Transcript Expansions to NOAA’s National Air Quality Forecast Capability Progress in 2008 October 7, 2008 Paula Davidson*, Rohit Mathur, Jeff McQueen, Scott Jackson, Ivanka Stajner, Ken.

Expansions to NOAA’s National Air
Quality Forecast Capability
Progress in 2008
October 7, 2008
Paula Davidson*, Rohit Mathur, Jeff McQueen, Scott
Jackson, Ivanka Stajner, Ken Carey
*NWS Manager, Air Quality Forecast Capability
1
Outline
Background on NAQFC
Progress in 2008
- Operational products:
- Experimental products
- Developmental testing
Coordination with Partners
Looking Ahead
2
National Air Quality Forecast Capability
Current and Planned Capabilities, 10/08
•
•
Improving the basis for AQ alerts
Providing AQ information for people at risk
FY08 Prediction Capabilities:
•
•
•
Operations:
Ozone, expanded from EUS to CONUS, 9/07
Smoke implemented over CONUS, 3/07
Experimental testing:
Ozone upgrades
Smoke predictions over AK
Developmental testing:
components for particulate matter (PM) forecasts
2007: O3,smoke
2005: O3
Near-term Operational Targets:
•
Ozone, smoke coverage extended Nationwide
Longer range:
•
•
•
Quantitative PM2.5 prediction
Extend air quality forecast range to 48-72 hours
Include broader range of significant pollutants
3
National Air Quality Forecast Capability
End-to-End Operational Capability
Model Components: Linked numerical
prediction system
Operationally integrated on NCEP’s supercomputer
•
NAM mesoscale NWP: WRF-NMM
•
CMAQ for AQ; HYSPLIT for smoke
Observational Input:
•
•
NWS weather observations; NESDIS fire locations
EPA emissions inventory
Gridded forecast guidance products
•
•
•
On NWS servers: www.weather.gov/aq and ftp-servers
On EPA servers
Updated 2x daily
Verification basis, near-real time:
•
•
AQI: Peak Oct
4
Ground-level AIRNow observations
Satellite smoke observations
Customer outreach/feedback
•
•
State & Local AQ forecasters coordinated with EPA
Public and Private Sector AQ constituents
4
Progress in 2008
Operational Products:
–
Ozone: Coast-to-Coast (CONUS) guidance implemented 9/07; 2008 updates for
emissions, WRF-NMM
–
Smoke: CONUS guidance implemented 3/07; 12/07 upgrade to full vertical resolution
Experimental Products:
–
Ozone: CB-05 chem mechanism, developing prototypes for AK, HI
–
Smoke: Expanded coverage to AK, 6/08
Developmental Products:
–
Aerosols: Developmental testing providing comprehensive dataset for diagnostic
evaluations. (CONUS)
•
CMAQ (aerosol option), testing CB05 chemical mechanism
–
Prototypes for AK, HI (ozone); HI (smoke)
–
Dust and smoke inputs: testing dust contributions to PM2.5 from global sources
–
•
Preliminary tests combining dust with CMAQ-aerosol
•
Case studies combining smoke inputs with CMAQ-aerosol
R&D efforts continuing in chemical data assimilation, real-time emissions sources,
advanced chemical mechanisms
5
Verification Statistics: Example
Max 8-hr O3 081507
Fraction
