NCEP Chemistry Modeling Overview and Status (With a focus on NEMS AQ development) Sarah Lu...
-
Upload
andrew-mcdonald -
Category
Documents
-
view
220 -
download
0
Transcript of NCEP Chemistry Modeling Overview and Status (With a focus on NEMS AQ development) Sarah Lu...
NCEP Chemistry Modeling NCEP Chemistry Modeling Overview and StatusOverview and Status
(With a focus on NEMS AQ development)(With a focus on NEMS AQ development)
Sarah LuNOAA/NWS/NCEP
Environmental Modeling Center
with acknowledgments to many colleagues and collaborators
National Central University, Chung-Li, Aug 26th, 2009
Acknowledgments:Acknowledgments:
EMC AQ group Jeff McQueen, Ho-Chun Huang, Youhua Tang, Dongchul Kim, Marina Tsidulko, Caterina Tassone
EMC UMIG group Mark Iredell, Henry Juang, Shrinivas Moorthi, Tom Black, Jun Wang, Weiyu Yang, Ratko Vasic, Ed Colon
EMC GMB Yu-Tai Hou, Suranjana Saha, Fanglin Yang, Xu Li, Jesse Meng, Yuejian Zhu, Jongil Han, John Ward
EMC GSI group John Derber, Russ Treadon, Daryl Kleist, Haixia Liu
CPC Craig Long, Shuntai Zhou
NWS OST Paula Davidson, Ivanka Stajner
OAR ARL Daewon Byun, Pius Lee, Roland Draxler, Ariel Stein, Hsin-Mu Lin, Daiwen Kang, Daniel Tong, Shao-cai Yu
GSFC Arlindo da Silva, Mian Chin, Thomas Diehl
EPA Ken Schere, Rohit Mathur, Jon Pleim
Howard University Everette Joseph, William Stockwell
NESDIS Shobha Kondragunta, Quanhua Liu, Yong Han, Brad Pierce
National Central University, Chung-Li, Aug 26th, 2009
Comparison of RAQMS OMI+TES reanalysis with IONS ozonesondes
(373 sondes, August, 2006)
PI: ANNE M. THOMPSON Penn State
Fishman, J et al., “Remote Sensing of Tropospheric Pollution from Space”, BAMS June 2008Pierce et al. “Impacts of background ozone production on Houston and Dallas, TX Air Quality during the TexAQS field mission”, Accepted JGR-Atmospheres, February, 2009
Tropospheric biases: +/- 20%
The TES+OMI assimilation results in significant reductions in column, tropospheric (>100mb), and stratospheric (<100mb) biases (all less then 1%)
However, the low tropospheric biases are the result of compensating errors in the upper and lower troposphere.
Extensive chemistry modeling efforts within NOAA Research Laboratories (e.g., ESRL, ARL, GFDL) and NESDIS.
