO skill in this metric and AC
Meteo-FrCMCNCEPUKMOECMWF
2011 2012Solid line lower than dashed
indicates improvement between 2011 and 2012
NCEP Only System to show improvement between 2011 and 2012
WoF Modeling Meeting 2/4/13 12
Gridpoint Statistical Interpolation(GSI) Data Assimilation System
Unified variational data assimilation (DA) system• Global and regional applications• Weather and climate
Operational system being used by • NOAA (GFS, NAM, RTMA, HWRF, RAP…)• NASA (GMAO global) • AFWA (Testing in progress)
Distributed development:• NCEP/EMC, NASA/GMAO, NOAA/ESRL, NCAR/MMM, …
A community code used in NOAA operations and research• Supported by NCEP, NASA, DTC• Disciplined code management policy and procedures required to
incorporate changes across user organizations
http://www.dtcenter.org/com-GSI/users/index.php
WoF Modeling Meeting 2/4/13 13
Goals of Community GSI Efforts
Provide current operational GSI capabilities to the research community (O2R)
Provide a framework for distributed development of new capabilities & advances in data assimilation
Provide a pathway for data assimilation research to operations process (R2O)
Provide rational basis to operational centers and research community for enhancement of data assimilation systems
WoF Modeling Meeting 2/4/13
Adding New Analysis Variables to GSI
• No longer a difficult procedureo Thanks to GSI Bundle infrastructure (via NCEP/NASA
collaboration)
• Generalized structure for analysis, guess, state, and control variableso e.g., hydrometeors, vertical velocity, double moment
microphysics variables, etc.
• Control through table in the anavinfo file
• Greatly reduces the number of routines needed to be modified• Modify table(s) + possible additional model I/O steps +
observation ingest + forward operator + background error• Can also disable the conventional large-scale balance
constraints present in the NMC derived background errors
WoF Modeling Meeting 2/4/13 15
North America Model (NAM)• Implemented 18 October 2011• NEMS based NMM• Outer grid at 12 km to 84hr• Multiple Nests Run to ~60hr
– 4 km CONUS nest– 6 km Alaska nest– 3 km HI & PR nests– 1.3km DHS/FireWeather/IMET (to
36hr)
Rapid Refresh (RAP)• Implemented 1 May 2012• WRF-based ARW• Use of GSI analysis• Expanded 13 km Domain to
include Alaska• Experimental 3 km HRRR
RUC-13 CONUS domain
WRF-Rapid Refresh domain – 2010
Original CONUS domain
Experimental 3 km HRRR
NOAA Operational MesoscaleModeling for CONUS:
WoF Modeling Meeting 2/4/13
Numerical Guidance to Support a NWS Operational Warn on Forecast Capability
S. Lord – NWS/OSTand
NCEP/EMC StaffV4.0
WoF Modeling Meeting 2/4/13
Introduction• Goal: field an operational, credible and skillful WoF capability by 2025
• Combines forecast requirements for– NWS severe weather forecasts and warnings– Aviation (NEXTGEN)– Daily general weather guidance to 3 days, including QPF
• Closely allied with Hurricane forecast technology (HFIP)
• Effort exceeds decade-long development and implementation cycle– Current operational system provides look at the future pieces– Incremental system development provides a risk-managed path for operations– With sustained development of the operational pieces, and sufficient HPC, NOAA (OAR and NWS) can
achieve the WoF goal• Development of operational systems provides enhancement and generalized capabilities to incorporate R&D and
improve capabilities (including forecast skill)• Improved skill must be documented with appropriate performance measures (TBD)
• HPC capability growth is essential to meet goal– Adequate resolution for predicting high impact weather events poses largest operational HPC requirements
• Forecast model• Data assimilation• Ensemble & postprocessing system
– Advanced nature of severe weather problem requires larger ratio of R&D HPC to operational HPC– Constant $$ HPC growth and new technology (e.g. GPU, MIC, etc) will be necessary but insufficient