Correct: 0.92
CONUS
N
Obs
Mean
Model
Mean
RMSE
(ppb)
NME
(%)
MB
(ppb)
1094
56.62
62.29
14.54
19.73
5.66
Eder et al. 2008
NMB
(%)
10.00
r
0.75
Max 8-hr O3 081507
6
Progress from 2005 to 2007:
Ozone Prediction Summary Verification
IOC Domain: Daily Verification: 8-hr Avg NEUS Domain
1.00
2005
0.90
0.80
Initial Operational Capability (IOC)
Fraction
Correct
Hit Accuracy
0.70
Target
NEUS
0.60
Operational
Operational, NE US Domain
0.50
Day 1-Jun
8-Jun
15-Jun
22-Jun
29-Jun
6-Jul
13-Jul
20-Jul
27-Jul
Experimental Domain: Daily Verification: 8-hr Avg Full EUS Domain 2005
2005
1.00
0.95
Experimental
Experimental, Eastern US
0.90
0.80
1-Jun
1-Jul
Approved 8/05 to replace IOC
(NE US) in operations
Fraction Correct
EUS
0.85
Target
1-Aug
1-Sep
1-Oct
Fraction Correct, 2006: 3X 8-hr avg
0.976
0.994
0.976
0.985
0.996
1
2006
0.95
Operational
Operational, Eastern US
0.9
Fraction Correct
Target
EUS
0.85
0.8
5/1/2006
Day
5/31/2006
6/30/2006
Monthly Cum
7/30/2006
8/29/2006
9/28/2006
Fraction Correct, 2007: 5X 8-hr avg for CONUS
0.964
0.981
2007
1
Experimental
0.998
0.95
0.985
Experimental, Contiguous US
0.976
0.9
Approved 9/07 to replace Eastern
US config in operations
Fraction Correct
0.85
CONUS
Target
Monthly Cum
0.8
5/1/07
5/15/07
5/29/07
6/12/07
6/26/07
7/10/07
7/24/07
8/7/07
8/21/07
9/4/07
9/18/07
7
Progress from 2007 to 2008:
CONUS O3 Prediction Summary Verification
Fraction Correct, 2007: 5X 8-hr avg for CONUS
2007
0.964
0.981
1
Experimental
0.998
0.95
0.985
0.976
Contiguous US (CONUS)
Implemented 9/07 to
replace Eastern US config
in operations JJA: 0.974
0.9
CONUS
Fraction Correct
0.85
Target
Monthly Cum
0.8
5/1/07
5/15/07
5/29/07
6/12/07
6/26/07
7/10/07
7/24/07
8/7/07
8/21/07
9/4/07
9/18/07
OPNL Predictions Fraction Correct, from 4/08:
5X 8-hr avg CONUS 85 ppb THRESHOLD
1
0.987
0.979
0.999
2008
Operational
0.975
0.997
0.95
Fraction Correct 85ppb
0.9
Target
0.85
0.8
4/1/08
CONUS, wrt 85ppb
Threshold
JJA: 0.980
CONUS
Monthly Cum 85-Threshold
4/16/08
5/1/08
5/16/08
5/31/08
6/15/08
6/30/08
7/15/08
7/30/08
8/14/08
8/29/08
8
Prediction Accuracy and Ozone Thresholds
Fraction Correct, from 4/08:
5X 8-hr avg CONUS 85 and 76 ppb THRESHOLDS
1
0.987
0.979
0.999
0.95
0.975
0.997
0.987
0.960
0.947
0.989
0.932
0.9
Fraction Correct 85ppb
Fraction Correct 75ppb
0.85
Monthly Cum 75-Threshold
Monthly Cum 85-Threshold
Target
0.8
4/1/2008
4/16/2008
5/1/2008
5/16/2008
5/31/2008
6/15/2008
6/30/2008
7/15/2008
7/30/2008
8/14/2008
8/29/2008
Effect on FC: moderate reduction
9
Smoke Predictions Summary:
Warm Season, 2007 and 2008
Threat Scores, 2007: Smoke Column > 1 microgam/m3
2007
0.5
0.4
Fraction Correct, 2007: Smoke Column > 1 microgam/m3
Threat Score
1
Target
Monthly Cum
0.9
0.848
0.783
0.755
0.280
0.177
0.3
0.8
0.7
0.150
0.099
0.2
0.780
0.798
0.6
0.5
0.1
0.4
0
0.082
Fraction Correct
FC Target: O3 Predictions
Monthly Cum
0.3
-0.1
0.2
4/1/07
4/15/07
4/29/07
5/13/07
