S.T. Rao, Rohit Mathur, Rob Pinder, Jon Pleim,Shawn Roselle, Dev Roy

Atmospheric Sciences Modeling DivisionARL/NOAA and NERL/EPA

Use of Remote Sensing Air QualityInformation in Regional Scale Air

Pollution Modeling:Current Use and Requirements

Motivation• Applications of regional AQ models are

continuously being extended to address pollutionphenomenon from local to hemispheric spatialscales over episodic to annual time scales

• The need to represent interactions between physical and chemical processes at these disparate spatial and temporal scales requires use of observational data beyond traditional surface networks

Use of Remote Sensing Informationin Regional AQMs

• Evaluation/Verification of model results High spatial resolution over large geographic regions of

remote sensing data is attractive• Improve estimates of model parameters

Location and effects of clouds (e.g., photolysis) Key meteorological parameters (e.g., PBL) Lateral Boundary conditions (LRT effects) Emissions (e.g, wildland fires, accountability)

• Chemical data assimilation Improving short-term air quality forecasts Identification of model deficiencies

Improving Key Meteorological Parameter EstimatesImpact of assimilation of Solar Radiation from GOES (7/31/1988; 1900z)

MM5 MM5 w/GOES Solar

Surface Radiation (W/m2)Large areas with reduced cloudeffects

Ground Temperature (oK)Up to 6 oK increase in surfacetemperature in affected areas

PBL Height (m)Up to 1000m higher PBL inaffected areas

Improving Model Parameter EstimatesPhotolysis attenuation by clouds

• J = Jclear f(transmissivity,fraction)

0

1Model (MM5) Satellite Derived (GOES)

Cloud Fraction

Transmissivity

Photolysis attenuation by clouds: CMAQ (contd.)Base Satellite Based

JNO2 (s-1)

O3

Base Satellite Based Observed

Courtesy: A. Biazar, Univ. of Alabama21Z on 8/24/2000

Photolysis attenuation by clouds: CMAQ (contd.)Base Satellite Based

JNO2 (s-1)

O3

Base Satellite Based Observed

Courtesy: A. Biazar, Univ. of Alabama21Z on 8/24/2000

Photolysis attenuation by clouds (contd.)

Courtesy: A. Biazar, Univ. of Alabama

Observed O3 vs Model Predictions

(South MISS., lon=-89.57, lat=30.23)

-40

-20

0

20

40

60

80

100

8/30/00 0:00 8/30/00 6:00 8/30/00 12:00 8/30/00 18:00 8/31/00 0:00 8/31/00 6:00 8/31/00 12:00 8/31/00 18:00

Date/Time (GMT)

Ozo

ne

Co

nc

en

tra

tio

n (

pp

b)

Observed O3

Model (cntrl)

Model (satcld)

(CNTRL-SATCLD)

Improving Model Parameter EstimatesWildfire Emission Specification in CMAQ

0

3

6

9

12

0 3 6 9 12

IMPROVE Average Total Carbon

Mo

de

led

A

ve

ra

ge

T

ota

l C

arb

on

0

3

6

9

12

0 3 6 9 12

IMPROVE Average Total Carbon

Mo

de

led

Av

era

ge

T

ota

l C

arb

on

0

2

4

6

8

10

12

14

16

18

20

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31

May 2001

Base Run

Reallocated Run

IMPROVE Obs.