Brad Pierce (NESDIS/STAR)
National Central University, Chung-Li, Aug 26th, 2009
OutlineOutline
NCEP current weather-air quality capabilitiesNational AQ Forecast CapabilityGlobal ozone assimilation
NCEP R&D activities National Environmental Modeling SystemNEMS Interactive atmosphere-chemistry modeling
Proposed enhancementsImpact of dynamic lateral BCs on AQ forecastsImpact of aerosols on weather forecasts
Conclusions
National Central University, Chung-Li, Aug 26th, 2009
NCEP Current Weather-AQ CapabilitiesNCEP Current Weather-AQ Capabilities
6
Model Components: Linked numerical prediction systemOperationally integrated on NCEP’s supercomputer
NCEP mesoscale NWP: WRF-NMMNOAA/EPA community model for AQ: CMAQ
Observational Input: NWS weather observations; NESDIS fire locationsEPA emissions inventory
National Air Quality Forecast National Air Quality Forecast CapabilityCapability
End-to-End Operational CapabilityEnd-to-End Operational Capability
Gridded forecast guidance productsOn NWS servers: www.weather.gov/aq and ftp-serversOn EPA serversUpdated 2x daily
Verification basis, near-real time: Ground-level AIRNow observations Satellite smoke observations
Customer outreach/feedbackState & Local AQ forecasters coordinated with EPAPublic and Private Sector AQ constituentsWebsite monitoring
AQI: Peak Oct AQI: Peak Oct 44
EPA Monitoring Network
Paula Davidson (NWS OST)
National Central University, Chung-Li, Aug 26th, 2009
442 grid cells
265gridcells
268 grid cells
259gridcells166
Grid cells
142142gridcells
CONUS “5x” Domain1. OPS: AQFC Sept. 07 2. EXP: AQFC/CB05 June. 083. DEV: AQFC/CB05-AERO-4
Eastern “3x” DomainSept 05
Northeast US “1x” Domain
Sept 04
Expansion of coverageExpansion of coverage
Jeff McQueen (EMC)
National Central University, Chung-Li, Aug 26th, 2009
NCEP Air Quality Forecast VerificationNCEP Air Quality Forecast Verification
Bia
s (p
pb)
Bia
s (p
pb)
-10
20
Almost the same for NW and Mid WestHigher for NE, SE and Low Miss Valley (increase positive bias)Higher for SW (improve negative bias)
Production Experimental
Jeff McQueen (EMC)
http://www.emc.ncep.noaa.gov/mmb/aq
8 h Avg Ozone Obs vs Fcst8 h Avg Ozone Obs vs Fcst
National Central University, Chung-Li, Aug 26th, 2009
Global Ozone Assimilation in GSIGlobal Ozone Assimilation in GSIWhy assimilate ozone
Ozone forecastsUV Index Forecasts Air Quality Forecasts
Needed for assimilating radiances from IR instruments (e.g. HIRS, AIRS) where ozone influences the accuracy of determining temperatures.
Parameterized ozone physics in GFSProduction and destruction are parameterized from monthly and zonal mean dataset derived from NRL 2D ozone chemistry model
Current and future ozone products to be assimilated at NCEPGFS currently assimilating only NOAA-17 SBUV/2 (nadir obs)Probable data update to NOAA-18 and possible for NOAA-19OMI and GOME-2 total ozone being tested in parallel
offers greater horizontal and latitudinal coverageNRT MLS ozone profile product is being evaluated.OMPS (NPP and NPOESS)
Craig Long (CPC)
National Central University, Chung-Li, Aug 26th, 2009
Total Ozone Analysis Improvements byTotal Ozone Analysis Improvements byAssimilating OMI TOz in addition to SBUV/2Assimilating OMI TOz in addition to SBUV/2
More Structure
Tighter Gradients
Craig Long and Shuntai Zhou (CPC)
National Central University, Chung-Li, Aug 26th, 2009
An Overview of National Environmental An Overview of National Environmental Modeling System (NEMS) Modeling System (NEMS)
National Central University, Chung-Li, Aug 26th, 2009
Earth System Modeling FrameworkEarth System Modeling Framework Modeling framework for the geo-science
community
A software infrastructure that enables different weather, climate, and data assimilation components to operate together on a variety of platformsEarth system models that can be built, assembled and reconfigured easily, using shared toolkits (e.g., data communications, time management, message logging, re-gridding, and error handling) and standard interfaces A growing pool of Earth system modeling components that, through their broad distribution and ability to interoperate, promotes the rapid transfer of knowledge. Community effort, partially supported by NOAAESMF superstructure (grid component, state, and coupler) required for all NEMS componentsESMF infrastructure optional
National Central University, Chung-Li, Aug 26th, 2009
National Environmental Modeling System National Environmental Modeling System (NEMS)(NEMS)
Earth Science Modeling Framework (ESMF)http://www.esmf.ucar.edu
NEMS atmosphereWrite history and Post processorNestingAerosols and ChemistryLandOcean, waves and sea iceIonosphereEnsembleData assimilation
Unified Modeling Infrastructure Group, led by Mark Iredell
NCEP UMIG group routinely meets with GSD and GFDL groups
National Central University, Chung-Li, Aug 26th, 2009
NEMS AtmosphereNEMS Atmosphere
Atmosphere
Dynamics PhysicsDyn-PhyCoupler
NMM-B
Spectral
FIM
Color Key
Generic Component
Generic Coupler
Completed Instance
Under Development
NAM Phy
GFS Phydo nada
unified atmosphereIncluding digital filter
Future Development
ARW
FVCORE
FISL
Navy
Navy
adjoints
Mark Iredell (EMC)
The goal is one unified atmospheric component that can invoke multiple dynamics and physics.