5/27/07
6/10/07
6/24/07
7/8/07
7/22/07
8/5/07
8/19/07
9/2/07
0.1
4/1/07
4/15/07
4/29/07
5/13/07
5/27/07
6/10/07
6/24/07
7/8/07
7/22/07
8/5/07
8/19/07
9/2/07
• FC generally > 0.7
• TS cum = 0.159; Target = 0.08 (Column verification)
• FC Target not established for qualitative smoke tool
• Based on satellite AOD; column verification only
Threat Scores, 2008: Smoke Column > 1 microgram/m3
Season:
0.133
Fraction Correct, 2008: Smoke Column > 1 microgram/m3
1
0.5
Threat Score
Target
Monthly Cum
0.4
0.3
0.9
0.157
0.135
0.658
0.7
0.143
0.770
0.6
0.143
0.2
0.8
0.694
0.716
0.5
0.4
0.1
0.3
0
0.2
0.106
0.1
-0.1
4/1/08
0.643
Fraction Correct
FC Target: O3 Predictions
Monthly Cum
5/1/08
5/31/08
• TS cum = 0.133
6/30/08
7/30/08
8/29/08
4/1/08
5/1/08
5/31/08
6/30/08
7/30/08
8/29/08
• FC generally > 0.6
10
Real-time Testing, Summer 2008:
Experimental Testing
Experimental Predictions
Experimental
Operational
Publicly available, real-time
Ozone:
 CMAQ with advanced gas-phase
chemical mechanism CB05
– more comprehensive Volatile Organic
Compound (VOC) reactions
weather.gov/aq-expr
weather.gov/aq
– challenge: more O3 with CB05
– regional implications: CA, NE US
Smoke:
 Testing over AK domain
– new GOES-W smoke verification in
development
– challenge: little fire activity in 2008
11
Real-time Testing, Summer 2008:
Experimental vs. Operational O3 at 85 ppb
Experimental
Fraction Correct, Ozone Predictions, 1200 UTC
Daily Maximum of 8-h avg, Full 5X Domain, Th=85 ppb
1.00
0.90
0.80
0.70
Fraction Correct
0.60
0.50
14-Jun
Target
21-Jun
28-Jun
5-Jul
12-Jul
19-Jul
26-Jul
2-Aug
9-Aug
16-Aug
OPNL Predictions Fraction Correct, from 4/08:
5X 8-hr avg CONUS 85 ppb THRESHOLD
23-Aug
30-Aug
6-Sep
Operational
1
0.987
0.979
0.999
0.975
0.997
0.95
Fraction Correct 85ppb
0.9
Monthly Cum 85-Threshold
Target
0.85
0.8
4/1/08
4/16/08
5/1/08
5/16/08
5/31/08
6/15/08
6/30/08
7/15/08
7/30/08
8/14/08
8/29/08
Experimental vs Operational, 85ppb:
FC decreases in exptl predictions
12
Real-time Testing, Summer 2008:
Experimental vs Operational O3 at 76 ppb
Experimental
Fraction Correct, Ozone Predictions, 1200 UTC
Daily Maximum of 8-h avg, Full 5X Domain, Th=76 ppb
1.00
0.90
0.80
0.70
0.60
0.50
14-Jun
Fraction Correct
Target
21-Jun
28-Jun
5-Jul
12-Jul
19-Jul
26-Jul
2-Aug
9-Aug
16-Aug
23-Aug
30-Aug
6-Sep
Operational
OPNL Predictions Fraction Correct, from 4/08:
5X 8-hr avg CONUS 76 ppb THRESHOLD
1
0.960
0.95
0.987
0.947
0.989
0.932
0.9
Fraction Correct 75ppb
Monthly Cum 75-Threshold
0.85
Target
0.8
4/1/08
4/16/08
5/1/08
5/16/08
5/31/08
6/15/08
6/30/08
7/15/08
7/30/08
8/14/08
8/29/08
Experimental vs. Operational, 76 ppb:
FC decreases in exptl predictions
13
Developmental Testing, Summer 2008
Developmental Predictions:
Focus group access only, real-time as
resources permit
fine particles
PM2.5
Real-time Testing, Aerosols from NEI
sources:
 CMAQ, Gas-phase CB05 June 10