OC

+E

C (

_g

m-3)

• Spatial allocation of emissions based on forest surrogates leads to unrealistic spatial distributions

• Reallocate NEI prescribed and wildfire emissions using MODIS Rapid Response Fire pixel count

Reallocation helps reduce bias and improves correlationin total carbon predictions

MODIS Fire Detect

Bas

eR

eallo

cate

d

CMAQ SCIAMACHY

On-going efforts:• Model Evaluation (2004)• Test and Improve NOx Emission Inventories• Accountability studies – Track impact of regulations on observed regional and local• AQ over time using both model and observations

molec/cm2

Evaluation of Modeled Spatial DistributionsNO2 Columns: Summer 2004

(Courtesy: R. Martin)

Similar discrepancies at surface

Comparable spatial distributionsSCIAMACHY higher in rural areas • higher regional background• missing source (lightning), or • chemistry (NOxNOy) too rapid CMAQ higher downwind of urban areas (e.g., Atlanta, St. Louis), Point sources• Opposite trend compared to GEOS-CHEM (resolution/chemistry)• air mass factor from GEOSCHEM (NOx lifetime difference due to resolution)

Evaluation of NO2 Spatial Distributions (contd.)M

odel

- S

CIA

MA

CH

Y

7/18/04 7/19/04 7/20/04

7/21/04 7/22/04 7/23/04

5 10 15 20 25 30 35µg/m3

Diagnosing Model Performance: Eta-CMAQ PM2.5 Forecasts

7/16 7/17 7/18 7/19 7/20 7/22

July 16-22, 2004: Evidence of Effects of Long Range Transport Originating from Outside the Modeled Domain Evolution of Model and Observed Aerosol Optical Depth

MODIS

Model

Transport from outside the domain influences observed PM concentrations whichare grossly under-predicted during this period

• Model picks up spatial signatures ahead of the front • Under predictions behind the front (due to LBCs)

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 AOD

Distribution of measured carbonaceousaerosol at STN sites within domain Regional enhancement in TCM on July 17-20 suggests influence of wildfires on air masses advected into the domain

Further Evidence7/14/04 7/15/04 7/16/047/13/04

Long Range Transport of AlaskanPlume

Can AOD assimilation improve model PMforecast?

MODIS AOD Assimilation: Impact on Surface PM2.5 Model Performance

• Reduced Bias/Error• Improved Correlation

AIR

NO

WJu

ly 1

9-23

, 04

STN

July

20,

04

Domain median surface levels enhanced by 23 - 42% due to Alaskan fires on different days

Summary• Air quality remote sensing data is useful for model

evaluation and improvements What level of quantitative agreement is acceptable? Need for harmonization between assumptions used in

retrieval and CTM process algorithms (e.g., AOD, NO2columns) for more rigorous quantitative use

• Potential for use in chemical data assimilation Simultaneous information on multiple chemical species Use in air quality forecasting would require availability in

near real-time

• Columnar distributions are a good starting point, butthere is a need for better vertical resolution Discern between BL and FT Help improve FT predictions in regional AQMs

• Linking with global models that assimilate satellite data• Direct assimilation in regional AQMs

Acknowledgements

• Collaborators: Arastoo Biazar and Dick McNider Randall Martin James Szykman Shobha Kondragunta

Disclaimer: The research presented here was performed under theDisclaimer: The research presented here was performed under theMemorandum of Understanding between the U.S. Environmental ProtectionMemorandum of Understanding between the U.S. Environmental ProtectionAgency (EPA) and the U.S. Department of CommerceAgency (EPA) and the U.S. Department of Commerce’’s National Oceanic ands National Oceanic andAtmospheric Administration (NOAA) and under agreement numberAtmospheric Administration (NOAA) and under agreement numberDW13921548. This work constitutes a contribution to the NOAA Air QualityDW13921548. This work constitutes a contribution to the NOAA Air QualityProgram. Although it has been reviewed by EPA and NOAA and approved forProgram. Although it has been reviewed by EPA and NOAA and approved forpublication, it does not necessarily reflect their views or policies.publication, it does not necessarily reflect their views or policies.

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

Model and Observed Aerosol Optical Depths (AOD)

Model (top panels) MODIS (bottom panels)

8/17/048/16/048/15/04

AERONET

Modeled AOD at 550nmAERONET Data at 440 and 670nm

Reasonable model simulation of spatial and temporal variability in AOD is possible Can AOD assimilation improve model PM forecasts?