At this time, dynamics and physics run on the same grid in the same decomposition, so the coupler literally does nothing.
FY2010 operational implementation for NEMS NMM-B
Chemistry
GOCART
AQF chem
reduced chemistry
National Central University, Chung-Li, Aug 26th, 2009
Developing an interactive atmosphere-Developing an interactive atmosphere-chemistry forecast systemchemistry forecast system
In-line chemistry advantageConsistent: no spatial-temporal interpolation, same physics parameterizationEfficient: lower overall CPU costsEasy data managementAllows for feedback to meteorology
Requirements:Meteorology and chemistry should be initialized with GSIConform to NCO CCS computer architecture Conform to NCO software & I/O standards (GRIB/BUFR)
NEMS AQ development:NMM-B Chem
In support of regional AQF systemGFS coupled with GOCART
Potential for improving weather forecasts (by improving aerosol-radiation feedback in GFS and atmospheric correction in GSI)Providing LBCs for regional AQF aerosol predictions
National Central University, Chung-Li, Aug 26th, 2009
NEMS Tracer Experiments: NEMS Tracer Experiments:
NMM-B and GFSNMM-B and GFS
National Central University, Chung-Li, Aug 26th, 2009
NEMS NMM-B tracer experimentNEMS NMM-B tracer experiment
Youhua Tang (EMC)
National Central University, Chung-Li, Aug 26th, 2009
GB EAS WAF SAM NAM
T62 L64 30-day experiments: CTR, CLD (Ferrier cloud microphysics), DYN (Adiabatic), SAS (Simplified Arakawa-Schubert convection), TVD (Flux-limited vertical advection)
GLB_SFC GLB_UTLS GLB_ALL
Change in global sum
-2.5
-2
-1.5
-1
-0.5
0
0.5
1 2 3CTRL
CLD
DYN
SAS
TVD-1.37% 0.03% (diffusion off)
NEMS GFS tracer experimentNEMS GFS tracer experiment
IC = 2009/01/01 00Z
Change in total mass loading (scaled by initial values)
National Central University, Chung-Li, Aug 26th, 2009
Zonal mean cross section for SAM_SFC & SAM_UTLS (IC=20090101)Flux-limited vertical advection reduces (but does not eliminate) negative tracer values
NEMS GFS tracer experimentNEMS GFS tracer experiment
National Central University, Chung-Li, Aug 26th, 2009
Global aerosol forecast and Global aerosol forecast and analysis system (GFS-GOCART)analysis system (GFS-GOCART)
National Central University, Chung-Li, Aug 26th, 2009
Goddard Chemistry Aerosol Radiation and Transport Model Goddard Chemistry Aerosol Radiation and Transport Model (GOCART)(GOCART)
National Central University, Chung-Li, Aug 26th, 2009
Global Forecast System (GFS)Global Forecast System (GFS)
Global spectrum model for NCEP operational medium range forecasts
RESOLUTION T382 horizontal resolution (~ 37 km)64 vertical levels (from surface to 0.2 mb)
MODEL PHYSICS AND DYNAMICSVertical coordinate changed from sigma to hybrid sigma-pressureNon-local vertical diffusionSimplified Arakawa-Schubert convection schemeRRTM LW radiation schemeMD Chou SW radiation schemeExplicit cloud microphysicsNoah LSM (4 soil layers: 10, 40, 100, 200 cm depth)
INITIAL CONDITIONS (both atmosphere and land states)NCEP Global Data Assimilation System (GDAS)
National Central University, Chung-Li, Aug 26th, 2009