 aerosol chemical reactions (AERO4) with heterogeneous pathways
 sea salt emissions and chemistry
included in aerosol module
CB-05
Expanded domains:
 HI smoke
 Exploring HI, AK ozone
Developmental
14
Developmental Aerosol Predictions:
Summary Verification, 2008
Fraction Correct, Aerosol Predictions, 0600 UTC
Emissions
Correction
Daily Maximum of 1-h avg, Full 5X Domain, Th=40 mg/m
3
1.00
0.90
0.80
0.70
Fraction Correct
0.60
Target
0.50
7Jan
14Jan
21Jan
28Jan
August 16, 2008
4- 11- 18- 25- 3Feb Feb Feb Feb Mar
10- 17Mar Mar
24- 31Mar Mar
7Apr
14Apr
21Apr
228- 5- 12- 19- 26Apr May May May May Jun
9Jun
16Jun
23Jun
30Jun
7Jul
14Jul
21Jul
28Jul
11- 18- 25- 14Aug Aug Aug Aug Sep
August 17, 2007
15
Aerosol Summary: 2008
Fraction Correct, Aerosol Predictions, 0600 UTC
Lower threshold
Daily Maximum of 1-h avg, Full 5X Domain, Th=35 mg/m
3
1.00
0.90
0.80
0.70
Fraction Correct
0.60
Target
0.50
7Jan
14Jan
21Jan
28Jan
4- 11- 18- 25- 3- 10- 17- 24- 31- 7Feb Feb Feb Feb Mar Mar Mar Mar Mar Apr
14Apr
21Apr
28- 5- 12- 19- 26- 2Apr May May May May Jun
9Jun
16Jun
23Jun
30Jun
7Jul
14Jul
21Jul
28- 4- 11- 18- 25- 1Jul Aug Aug Aug Aug Sep
9Jun
16Jun
23Jun
30Jun
7Jul
14Jul
21Jul
28- 4- 11- 18- 25- 1Jul Aug Aug Aug Aug Sep
Fraction Correct, Aerosol Predictions, 0600 UTC
Daily Maximum of 1-h avg, Full 5X Domain, Th=40 mg/m
3
1.00
0.90
0.80
0.70
Fraction Correct
0.60
Target
0.50
7Jan
14Jan
21Jan
28Jan
4- 11- 18- 25- 3- 10- 17- 24- 31- 7Feb Feb Feb Feb Mar Mar Mar Mar Mar Apr
14Apr
21Apr
28- 5- 12- 19- 26- 2Apr May May May May Jun
Fraction Correct
Fraction Correct, Aerosol Predictions, 0600 UTC
Daily Maximum of 24-h Running Average, Full 5X Domain, Th=40 mg/m
Target
3
1.00
0.90
0.80
0.70
0.60
0.50
7Jan
14Jan
21Jan
28Jan
4- 11- 18- 25- 3- 10- 17- 24- 31- 7Feb Feb Feb Feb Mar Mar Mar Mar Mar Apr
14Apr
21Apr
28- 5- 12- 19- 26- 2Apr May May May May Jun
9Jun
16Jun
23Jun
30Jun
7Jul
14Jul
21Jul
28- 4- 11- 18- 25- 1Jul Aug Aug Aug Aug Sep
16
Partnering with AQ Forecasters
Focus group of state and local AQ
forecasters:
• Participate in real-time developmental testing of
new capabilities, e.g. aerosol predictions
• Provide feedback on reliability, utility of test
products
• Emphasize local episodes/case studies
• Meet regularly to examine test predictions,
discuss feedback, potential improvements
• Work together with EPA’s AIRNow & NOAA
Forecaster Coordination:
• WFO and NCEP/HPC forecasters provide
weather information for partner AQ forecasters
– Web-site for AQ forecasters, interactive discussion on
event-driven basis
17
PHL O3 and aerosol predictions -- Ryan, 2008
Experimental Ozone Guidance, 6/17 -9/10, 1200 UTC
Observed Mean: 64.5 ppbv
Operational Bias: +4.8 ppbv
Experimental Bias: +8.3 ppbv
Increased over-prediction consistent day-to-day as
weather conditions changed.
50
Developmental aerosol guidance,
7/12 -9/10, 1200 UTC
Daily 24-hr max
PM2.5 24-h Average (microg/m3)
45
40
35
30
OBS?
