Upload
betha
View
38
Download
0
Tags:
Embed Size (px)
DESCRIPTION
NCEP Update: Opportunities for Collaboration. “Where America’s Climate, Weather and Ocean Prediction Services Begin”. Louis W. Uccellini Director, NCEP Visit to Central Region HQ Kansas City, MO October 24, 2007. Overview. Background Recent Model Implementations Collaboration - PowerPoint PPT Presentation
Citation preview
NCEP Update:
Opportunities for Collaboration
“Where America’s Climate, Weather and Ocean Prediction Services Begin”
Louis W. UccelliniDirector, NCEP
Visit to Central Region HQKansas City, MO
October 24, 2007
2
Overview
• Background
• Recent Model Implementations
• Collaboration– Model requirements – Annual Operating Plan– Model Evaluation – new NCO process– Service Center Collaboration Efforts
• Space Weather Prediction Center
• Summary
3
Background
• Organization• Test Beds
4
5
Service – Science Linkagewith the Outside Community
• EMC WRF Developmental Test Center, Joint Center for Satellite Data Assimilation
• CPC Climate Test Bed
• TPC Joint Hurricane Test Bed
• HPC Hydrometeorological Test Bed
• SPC Hazardous Weather Test Bed with NSSL
• SWPC Solar Test Bed
• AWC FAA Aviation Test Bed with NCAR RAP
• OPC linked with EMC’s Marine Modeling and Analysis Branch
6
Recent Model Implementations
• Model Production Suite• 2007 Implementations• Ensemble Systems
7
Air Quality
WRF NMM/ARWWorkstation WRF
WRF: ARW, NMMETA, RSM GFS, Canadian Global Model
Satellites99.9%
Regional NAMWRF NMM
North American Ensemble Forecast System
Hurricane GFDLHWRF
GlobalForecastSystem
Dispersion
ARL/HYSPLIT
Forecast
Severe Weather
Rapid Updatefor Aviation
ClimateCFS
1.7B Obs/Day
Short-RangeEnsemble Forecast
NOAA’s NWS Model Production Suite
MOM3
NOAH Land Surface Model
Coupled
Global DataAssimilation
OceansHYCOM
WaveWatch III
NAM/CMAQ
8
2007 Implementations• Global Forecast System (GFS)
– Hybrid Sigma-Pressure GFS – 3rd Qtr FY2007
– Gridpoint Statistical Interpolation (GSI) Analysis – 3rd Qtr FY2007
• NAM– Low level model moisture improvement - 3rd Q FY2007
• Real-Time Mesoscale Analysis – 4th Qtr FY2007– RTMA upgrade
• NAEFS– Increase U.S. Membership from 15 20 Members – 2rd
Qtr FY2007 • U.S. – 80 members, Canadian – 20 members (inc. to 40 in
July’07), FNMOC – 20 members
9
2007 Implementations (cont.)• HiRes Window
– Higher resolution and expanded domains – 4th Qtr FY2007
• Air Quality Forecast– Expanded operational domain to
CONUS – 4th Qtr FY2007• Ocean Modeling
– Multi-scale Wave Model – 4th Qtr FY2007 (including Great Lakes)
– HYCOM data assimilation upgrade – 3rd Qtr FY2007
• Hurricane Modeling – 3rd Qtr FY2007– Hurricane WRF in parallel
operations– Probabilistic Storm Surge
Ocean Heat Content - Katrina
10
Wave Watch III for Great Lakes
• NCEP’s Great Lakes Wave model became operational in August 2006, and consists of:– A single grid with a resolution of 0.035x0.05
(4km) covering all of the Great Lakes– Wind forcing is taken from the 1/12 NAM model,
as is the dynamically updated ice coverage• 6h hindcast for continuity between model runs
• 84h forecast based on availability of hourly NAM winds
• Four model cycles per day
11
In Development• September, 2007, the ocean wave models converted to a
new mosaic approach to wave modeling, with multiple grids with different resolutions working as a single integrated model.
• The GLW model is slated to convert to the new wave model software in FY2008-Q2.– For now still with a single grid.
– Exploring downscaling techniques for NAM winds.
– Intend to start parallel runs with a second GLW model using NDFD winds
• Parallel for FY2008-Q2, no tentative operational date yet.
12
North American Ensemble Forecast System
• NAEFS basic product list– 11 functionalities
• Ensemble mean, spread, probabilities, etc.