Gridpoint Statistical Interpolation (GSI)Gridpoint Statistical Interpolation (GSI)
Global/regional analysis system for operational weather forecasts
NCEP 3DVAR ASSIMILATION SYSTEMImplemented with WRF-NMM into the NAM system in June, 2006Implemented for replacement of SSI in the GFS system in May, 2007
SCIENTIFIC ADVANCESGrid point definition of background errors Inclusion of new types of data (e.g., AIRS radiance, COSMIC GPS)Advanced data assimilation techniques (e.g., improved balance constraints)New analysis variables (e.g., SST)
CODE DEVELOPMENTGMAO collaboration through NASA-NOAA-DOD Joint Center for Satellite Data Assimilation (JCSDA)Evolution to Earth System Modeling Framework (ESMF)
National Central University, Chung-Li, Aug 26th, 2009
Nick Nalli (NESDIS)
Impact of aerosols on AVHRR Pathfinder Atmospheres (PATMOS) OI multichannel SST (MCSST) retrievals
National Central University, Chung-Li, Aug 26th, 2009
Aerosol effect on HIRS brightness
temperature retrieval
Aerosol Effect on hirs3_n17
-5
-4
-3
-2
-1
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
BT
dif
fere
nc
e (
K)
.
Quanhua Liu (NESDIS)
National Central University, Chung-Li, Aug 26th, 2009
Global aerosol forecast and analysis systemGlobal aerosol forecast and analysis system
Global forecast and analysis systemModeling
Emissions
Data Assimilation
Regional AQF
Dynamic LBCs
SST Analysis
Atmos. Correction
Validation
Various datasetsAERONET, OMI, CALIPSO
Satellite data
Algorithm
GOCART
MODIS fire emissions
NASA obs and tech
ROSE project
NCEP DSSs
Color key
Goal: Improving weather and air quality forecasts by incorporating prognostic aerosols in GFS and assimilating global aerosol information in GSI via NCEP-
NASA/GSFC-Howard University collaborations
National Central University, Chung-Li, Aug 26th, 2009
Multiple, complementary approaches:On-line systems including GOCART:
GFS/GOCART: new capability being developed GEOS-5/GOCART: NASA/GMAO real-time system GFS~GEOS-5/GOCART: Hybrid model (GEOS-5 dynamics + GFS physics)
Off-line GOCART CTM Driven by GFS meteorology
Phased development:Development of prototype systemTransition to real time systemTransition to operational productionPrototype system extended to include ozone chemistry (if resources available)Transition to NCEP’s climate system (if resources available)
NEMS/GFS-GOCART
Dust-only offline GFS-GOCART
Global aerosol forecast and analysis systemGlobal aerosol forecast and analysis system(-cont’d)(-cont’d)
National Central University, Chung-Li, Aug 26th, 2009
http://www.emc.ncep.noaa.gov/gc_wmb/dkim/web/html/dust_day.html
National Central University, Chung-Li, Aug 26th, 2009
-14.6E 23.5N, 2006072214
0
2
4
6
8
0 0.1 0.2 0.3 0.4 0.5
Aerosol Extinction (1/km)
Hig
ht
(km
)
CALIPSO
MODEL
Comparisons between Model and CALIPSO
(2006072214)
-12.6E 14.9N, 2006072214
0
2
4
6
8
0 0.05 0.1 0.15 0.2 0.25 0.3
Aerosol Extinction (1/km)
Hig
ht
(km
)
CALIPSO
MODEL
Dongchul Kim (EMC)
National Central University, Chung-Li, Aug 26th, 2009
Comparisons between Model and CALIPSO
(2006072705)
-53.3E 21.4N, 2006072705
0
2
4
6