EXPR
25
20
15
10
5
0
12
7/
19
7/
26
7/
2
8/
9
8/
16
8/
Date (2008)
23
8/
30
8/
6
9/
National Air Quality Forecast Capability
Looking Ahead
Nationwide ozone and particulate
matter predictions
• Expanding ozone & smoke to
nationwide coverage, Target: FY10
and
• Begin quantitative particulate
matter predictions, Target: FY14
•Providing information Nationwide on when/where poor AQ is expected
•Reducing losses to life (50,000) each year from poor AQ
•Reducing economic losses ($150B each year) from poor AQ
19
National AQF Capability:
Next Steps
Developing Particulate matter components:
• Smoke from large fires: experimental testing in AK, HI
• Components for quantitative PM forecast capability:
– Objective satellite products for verification (ongoing)
– Aerosols predictions from anthropogenic source emissions in inventories: continued
development/testing/analysis– testing advanced chemical mechanisms
– Further component development, chemical data assimilation, dust, speciated fire emissions,
“in-line” coupling of weather and AQ simulation
– Developmental and experimental testing, integrated quantitative PM capability
– Target operational implementation for initial PM forecasts, NE US: FY14
Expanding, Improving Ozone forecast guidance
• Closer coupling of AQ with NAM; treatments/resolution, horizontal boundary
conditions…
• Development of AK, HI capabilities; target operational implementation in FY10
• Extend forecast range to Day 2 and beyond
20
Acknowledgments:
AQF Implementation Team Members
OCWWS
OCIO
OST/MDL
OST/MDL
Jannie Ferrell
Outreach, Feedback
Cindy Cromwell, Allan Darling, Bob Bunge
Data Communications
Jerry Gorline
Dev. Verification
OST
Marc Saccucci, Tim Boyer, Dave Ruth
Ken Carey, Ivanka Stajner
NESDIS/NCDC
Alan Hall
NDGD Product Development
Program Support
Product Archiving
NOAA/OAR Jim Meagher
NOAA AQ Matrix Manager
NCEP
Jeff McQueen, Pius Lee, Marina Tsidulko, Youhua Tang,
Ho-Chun Huang, Dongchul Kim
*Sarah Lu
*Brad Ferrier, *Dan Johnson, *Eric Rogers, *Hui-Ya Chuang
Geoff Manikin
John Ward, Brent Gordon, Chris Magee
Robert Kelly, Bob Bodner, Andrew Orrison
* Guest Contributors
EMC AQF model interface development, testing and integration
Global data assimilation and feedback testing (NASA, NESDIS)
WRF/NAM coordination
Smoke Product testing and integration
NCO transition and systems testing
HPC coordination and AQF webdrawer
NOAA/OAR
Daewon Byun, Shaocai Yu, Daiwen Kang, Hsin-Mu Lin,
David Wong, Daniel Tong, Tianfeng Chai
Roland Draxler, Glenn Rolph, Ariel Stein
CMAQ development, adaptation of AQ simulations for AQF
HYSPLIT adaptations
NOAA/NESDIS
Shobha Kondragunta, Jian Zeng
Matt Seybold, Mark Ruminski
Smoke Verification product development
HMS product integration with smoke forecast tool
EPA/OAQPS
Chet Wayland, Phil Dickerson, Scott Jackson, Brad Johns
AIRNow development, coordination with NAQFC
EPA/ORD
Rohit Mathur, Ken Schere, Jon Pleim, Tanya Otte,
Jeff Young, George Pouliot, Brian Eder
CMAQ development, adaptation of AQ simulations for AQF
21
Operational AQ forecast guidance
www.weather.gov/aq
CONUS Ozone
Expansion Implemented September, 2007
Smoke Products
Implemented March, 2007
Further information: www.nws.noaa.gov/ost/air_quality
22
Appendix
23
Updates in 2008
Operational Products
WRF update (March, 2008)
– Size: Expanded domain by 18%
– Model Parameterizations: Implemented gravity wave mountain drag parameterization,
modified horizontal advection for better mass conservation, Improved surface longwave
radiation calculation, Improved soil moisture calculations OCONUS
– Data assimilation: Upgraded GSI with NMM bkgd error covariances, more satellite obs
Ozone Predictions: Emissions Updates (May, 2008)
– Point, area and mobile source emissions: updated based on NEI (2005) and projected for
the current year.
•
EPA Office of Transportation and Air Quality on-road emissions estimates
•
EGU sources: 2006 CEM data projected for 2008.