– 50 variables• U,v,t,z,CAPE,
precip type, etc.– 7 domains
• Global, NH, NA, CONUS, SA, Caribbean, Africa• Over 600 products requested by users (will be supplied via priority order)
– Graphics• NAWIPS (interactively available)• NCEP Model Web Page – beginning July 07 on experimental basis
– Grids• NAWIPS• ftp site (GRIB2)• NDGD in planning phase (1stQ 08)
• Still working availability issues - first priority: “intermediate” users
Canada contributes 40 runs/day to 16 days, US contributes 80/day to 16 days
13
• 21 members 4 times per day; 87 hours from 9 and 21Z
• WRF/NMM, WRF/ARW, Eta, RSM; Resolution 32-45 km and 60 levels
• Added six members twice per day (WRF ARW & WRF NMM) (Dec ‘05)
• Run SREF 4 times per day (03, 09, 15 and 21 UTC) (Jun ‘06)
• Implement Grid Based Bias Correction (Q1, FY2008)
• Add WRF BUFR Files (Q1, FY2008)
• Implement ensemble mean BUFR files (Q4, FY2008)
• All members at 32 km resolution (Q4, FY2008)
Short Range Ensemble Forecast
SREF Upgrades
14
Winter Weather Impact Graphics from SREF
• Probability event will last more than 12, 24, 48 hrs • Probability visibility will be reduced to 1/2, 1/4, or 1/8
mile in winter precip• Probability winter precip rate will be more than 1, 2, or
3" per hr • Probability road sensor will detect winter precip
(relative to normal) “snow on road”• Probability Blizzard Criteria will be met • Probability Freezing Rain .01" or more will accumulate
on any surface • Probability NWS Winter Storm Warning criteria will be
met (under construction)
http://www.hpc.ncep.noaa.gov/wwd/impactgraphics/
15
High Impact SREFGuidance at the SPC
http://www.spc.noaa.gov/exper/sref• Hundreds of images
available online• Overview fields • Specialized high-
impact guidance– Fire weather– Winter weather– Convection
• Post-processed calibrated guidance
– Probability of thunderstorms
– Probability of severe storms
– Probability of dry lightning
Expandable Menus:
16
Collaboration Opportunities
• Model Requirements• New Model Assessment Procedures• Service Centers
17
Model RequirementsAnnual Operating Plan
• EMC Meeting held each December• All SSD Chiefs attend
– Assessments – Prioritized requirements (collected from regions)– Issues/process modification
• What were major items from last year?– Field evaluation of the WRF-NAM – Formation of NWP Team to enhance products and services (NCEP,
Regions, AWIPS, OAR?)• Transmission of grids via NAWIPS • Increased products (e.g. ecosystems, forecasts)• Probabilistic guidance
• Opportunities for technical involvement/training – Ensemble products
• Rich Grumm (State College)• Mike Evans (Binghamton)• Steve Zubrick (Baltimore/Washington)• David Bright (SPC)
18
New Model Assessment Procedures
• Expanded participation– NCEP Centers– NWS Regions– University Community– Private Sector
• Mailing list for notifications of upcoming major changeshttps://lstsrv.ncep.noaa.gov/mailman/listinfo/ncep.list.modelevalinfo
• Notification is provided to mailing list of upcoming evaluation periods including due dates
• Parallel data sets are made available to those who agree to participate
• Completed evaluations submitted by due date are used as input to decision on implementation
19
Increasing Collaboration Within NCEP Service Centers and the NWS
• SPC/WFO– Watch “by county”
• HPC/WFO– Winter Weather Desks– Medium Range: Days 4-7– Hurricane “Hot Line”
• AWC/WFO/CWSU– Collaborative Convective Forecast
Product• CPC/ Regions/WFO/RFC
– Hazard Assessment– Seasonal outlooks
• OPC/WFO– Near shore High seas
• HPC-TPC-OPC-AR-PR– Unified Surface Analysis!
• SEC – AWC– Solar/Aviation Products
20
Space Weather Prediction Center
Space Weather in your Neighborhood
21
Space Weather Prediction Center• Space Environment Center formally joined NCEP/NWS/NOAA on
January 9, 2005 from OAR/NOAA– Their addition helps foster a seamless suite of operational products from the “Sun
to the Sea”
• Provides space weather alerts and warnings for disturbances that can affect people and equipment working in space and on Earth
• Prediction /Monitoring products include:– Onset of solar storms and magnitude of
resultant geomagnetic disturbances with impacts on GPS, aircraft, communications, satellite systems, utilities, aurora,…
• Next Solar Max – 2011!