8
0 0.1 0.2 0.3 0.4 0.5
Aerosol Extinction (1/km)
Hig
ht
(km
)
CALIPSO
MODEL
-52.5E 25.0N, 2006072705
0
2
4
6
8
0 0.1 0.2 0.3 0.4 0.5
Aerosol Extinction (1/km)
Hig
ht
(km
)
CALIPSO
MODEL
Dongchul Kim (EMC)
National Central University, Chung-Li, Aug 26th, 2009
Resources !! Code optimization needed
The inclusion of 15 passive tracers leads to ~45% increase in wall time
The 3d atmosphere file sizes increased by the factor of 2.4-2.7
Needed capabilities
Convective transport (under testing for RAS)
Tracer scavenging
Positive definite advection with mass conserving
Challenges for incorporating chemistry Challenges for incorporating chemistry component into NEMS GFS:component into NEMS GFS:
The chemistry modeling efforts will lead to scientific advances and technical upgrades in the NEMS
National Central University, Chung-Li, Aug 26th, 2009
NOAA medium range weather forecasts
Climatology-based aerosol distributions are used in the GFS and background aerosol conditions are assumed in the GSI Community Radiative Transfer Model (CRTM)
Global aerosol products will improve the representation of aerosol distributions and variations within the GFS/GSI system
NOAA air quality forecasts
Default static boundary conditions are used for the developmental aerosol air quality predictions
Global aerosol products will provide improved aerosol lateral boundary conditions for the AQF system and, consequently, improve AQF aerosol forecasts
Proposed EnhancementsProposed Enhancements
National Central University, Chung-Li, Aug 26th, 2009
The impact of aerosols on medium The impact of aerosols on medium range weather forecastsrange weather forecasts
National Central University, Chung-Li, Aug 26th, 2009
U-wind Cross Section at 10W
The intensity and location of African Easterly Jet are affected by background aerosol loading (via direct radiative effect)
Climate Forecast System (CFS):Climate Forecast System (CFS):GFS coupled with GFDL MOM3GFS coupled with GFDL MOM3
OPAC climo. GOCART climo.
National Central University, Chung-Li, Aug 26th, 2009
RMS errors of NH temp for 00Z forecasts
Pressure
Forecast hours
GDAS experiments with different aerosol representations:T126 L64; PRC (climatology) vs PRG (time varying)
RMSE reduced
RMSE increased
National Central University, Chung-Li, Aug 26th, 2009
North America temperature verification
Temperature biases reduced by ~ 10% in lower atmosphere
Climo.
Time-varying
National Central University, Chung-Li, Aug 26th, 2009
The impact of lateral boundary The impact of lateral boundary conditions on air quality forecastsconditions on air quality forecasts
National Central University, Chung-Li, Aug 26th, 2009
4 0 6 0 8 0 1 0 0 1 2 0 1 4 0O 3 (p p b v)
0
4000
8000
12000
16000
Alt
itud
e ab
ove
Sea
Lev
el (
m)
ObservedFixed LBCRAQMS LBCMOZART LBCGFS-O3 LBCIONS LBC1IONS LBC2
Beltsville 20060803 17.92 UTC
4 0 6 0 8 0 1 0 0 1 2 0 1 4 0O 3 (p p b v)
0
4000
8000
12000
16000
Alt
itud
e ab
ove
Sea
Lev
el (
m)
ObservedFixed LBCRAQMS LBCMOZART LBCGFS-O3 LBCIONS LBC1IONS LBC2
Boulder 20060803 19.33 UTC
0 100 200 300 400O 3 (p p b v)
0
