in Air Quality Decision Models AQAST Project Physical Atmosphere Panel Meeting April 2526 2012 Atlanta GA Richard McNider Arastoo Pour Biazar or Arastoo McBiazar University of Alabama in Huntsville ID: 558958
Download Presentation The PPT/PDF document "Use of Satellite Data to Improve the Phy..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Use of Satellite Data to Improve the Physical Atmosphere
in
Air Quality Decision
Models
AQAST Project
Physical Atmosphere Panel Meeting
April 25-26, 2012
Atlanta, GA
Richard
McNider
Arastoo
Pour
Biazar
(or
Arastoo
McBiazar
)
University of Alabama in HuntsvilleSlide2
Physical Atmosphere Advisory Team
Wayne
Angevine
- NOAA – Boundary Layer
Observations
Bright
Dornblauser
– State of Texas – Regulator Model Evaluation
Mike
Ek
/Jeff McQueen
– NOAA – Land Surface Modeling
Georg
Grell
–
NOAA
–
Clouds and Modeling
John Nielsen-Gammon
– Texas A&M – Model Evaluation
Brian Lamb
– Washington State University – Emissions/ Model Evaluation
Pius Lee
– NOAA – Air Resources Laboratory – Air Quality Forecasting
Jon
Pleim
– US EPA – Boundary Layer Modeling
Nelsen Seaman
– Penn State University – Meteorological Modeling
Saffett
Tanrikulu
-
SF Bay Area Air Quality District – Meteorological Modeling Slide3
Also had participation from Local and Regional Air Quality Community in and around Atlanta
Brenda Johnson – EPA Region IV
Richard
Monteith
– EPA Region IV
Steve Mueller – Tennessee Valley Authority
Justin Walters – Southern Company
Jim
Boylan
– Georgia Environmental Protection Division
Tao
Zeng
- Georgia Environmental Protection Division
Lacy Brent – Discovery AQ/U. Maryland
Kiran
Alapaty
– EPA-NERL
Jim
Szykman
- EPA- NERL
Ted Russell - Georgia Tech
Talat
Odman
– Georgia Tech
Maudood
Khan – University Space Research Association
Scot t
Goodrick
– U.S. Forest Service Slide4
Physical Atmosphere Can Significantly Impact Atmospheric Chemistry and Resulting Air Quality
Most Importantly the Physical Atmosphere Can Impact Control Strategy Efficacy and Response
Temperature, Clouds, Mixing Heights, Humidity and Turbulence Can All Impact Air Quality
Satellite
Observation
Temperature
Mixing Heights
CloudsSlide5
AGENDA
AQAST PHYSICAL ATMOSPHERE MEETING
April 25-26, Atlanta Georgia
April 25
12:30 PM Lunch
2:00 PM Introductions
2:15 PM Background and Charge – Dick McNider 2:45 PM Physical Issues and Shortcomings in Physical Atmosphere Modeling for SIP or Forecasting (10-15 minute presentations) General Nelson Seaman
Saffet Tanrikulu
James Boylan
Scott Goodrick Wayne
Angevine 4:00 Break Slide6
4:15 Physical Issues and Shortcomings in Physical Atmosphere Modeling for SIP or Forecasting (continued)
General
Pius Lee
Bright
Dornblauser
Jeff McQueen Steve Mueller Lacey Brent (Discovery AQ)
Maudood Khan Clouds and Photolysis Arastoo
Biazar Kiran
Alapaty
6:00 PM Recap and Adjourn6:30 -8:30 PM Reception Slide7
April 26
8:00AM -8:30AM Continental Breakfast
8:30 AM Physical Issues and Shortcomings in Physical Atmosphere Modeling for SIP or Forecasting (continued)
Land Surface –PBL - Emissions
Jon
Pleim John Nielsen-Gammon
Ted Russell Brian Lamb9:30 AM Discussion of Use of Satellite Information to Improve the Physical Atmosphere Overview – Dick McNider
Land Surface – Jon Pleim, Jeff McQueen, Maudood
Khan Clouds and Photolysis– Arastoo
Biazar, Kiran
Alapaty, Saffet
Tanrikulu
Winds – Bill
Murphrey
/ Dick
McNider
/Seaman
10:30 AM Break Slide8
10:45 AM Discussion of Use of Satellite Information to Improve the Physical Atmosphere
General ( Participation by all)
12:00
NOON
Lunch
1:00 PM Selection of Priorities – Lead (Dick
McNider) Participation by All 2:00 PM Formation of Application Paths and Team Formation 3:00 PM Recap and AdjournSlide9
The presentations by both members of the panel and by local participants brought up a wide variety of topics
Coastal clouds in California
Nighttime Mixing in Houston and Atlanta
Winds for forest fire smoke transport in Georgia
Snow cover in Spring in West (photolysis and land surface
energetics)