22
• 268,762 RSF; includes 800+ Feds, contractors and visiting scientists • 5 NCEP Centers• NESDIS research and satellite services• OAR Air Resources Laboratory
• 40 spaces for visiting scientists
NOAA Center for Weather and Climate Predictionat the UMD Research Park (M-Square)
Date
Construction Start May 9, 2007
Move Start Dec 2008
Move Complete July 2009
Construction Schedule
23
NOAA Center for Weather and Climate Prediction
Recent Activities
Construction photographs – October 2, 2007
24
NOAA Center for Weather and
Climate Prediction
NORTH
First New UMDBuilding
25
Summary
• NCEP spanning “Sun to the Sea”; many new programmatic areas (oceans, air quality, space weather…)
• Many science-service opportunities/challenges exist• NCEP is uniquely positioned to handle transition issues
from research to operations – Test beds• NCEP is actively pursuing opportunities for collaboration
in research, transition to operations and operational production and delivery of services
• Still consider NCEP to be an underutilized entity by the research and operational communities
26
Appendix
27Five Order of Magnitude Increase in Satellite Data Over Next Ten YearsFive Order of Magnitude Increase in Satellite Data Over Next Ten YearsFive Order of Magnitude Increase in Satellite Data Over Next Ten YearsFive Order of Magnitude Increase in Satellite Data Over Next Ten Years
Count
(Mill
ions)
Daily Satellite & Radar Observation Count
20001990 2010 2010-10%of obs
2002 100 M obs
2003-4 125 M obs
Level 2 radar data 2 B
2005 210 M obs
Joint Center for Satellite Data Assimilation
2007 1.1B obs
28
JCSDA Mission and Vision
• Mission: Accelerate and improve the quantitative use of research and operational satellite data in weather and climate analysis and prediction models
• Near-term Vision: A weather and climate analysis and prediction community empowered to effectively assimilate increasing amounts of advanced satellite observations
• Long-term Vision: An environmental analysis and prediction community empowered to effectively use the integrated observations of the GEOSS – and be ready for NPOESS at “Day 1” after launch
29
The Joint Center for Satellite Data Assimilation
• Formed in 2001(NASA, NOAA, DoD)
• Infrastructure for real-time access to operational and research satellite data from GOES, AMSU, Quikscat, AIRS, MODIS,…
• Community fast forward radiative transfer scheme … operational data assimilation and model forecast systems available to research and forecast communities
• Supports “internal” and “external research” and data assessments on NOAA/NCEP computers
The Research Community is now using the operational infrastructure.
The Operational Community is now accelerating use of satellite data.
30
JCSDA Major Accomplishments• Common assimilation infrastructure at NOAA and NASA • Community radiative transfer model V2 released• Common NOAA/NASA land data assimilation system• Interfaces between JCSDA models and external researchers• Operational Implementations Include:• Snow/sea ice emissivity model – permits 300% increase in sounding data
usage over high latitudes – improved forecasts• MODIS winds, polar regions, - improved forecasts• AIRS radiances – improved forecasts• New generation, physically based SST analysis - Improved SST• Preparation for advanced satellite data such as METOP (IASI/AMSU/MHS),
DMSP (SSMIS), COSMIC GPS data, EOS AMSR-E, GOES-R • Impact studies of POES MHS, EOS AIRS/MODIS, Windsat, DMSP
SSMIS……. on NWP through EMC parallel experiments
31
How HPC adds value to Day 3 - 7
Adjusting approximately 35% of the points
33
How HPC adds value to Day 3 - 7
Adjusting approximately 35% of the points
34
The NCEP coupled Climate Forecast System (implemented August 24, 2004)
1. Atmospheric component• Global Forecast System 2003 (GFS03)
• T62 in horizontal; 64 layers in vertical
• Recent upgrades in model physics Solar radiation (Hou, 1996)
cumulus convection (Hong and Pan, 1998)
gravity wave drag (Kim and Arakawa, 1995)
cloud water/ice (Zhao and Carr,1997)
2. Oceanic component• GFDL MOM3 (Pacanowski and Griffies, 1998)
• 1/3°x1° in tropics; 1°x1° in extratropics; 40 layers
• Quasi-global domain (74°S to 64°N)
• Free surface3. Coupled model
• Once-a-day coupling
• Sea ice extent taken as observed climatology
4. Calibrated on past 38 years
35
Climate Forecast System Availability
7 day average centered on Dec 27
• Real-time 2x daily, 9-month forecasts, monthly ensemble of 40-60 members.