4000
8000
12000
16000
Alt
itud
e ab
ove
Sea
Lev
el (
m)
ObservedFixed LBCRAQMS LBCMOZART LBCGFS-O3 LBCIONS LBC1IONS LBC2
Trinidad Head 20060803 21.02 UTC
4 0 6 0 8 0 1 0 0 1 2 0 1 4 0O 3 (p p b v)
0
4000
8000
12000
16000
Alt
itud
e ab
ove
Sea
Lev
el (
m)
ObservedFixed LBCRAQMS LBCMOZART LBCGFS-O3 LBCIONS LBC1IONS LBC2
Huntsville 20060803 17.53 UTC
0 100 200 300 400 500O 3 (p p b v)
0
4000
8000
12000
16000A
ltit
ude
abov
e Se
a L
evel
(m
)
ObservedFixed LBCRAQMS LBCMOZART LBCGFS-O3 LBCIONS LBC1IONS LBC2
Bratt's Lake 20060803 21 UTC
Youhua Tang (EMC)
Ozone Lateral Boundary Conditions Tests Ozone Lateral Boundary Conditions Tests
Tang et al., The impact of chemical lateral boundary conditions on CMAQ predictions of tropospheric ozone over the continental United States, Environmental Fluid Mechanics, 2008
Obs (IONS), Obs (IONS), FixedFixed, , RAQMSRAQMS, , MOZARTMOZART, , GFS-O3GFS-O3
National Central University, Chung-Li, Aug 26th, 2009
Aerosol Lateral Boundary Conditions Tests: Aerosol Lateral Boundary Conditions Tests: Trans-Atlantic dust TransportTrans-Atlantic dust Transport
During Texas Air Quality Study 2006, the model inter-comparison team found all 7 regional air quality models missed some high-PM events, due to trans-Atlantic Saharan dust storms.
These events are re-visited here, using dynamic lateral aerosol boundary conditions provided from dust-only off-line GFS-GOCART.
Corpus Christi - Nat, TX 2006 Observed
CO
NC
(ug/
m3)
60 50 40 30 20 10 0
Corpus Christi - Nat, TX 2006CMAQ base run
CO
NC
(ug/
m3)
60 50 40 30 20 10 0
Corpus Christi - Nat, TX 2006
CMAQ+GFS-GOCART LBC
CO
NC
(ug/
m3)
60 50 40 30 20 10 0
Thomas Jefferson Sch, TX 2006 Observed
CO
NC
(ug/
m3)
80 70 60 50 40 30 20 10 0
Thomas Jefferson Sch, TX 2006 Observed
CO
NC
(ug/
m3)
80 70 60 50 40 30 20 10 0
Thomas Jefferson Sch, TX 2006CMAQ base run
CO
NC
(ug/
m3)
80 70 60 50 40 30 20 10 0
Thomas Jefferson Sch, TX 2006
CMAQ+GFS-GOCART LBC
CO
NC
(ug/
m3)
80 70 60 50 40 30 20 10 0
Karnack C85, TX 2006 Observed
CO
NC
(ug/
m3)
60 50 40 30 20 10 0
29JUL 31JUL 02AUG 04AUG 06AUG 08AUG 10AUG
Karnack C85, TX 2006 Observed
CO
NC
(ug/
m3)
60 50 40 30 20 10 0
29JUL 31JUL 02AUG 04AUG 06AUG 08AUG 10AUG
Karnack C85, TX 2006CMAQ base run
CO
NC
(ug/
m3)
60 50 40 30 20 10 0
29JUL 31JUL 02AUG 04AUG 06AUG 08AUG 10AUG
Karnack C85, TX 2006
CMAQ+GFS-GOCART LBCC
ON
C (u
g/m
3)
60 50 40 30 20 10 0
29JUL 31JUL 02AUG 04AUG 06AUG 08AUG 10AUG
Youhua Tang and Ho-Chun Huang (EMC)
National Central University, Chung-Li, Aug 26th, 2009
NCEP is developing NEMS as next-generation weather forecast system
NEMS R & D efforts continue in interactive atmosphere-chemistry modeling system
NMM-B + Chem
GFS-GOCART
NCEP modeling efforts leverage common modeling framework (ESMF), shared software development (via NOAA-NASA-DOD JCSDA), and research collaborations, such as
GSI ozone and aerosol data assimilation working group (EMC AQ group)
Co-Ops Biomass Burning Emission Committee (Jeff Reid and Shobha Kondragunta)
AeroCOM (Michael Shulz, Stefan Kinne, and Mian Chin)
GEMS/MACC community
In ConclusionIn Conclusion