Tropospheric/Stratospheric exchange for background ozone in the Pacific NorthwestTopographic effects on 8 hour standardsUrban/Rural bias in NO2 which may be related to physical atmosphere in Mid-AtlanticRepresentativeness of SIP Meteorology in GeorgiaSlide10
Categorization Summary
Clouds for Photolysis, Vertical Mixing and Aqueous Chemistry
‐
Angevine
, Tanrikulu, Biazar
, Alapaty, BoylanStable Boundary Layer for Vertical Mixing, Winds, Cold Pooling -
Seaman, Boylan, Lee, Russell, LambLand Surface for Fluxes and Deposition - Angevine, DornBluaser, Pleim
, Tanrikulu, LeeWinds for Transport and Dilution
- Dornblaser, Lee, Odman
Mixing Heights for Dilution and Plume Rise - McQueen, GoodrickTopography
– Seaman, Mueller, LambSnow Cover for Land Surface and Photolysis ‐ Tanrikulu
Tropospheric
/Stratospheric Exchange for Ozone Background -
Lamb,
Biazar
Slide11
Potential For Use of Satellite Data For Improvement and/or Verification
Clouds for Photolysis, Vertical Mixing and Aqueous Chemistry
‐
Angevine
, Tanrikulu, Biazar
, Alapaty, Boylan -
VERY HIGHStable Boundary Layer for Vertical Mixing, Winds, Cold Pooling - Seaman, Boylan, Lee, Russell, Lamb - MODERATE Land Surface for Fluxes and Deposition -
Angevine, DornBluaser, Pleim,
Tanrikulu, Lee - HIGH Winds
for Transport and Dilution - Dornblauser, Lee, Odman
– MODERATE Mixing Heights for Dilution and Plume Rise
- McQueen,
Goodrick
–
LOW/MODERATE
Topography
– Seaman, Mueller, Lamb -
LOW
Snow Cover for Land Surface and Photolysis
‐
Tanrikulu
VERY HIGH
Tropospheric
/Stratospheric Exchange for Ozone Background -
Lamb,
Biazar
–
MODERATE/HIGH
Slide12
Based on Importance to Physical Atmosphere and Potential for Use of Satellite Data Selected Three Major Themes
Clouds
Stable Boundary Layer
3. Land Surface
Slide13
Stable Nocturnal Boundary LayerSlide14
Time series
O
3
O
3
Mon Tue Wed Thur Fri Sat Sun
James
BoylanSlide15
Ozone in Houston
Original
K
zz
8-h Ozone Concentration (ppm)
9525 1/2 Clinton Dr
Date (July 2006 CDT)
Date (July 2006 CDT)
9525 1/2 Clinton Dr
Modified
K
zz
Driven by reanalysis of nocturnal boundary layer mixing
Russell-
OdmanSlide16
AQAST Physical Atmosphere Meeting, April 25-26, Atlanta GA
16
Model deficiency: mismatch in night time decoupling?
Stable regime
regional average surface wind speed for period June 4 – 12 (dark color)
and June 26 – July 3 (light color).
Night time high wind-speed bias
Occurred repeatedly for many days
right after sunset
Frequent surface wind-speed high-bias
Similarity theory for surface layer;
e.g. Ulrike Pechinger et al.
COST 710
, 1997
Texas –
Dornblaser
– Lee Slide17
1-h Ozone Concentration
Original K
zz
Modified K
zz
Russell-
OdmanSlide18
Ramifications
Significantly changes model performance
Less effect on peak ozone
Still non-zero
Major effect on primary/pseudo-primary species concentrations
EC, CO, NO2, PM2.5 New standards raise importance of NO
2.Use of models in health effects research raise importance of bias, diurnal variationTed Russell Slide19
Cold pool modeling
Routine application of prognostic meteorological models including the Fifth-Generation NCAR/Penn State
Mesoscale
Model (MM5) and Weather Research and Forecasting Model
(WRF) with a variety of different physics options, initialization input, vertical and horizontal resolutions, and nudging approaches have failed to replicate the degree and persistence of stagnant meteorological conditions
. (Baker et al., 2011, ES&T).
AIRPACT Forecasts don’t capture elevated wintertime PM2.5 levels
stagnant valley meteorology
woodstove emissions
from Avey, Utah DEQ)
Brian LambSlide20
20
Sub-Km Modeling of the Stable Boundary Layer
Combined modeling and observation
studies Nittany Valley, Central PA
10 km scale
WRF smallest domain (0.444 km horizontal resolution)
Observation NetworkSlide21
21
a) 0500-0700 UTC
b) 0600-0800 UTC
c) 0700-0900 UTC
e) 0600-0800
UTC
f) 0700-0900 UTC
d) 0500-0700 UTC
Releases at one-hour intervals from Site 9
at
5 m AGL
Sub-Km Modeling of the Stable Boundary
Layer
Tussey RidgeSlide22
Path Forward
Explore mixing formulations for stable boundary layer and role of resolution with MODIS skin temperatures as evaluation metric.