• 15-member reforecasts per month (1981–2005)– Calibration– Skill estimates– Analog and statistical forecasts
• The website for real time data retrieval is at: ftp://tgftp.nws.noaa.gov/SL.us008001/ST.opnl/MT.cfs_MR.fcst
• The climatological data is at: ftp://tgftp.nws.noaa.gov/SL.us008001/ST.opnl/MT.cfs_MR.clim/
• Complete documentation available at: http://www.emc.ncep.noaa.gov/gmb/ssaha/cfs_data/cfs_data.pdf
36
Next Generation CFS• Coupled atmosphere-ocean-land surface-cryosphere system
– Improvements to • ODA (MOM3 MOM4)• “Unified Physics”• Atmospheric model
– Sigma-pressure hybrid model– Upgrades to microphysics, radiation,…
• Atmospheric data assimilation (GSI)– ESMF-based coupling and model structure
• Reanalysis (1979-present)• Reforecast for
– Weather & Week2 (1-14 days)– Monthly (2 weeks to 2 months)– S/I (2-12 months)
• Estimated completion January 2010
37
38
Without skill mask
CFS SeasonalTemp Forecast
(deg K)
39
With skill mask
• If anomaly correlation between forecast and observed conditions over the 1982-2003 period is below 0.3, values are not shown
CFS SeasonalTemp Forecast
(deg K)
40
41
U.S. Seasonal Temperature - Skill0.5 Month Lead – 4 Year Running Average vs. GPRA Goal
• Climate Forecast System: First dynamic operational climate forecast model implemented August 2004• Climate Test Bed: established in 2005and focused on improving the Climate Forecast System and related seasonal forecast products• Unprecedented increases in the skill of CPC official seasonal outlooks: (20% or more; O’Lenic et al. 2007) due in part to CFS and to CTB activities (e.g. consolidation tool)
Proposed Time Line for CFS “Next”• April 2007: Implementation of the GSI scheme for GFS.
• Jun 2007: Testing ESMF coupler for GFS and MOM4.
• Aug 2007: Pilot studies for fully coupled reanalysis (GFS, GODAS & GLDAS)
• Jan 2008: Production and Evaluation of CFS Reanalysis for 1979 to present
• Jul 2008: Prepare CFS Retrospective Forecasts (2 initial months: October and April)
• Jan 2009: Complete CFS Retrospective Forecasts (remaining 10 months)
• Nov 2009: Compute calibration statistics for CFS daily, monthly and seasonal forecasts.
Prepare CFS Reanalysis & Retrospective Fcst data for public dissemination.
• Jan 2010: Operational implementation of the next CFS monthly & seasonal forecast suite.
************************
* Reanalysis, reforecast funding for NCEP provided in FY07 Pres. Budget ($800 K); Funding is critical to move beyond pilot studies
43
HPC
• Performance Metrics
• QPF Status
• Winter Weather Desk
• NDFD – Days 3-7
44
Impact of Models on Day 1 Precipitation Scores
0
0.05
0.1
0.15
0.2
0.25
0.3
0.3519
91
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
Th
reat
Sco
re
Human(HPC)
ETA
Linear(Human(HPC))Linear (ETA)
HPC Forecasters Add Value
Models provide basis for improvement Correlations
Of HPC with:
Eta: 0.99GFS: 0.74NGM: 0.85
(DOC GPRA goal)
45
46
In last 10 years, HPC forecasts have added 2+ days of skill
Day 7
Day 5
Day 3
48
Status of QPF Procedures
• Collaborative Procedure– WFOs notified when QPF is ready and when
changes are substantial from previous forecast– WFOs contact HPC for collaboration via 12Planet
and telephone
• Link to RFCs– HPC sends fcst files to RFCs electronically and are
available for collaboration via 12Planet and phone
• Awaiting final take on QPF White Paper
49
NWS Winter Weather Desk
Goals of 4 year experiment from 2001- 2004: Improve Winter Weather Services to the public
through coordination of the winter weather watches/warnings with National guidance products
Test short range ensemble for their applications to winter weather forecasting