WoF Modeling Meeting 2/4/13
International Thrust – UKMO 2014Model Resolution
(km)Forecast Length
Frequency (Members)
Other details
UK 1.5 36 8 3d var
UK ensemble 2.2 36 4(12) Global perturbed boundaries and init. Conditions
Global 17 14472
22
Hybrid 4-d Var
Global Ensemble
33 72 4(12) ETKF with Stochastic Physics
• UK is a very small country!– Area covers approximately Fire Wx nest (next slide)– Weather events less diverse– Regional high resolution NWP is more affordable computationally
• US global system (planned for 2014) has comparable resolution – SL T1148 (17 km)– Hybrid DA (27 km ensemble) enhances skill to comparability with UKMO
• US regional system lags in operational implementation (due to HPC constraints) but not technology level
WoF Modeling Meeting 2/4/13 19
Global Continental Local Nesting Strategy
• Discussion items– Nesting necessary to save computing; current NAM domain could be expanded to global 10 km– Advantages/disadvantages of same model for global and regional– Applicability (DA) of a multi-model ensemble-based covariance estimate with multiple bias characteristics– Data assimilation at resolution of 3 km or less poses challenges
• Advanced assimilation of radar reflectivity• Consistent integration of radar and satellite info• Proper dynamically balanced analysis increments at multiple high resolution scales (10, 3, 1 km) to minimize
spin-up/spin-down and maximize evolution of nascent disturbances
WoF Modeling Meeting 2/4/13
Numerical Guidance Attributes (1)• Analysis and uncertainty products
– Hourly refresh but sub-hourly for rapidly changing observables (reflectivity, hydrometeors, wind shear)– 1 km resolution for rapidly changing observables– Hybrid analysis provides analysis uncertainty information and ensemble perturbations– Multi-model ensemble-based background (technology not developed)– Analysis variables
• Surface (sensible wx) variables• Clouds• Radar reflectivity, hydrometeors• 3-D atmosphere• Aerosols• Skin temperature• Precipitation amount and hydrometeors• Land & hydrology• Eventual coupled system to add marine & coastal analysis (waves, water level, water quality)
– Products support forecaster Situational Awareness (SA)• MRMS products are an initial capability• Enhancement to Rapidly Updating Analysis (RUA)
– Combines obs. from» Radar» Satellite» In situ
– Expands to hydrological and coastal marine domain– Provides best and consistent geophysical picture from all available information
– Catchup cycle for late data produces “Analysis of Record” for verification
WoF Modeling Meeting 2/4/13
Numerical Guidance Attributes (2)
• Initialized from hourly analysis• Evolved through RAP, HRRR & NAM development
– North American Rapid Refresh Ensemble (NARRE)– High Resolution Rapid Refresh Ensemble (HRRRE)
• 1-way nested– Technology available in NAM as prototype– 1 km resolution where needed– Ensemble-based probabilistic products support
• Severe convective weather (incl. tornado environment)• Aviation and surface transportation• Landfalling hurricanes• Flash floods
– Eventual support for• Storm surge• Waves• Water quality
• High resolution analysis and nested forecast system provide consistency between real-time SA and forecast evolution
WoF Modeling Meeting 2/4/13
Current Ops (NAM-RR) 2025 WoF Ops
Parent Domain North America (NA) Global
Nested domains 4 children (CONUS, AK, HA, PR-H)1 grandchild (Fire Wx)
Same children (4)Grandchildren (27) - Fire Wx (2) - WoF (5) - Fixed airport (20)
Resolution 12-13 km (NA) 4 km CONUS, 6 km AK, 3 km HA, PR-H 1 km Fire Wx (grandchild)
10 km Global3 km children1 km grandchildren
Analysis update 6 hourly (NA)1 hour CONUS
3-6 hour (global)1 hour (child, grandchild)
Ensemble NA: Short-Range Ensemble Forecast (SREF, 21 members, 16 km)
Global: GEFS (30 members, 10 km)Child: 30 members, 3 kmG’child: 20 members, 1 km