F
h
(
Ri
)Ri
Coarse grid models
Theory Slide23
Use MODIS Skin Temperatures for Model Evaluation Slide24
GOES Derived Skin Temperature
MODIS Derived Skin TemperatureSlide25
Nocturnal boundary layer formation dependent on topography has implications for 8 hour attainment at high elevations.
Steve MuellerSlide26Slide27
CO profiles from P3 upwind, over, and downwind of Nashville (symbols)
Tracer profile from 1D cloud-aware PBL model (early version of TEMF)
Lower panel shows what happens when cloud-induced mixing is not present
CO profiles from P3 upwind, over, and downwind of Nashville (symbols)
Tracer profile from 1D cloud-aware PBL model (early version of TEMF)
Lower panel shows what happens when cloud-induced mixing is not present
Wayne
Angevine
Cloud Mixing Changes Effective PBL HeightSlide28
Southeast Land cells
10X10 cells
Over RDU
Surface
Insolation
Diff:
(KFC
-
BASE)
W/m^2
Kiran
AlapatySlide29
Tests in Texas showed changes in cloud locations and
radiative
properties can change ozone by 70ppb Slide30
Too Many Options Not Enough Information on Performance!Slide31
It Rains Cats & Dogs in a Clear Sky!!!
(for convective clouds in WRF)
Radiative
effects were not included for WRF
subgrid
scale clouds.
Kiran
AlapatySlide32
Inconsistency in Cloud Handling in Models
MM5/WRF do not consider sub-grid clouds in radiation calculations.
Clouds in MM5/WRF not used in CMAQ (clouds
rediagnosed
) for wet chemistry mixing.
CMAQ photolysis rates not based on CMAQ clouds but on MM5/WRF liquid water profiles.
These inconsistencies make correction difficult!Slide33
Satellite data can be used as a metric to test model cloud agreementSlide34
Path Forward
Insert satellite measures of
radiative
properties directly in models.
Use satellite derived measures of insolation
based on satellite clouds rather than modeled insolation using model clouds (McNider et al. 1995)
Use satellite cloud transmittance in photolysis calculations (Biazar et al. 2007)Improve physical parameterizations using satellite data as performance metric Correct model radiation (Alapaty et al. 2012)
Connect PBL and cloud schemes (Angevine 2012)
Assimilate satellite data to improve the location and timing of cloudProvide dynamical cloud support and cloud clearing (McNider and
Biazar 2012) Slide35Slide36Slide37Slide38
Land Surface Slide39
Factors controlling surface temperatures are complex and many models have created complex land use models that in the end require many ill defined parameters
. Slide40
Land surface
Top-level soil temperature and moisture
BLLAST, 30 June
2011, 14ZSlide41
Air Quality Simulations for SIPs Are Retrospective Studies
Allows use of observations to constrain forecast models
Simple Surface Models Constrained by Observations
Pleim
Xiu Scheme
McNider et al. 1994 / Norman et al. 1995 (ALEXI) Slide42
Pleim-Xiu
- Land surface energy budget
Soil moistureSlide43
Soil Moisture Nudging
Nudge according to model bias in 2-m T and RH compared to surface air analysisSlide44
T-2m bias relative to analysis for January 2006
Q
v
-2m bias relative to analysis for January 2006Slide45
Mean bias for 2m T – August 2006
12km domain:
Most around -1 to +1
Positive bias: N and W regions
Negative bias: S, E
4km:
Most around: -0.5 to +0.5
Negative bias: high along the coast
1km:
Most Negative within -0.5
Negative bias: high along the coast Slide46
McNider
et al. 1995 Surface
Energy Budget
Bulk Heat Capacity
Evaporative Heat Flux
Short-wave
radiation
obtained from Satellite
Morning
EveningSlide47
Satellite
Observation
Assimilation
Control
Satellite Data Can Provide Many More Opportunities for Data Skin Temperature Assimilation (GOES ~5 km and MODIS ~1 km).
Land characteristics especially in Eastern U.S. fine scale variations. Slide48
Model BL Heights (CNTRL)
Aug. 26, 2000, 19:00-21:00 GMT averaged
Model BL Heights
(ASSIMALATED)
Aug. 26, 2000, 19:00-21:00 GMT averagedSlide49
Path Forward
Use satellite skin temperatures in
Pleim
–
Xiu scheme rather than National Weather Service 2 m temperatures
Test McNider et al. scheme using new corrections (use of model skin temperatures and aerodynamic temperatures) suggested by
Mackaro Slide50
Pleim-Xiu
- Land surface energy budget
Soil moisture
Use satellite derived
albedo
and
insolation
Slide51
Soil Moisture Nudging
Nudge according to model bias in 2-m T and RH compared to surface air analysis
Use satellite skin temperatures rather than NWS temperaturesSlide52
Teams are being formed for priority areas
Clouds
( Pour-
Biazar,Alapaty,Nielsen
–Gammon)
Stable Boundary Layer (
McNider, Angevine, Russell,Lee)3. Land Surface – (Pleim
, Angevine, Tanrikulu, McQueen/Ek
)Next Meeting (12-18 mos
) will be on West Coast