Motivation: Jan 24-25, 2000; December 30, 2000: March 4-6, 2001
WWD “operational” September 15, 2004
50
NWS Winter Weather Desk
Time line: Sep 15 – May 15 Participants
NCEP HPC Provide SREF based Winter Wx guidance Collaborate with WFOs (Chat Room Technology)
WFOs All CONUS WFOs Use guidance from NCEP to produce coordinated Winter Storm Watches/Warnings
Products: http://www.hpc.ncep.noaa.gov/wwd/winter_wx.shtml 24 h probability (low, moderate, high) of meeting/exceeding 4”, 8”, 12”
snow, 0.25” freezing rain (for day 1, 2, 3) 72h Low tracks graphic and discussion
Regional Stats
* Oct - Mar
WR WWE2 02-03’
WWE303-04’
WWD04-05’
WWD05-06’
WWD06-07’
# WFO NA 10 ALL ALL ALL
POD NA .88 .88 .86 .79
FAR NA .27 .30 .36 .43
CSI NA .67 .64 .58 .49
LT Warn NA 14 16 16 15
SR WWE2 02-03’
WWE303-04’
WWD 04-05’
WWD*05-06’
WWD*06-07’
# WFO NA 11 ALL ALL ALL
POD NA .92 .90 .85 .93
FAR NA .38 .39 .48 .33
CSI NA .59 .57 .48 .64
LT Warn NA 9 9 11 19
ER WWE202-03’
WWE303-04’
WWD*04-05’
WWD*05-06’
WWD*06-07’
# WFO ALL ALL ALL ALL ALL
POD .90 .92 .92 .91 .94
FAR .30 .32 .30 .35 .28
CSI .65 M .66 .61 .69
LT Warn 15 18 21 19 22
CR WWE202-03’
WWE303-04’
WWD04-05’
WWD05-06’
WWD06-07’
# WFO 8 33 ALL ALL ALL
POD .90 .88 .92 .91 .93
FAR .40 .45 .32 .38 .26
CSI .57 .51 .65 .53 .70
LT Warn 13 13 17 17 18
06-07 Winter very dry in West
Factors influencing stats: collaborative WWD activity, better use of SREF, application of CSTAR and other R2O, active sharing of best practices, post storm service and science assessments
52
• Routine collaboration using 12Planet or by telephone (conference call)
• Experimental 2006-07 GoToMeeting PC/phone collaboration with subset of Western Region offices
2006-2007 Collaboration (right)
Green – 12 Planet only
Orange – Telephone Only
Red – Both 12Planet/Phone
Light Blue Outline – GoToMeeting
2005-2006 Collaboration (left)
HPC Winter Weather Desk Collaboration
53
HPC/NDFD Highlights• For decades HPC provided day 3 - 7 guidance
– Max Temp, Min Temp, 24 hr (later 12 hr) Pop, surface progs• Spring 2004 - HPC asked to provide forecasts for additional fields
including dew point, sky cover, wind direction and speed, and precipitation type at 5 km resolution
• HPC began providing these fields in June 2004• WFOs get HPC CONUS grids and display on AWIPS (with AWIPS OB 5)• National perspective available since October 2006 on HPC web site
54
Methodology
• HPC has 2 forecasters/shift preparing day 4 – 7 forecasts (one shift/day)– One prepares surface progs and writes the narrative– Second prepares the max/min temperatures and PoPs
• HPC uses N-AWIPS to generate forecasts– Forecasters can edit 384 stations; – Can edit individual stations or groups of stations– Can use a “model blender” to interactively generate forecasts
from various models with user selectable weighting• No new staff
– Algorithms to generate additional fields: dewpoint, cloud cover, winds, precip type
– Convert point forecasts to 5 km grids
55
How HPC adds Value to Day 3 – 7Minimum Temperature
56
How HPC adds Value to Day 3 – 7PoP
• HPC makes many more small changes to MOS than large changes, but percent
improvement is greater for large changes.
57
Issues
• Time of release of forecasts
– Some WFOs want HPC to issued grids earlier
• HPC starts with MOS rather than previous forecast
– Field prefers less “flip-flopping”
• How involved do forecasters need to be?
– Are post-processed grids good enough?
• How should HPC provide probabilistic info?
– QPF – exceedance values?
– Day 4 – 7 – ranges in addition to “best guess”?
• How is collaborative approach sustained?