Fcst length & frequency Parent: 4X/day to 84 h (NA)Child: 4X/day 60 h (child)G’child; 4X/day to 36 h
Global Parent: 4X/day to 240 h at full resolutionChild: 4X/day to 84 h; 20X/day to 36 hG’child: 4X/day to 36 h; 20X/day to 6 h
Data Assimilation 3-D Var Parent, Child: 4-D Hybrid (Var-EnKF)G’child: 3-D Hybrid (or more, if affordable)
Assimilated variables Primarily atmosphere Atmosphere, cloud, radar reflectivity& hydrometeors, aerosol, skin T, land & hydrology
Rapid Update Analysis (RUA) Hourly Surface (Real-Time Mesoscale Analysis, RTMA) 2-D fields at 2.5 km
5 min, 1 km 3-D RUA for clouds, radar reflectivity, 1 km 2-D quantitative precipitation estimate (QPE) & surface products; hourly 3-D atmosphere at 2.5 km; land & hydrology,
Required computing capability
1 Global: 670 XChild: 1734 XG’child: 2373 X
Warn on Forecast (WoF) System Comparison Summary
WoF Modeling Meeting 2/4/13
Current Ops (NAM-RR) 2025 WoF Ops
Required computing capability
1 Global: 670 XChild: 1734 XG’child: 2373 X
System Evolution
Expected computing Xfactor
2013 2015 2018 2021 2024
2.3 4 5 3 3Cumulative 2.3 9.2 46 138 414
HPC GapsEst. HPC
Req. HPC
Ratio Notes
Global 414 670 1.6 SL can mitigate gap
Child 1734 4.2 1) system too ambitious OR 2) add’l HPC cycles
G’child 2373 5.7
WoF Modeling Meeting 2/4/13 24
Runtime & optimal node apportionment for operationalNMMB nesting with a Fire Wx nest over CONUS (72 nodes)
3 km Puerto Rico nest 3/72 or 4%
1.33 km CONUS FireWx nest 12/72 or 17%
4 km CONUS nest 40/72 or 56%
24
3 km Hawaii nest 4/72 or 5%
12 km parent 8/72 or 11%
6 km Alaska nest 5.72 or 7%
1-way nests solved simultaneouslyWRF-NMM
takes 3.6 times longer than
NEMS-NMMB to run
Fire Wx nest
Increased nest
capability by adding
processors!
WoF Modeling Meeting 2/4/13
Summary• Ensemble-based probabilistic guidance for WoF is a
major challenge– Scientific development– Demonstration of system capabilities
• Current operational capabilities can provide basis for future system– But many changes necessary in response to R&D needs and
results• HPC is second major challenge
– Prototype system exceeds planned availability by factor of 4-6• R&D may suggest compromises and where to spend HPC resources
to achieve required skill affordably
WoF Modeling Meeting 2/4/13
Supplementaries
WoF Modeling Meeting 2/4/13
2015-2016
North American Rapid Refresh ENSEMBLE (NARRE)
• NMMB (from NCEP) & ARW (from ESRL) dynamic cores• Common use of NEMS infrastructure and GSI analysis• Common NAM parent domain at 10-12 km• Initially ~6 member ensemble made up of equal numbers of
NMMB- & ARW-based configurations• Hourly updated with forecasts to 24 hours• NMMB & ARW control data assimilation cycles with 3-6 hr
pre-forecast period (catch-up) with hourly updating• NAM & SREF 84 hr forecasts are extensions of the 00z,
06z, 12z, & 18z runs – for continuity sake.– SREF will be at same 10-12 km resolution as NARRE by then– SREF will have 21 members plus 6 from NARRE for total of 27
• NARRE requires an increase in current HPCC funding
27
DiMego
WoF Modeling Meeting 2/4/13 28
2017-2018High Resolution Rapid Refresh
ENSEMBLE (HRRRE)• Each member of NARRE contains 3 km nests
– CONUS, Alaska, Hawaii & Puerto Rico/Hispaniola nests– The two control runs initialized with radar data & other hi-res obs
• This capability puts NWS/NCEP[+OAR/ESRL] in a position to – Provide NextGen Enroute AND Terminal guidance (FWIS-like)– Provide PROBABILITY guidance with full Probability Density Function
specified, hence uncertainty information too– Provide a vehicle to improve assimilation capabilities using hybrid
(EnKF+4DVar) technique with current & future radar & satellite– Address Warn-on-Forecast as resolutions evolve towards ~1 km
• NAM nests are extensions of the 00z, 06z, 12z & 18Z runs.• HRRRE requires an increase in HPCC funding over
and above that required for the NARRE
DiMego