– Biogenic sources: updated with BEIS 3.13
Smoke Predictions (December, 2007):
– Increased vertical resolution: now at full NAM 60-layer resolution
– Increased computational efficiency, faster product delivery
24
Continuing Science Upgrades
Improvements to the expanding NAQFC
Continuing R&D required
• OAR and EPA working actively with NWS to provide prototype capabilities for pre-operational
development, testing experimental production, and implementation
Assuring quality with science peer reviews:
• Design review of major system upgrades (initial, yearly upgrades)
• Diagnostic evaluations with field campaigns and evaluations
• Publication of T&E in peer-reviewed literature
Ozone Capability
–
Otte et al. Weather and Forecasting, 20, 367-385 (2005)
–
Lee et al., J Applied Meteorology and Climatology, in press (2007)
–
Mckeen et al., J. Geophys. Res. 110, D21307 (2005)
–
Yu, et al. , J. Geophys. Res. (2007)
Smoke Tool
–
Prados et al., J. Geophys. Res. 112, D15201 (2007)
–
Kondragunta et al., submitted for publication
–
Rolph et al., submitted for publication
–
Zeng and Kondragunta, ms. in preparation
–
Ruminski, Kondragunta, Draxler and Zeng, in preparation
25
National Air Quality Forecast Capability
Major Components: IOC
NWP Model
NAM/Eta-12
Weather
Observations
NOAA/NWS
NWP Post-processors
for AQ Modules
EPA’s National
AQ Module:
Emissions Preprocessor
PREMAQ
NOAA/OAR and EPA/ORD
Emissions Inventory:
EPA/OAQPS
IT /Comms
NOAA/NWS
AQ Module:
Air Quality Reactive Transport
CMAQ
and
EPA/OAQPS
NOAA/OAR and EPA/ORD
26
Smoke Forecast Tool:
What is it?
Overview
• Passive transport/dispersion computed with HYSPLIT & WRF-NAM (or
GFS, OCONUS). 24-hr spin-up, 48-hour prediction made daily with 6Z
cycle
Fire Locations
• NESDIS/HMS: Filtered ABBA product (only fires with observed
associated smoke)
Emissions
• USFS’ BlueSky algorithm for emitted PM2.5
Smoke Transport/dispersion
• HYSPLIT (Lagrangian); plume rise based on combustion heat and
meteorology
Verification
• Based on satellite imagery for footprint of extent of observed smoke in
atmospheric column exceeding threshold of detection
27
Smoke Forecast Tool
Major Components
NWP Model
NAM/WRF-NMM
Weather
Observations
NOAA/NWS
NESDIS HMS
Fire Locations
NWP Post-processors
for AQ Modules
HYSPLIT Module:
NOAA/OAR
USFS’s BlueSky
Emissions Inventory:
USFS
Verification:
NESDIS/GASP Smoke
28
New Threshold and FC
Good at
Lower
Threshold
FC increase
FC decrease
Good at Lower
Threshold
Old Threshold
Miss at
Lower
Threshold
Good
Good
Good
Miss at Lower
Threshold
New Threshold
Prediction
Miss
Miss
Observation
29
Developmental Aerosol Predictions:
Summary Verification, 2007-2008
Fraction Correct, Aerosol Predictions, 0600 UTC
Daily Maximum of 1-h Avg, Full 5X Domain
2007
Hit Accuracy
Target
1.00
0.90
0.80
0.70
0.60
0.50
18- 15- 22- 29- 5- 12- 19- 26- 5- 12- 19- 26- 29- 16- 23- 30- 7- 14- 21- 28- 4- 11- 18- 25- 2Jan Jan Jan Jan Jan Feb Feb Feb Feb Mar Mar Mar Mar Apr Apr Apr Apr Apr May May May May Jun Jun Jun Jun Jul
Fraction Correct, Aerosol Predictions, 0600 UTC
9Jul
16- 23Jul Jul
Emissions
Correction
2008
Daily Maximum of 1-h avg, Full 5X Domain, Th=40 mg/m
3
30- 6- 13- 20- 27Jul Aug Aug Aug Aug
1.00
0.90
0.80
0.70
Fraction Correct
0.60
Target
0.50
7Jan
14Jan
21Jan
28Jan
411- 18Feb Feb Feb
253Feb Mar
10Mar
17Mar
24Mar
31Mar
7Apr
14Apr
21Apr
28Apr
512- 19- 26May May May May
2Jun
9Jun
16Jun
23Jun
30Jun
7Jul
14Jul
21Jul
28Jul
4Aug
11- 18Aug Aug
251Aug Sep
30
Daily Aerosol Verification
March 5 – 20, 2008
Note: sudden improvement from March 16
31