1 Using Hemispheric-CMAQ to Provide Initial and Boundary Conditions for Regional Modeling Joshua S. Fu 1, Xinyi Dong 1, Kan Huang 1, and Carey Jang 2 1.

1

Using Hemispheric-CMAQ to Provide Initial and Boundary Conditions for Regional

Modeling

Joshua S. Fu1, Xinyi Dong1, Kan Huang1, and Carey Jang2

1University of Tennessee, Knoxville

2Office of Air Quality Planning and Standards, USEPA

Oct 17, 2012

1111thth Annual CMAS Conference Annual CMAS Conference

15-1715-17th th October, 2012October, 2012

2

Objective

Study the potential applications of Hemispheric-scale CMAQ (H-CMAQ) to:

Provide initial and boundary conditions (IC/BC) for Provide initial and boundary conditions (IC/BC) for regional modelingregional modeling

Provide support for hemispheric scale studies with Provide support for hemispheric scale studies with CMAQCMAQ

intercontinental transportintercontinental transport public relevant background ozonepublic relevant background ozone Arctic black carbonArctic black carbon

3

Motivation Uncertainties remain in current methods to

produce IC/BC

Profile IC/BC: Limited horizontal and vertical variationsProfile IC/BC: Limited horizontal and vertical variations Ordinary downscaling IC/BC: excessive transport of OOrdinary downscaling IC/BC: excessive transport of O33 in the upper in the upper

tropospheretroposphere Dynamic downscaling: inconsistent modeling system, discontinuity Dynamic downscaling: inconsistent modeling system, discontinuity

between different chemical and transport schemesbetween different chemical and transport schemes

Uncertainties remain in current global-regional modeling application Introduce uncertainties during downscalingIntroduce uncertainties during downscaling Global and regional models are not using exactly the same Global and regional models are not using exactly the same

emissions over the same domainemissions over the same domain

4

Methodology OverviewHow to evaluate the capability of H-CMAQ to reproduce air pollutants concentrations, and provide IC/BC for regional CMAQ modeling ?

GEOS-Chem

Re-projection

Species mapping

Re-projection

Layer interpolation

Dynamic downscaling

H-CMAQ

Nest down

IC/BC IC/BC

Regional CMAQ Regional CMAQ

Part 1. Compare modeling results from two different models

Part 3. Compare modeling results from two different CMAQ simulations driven by two different sets of IC/BCs

Part 2. Produce IC/BC

5

Domain ConfigurationGEOS-Chem H-CMAQ

CMAQ GEOS-Chem H-CMAQ CMAQ

Model Version V9-01-02 V4.7.1 V4.7.1

Horizontal Grid Spacing

2 x 2.5° 108 x 108 km 36 x 36 km

Vertical Grid Spacing

48 layers (0.01mb top)

34 layers (100mb top)

34 layers (100mb top)

Projection Geographic Lat/Lon Polar Lambert Conformal Conic

Met Field GEOS-5 WRFv3.2.1 WRFv3.2.1

3 Domains:- Global:- Global:

GEOS-ChemGEOS-Chem- North Hemisphere: - North Hemisphere:

H-CMAQH-CMAQ- CONUS:- CONUS:

CMAQCMAQ

Model Configurations

6

H-CMAQ Emission Preparation Base data: 2008 Annual Emission Database for Glob

al Atmospheric Research (EDGARv4.2) Species: SO2, NOx, CO, NMVOC, NH3, PM10, PM2.5, OC, BC Resolution: 16 sectors, annual emission, 0.1x0.1°grid spaci

ng Three processing steps:

Temporal Allocation Speciation Vertical Distribution

0

100

200

300

400

500

600

700

800

900

1000

0 5 10 15 20

Percentage (%)

Hei

ght (

met

ers)

Power Generation

Hourly Scaling Factors

Vertical Profile

0

0.5

1

1.5

2

2.5

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24Hour

Scal

ing

Fact

or

Industry

Power Generation

Residential

Transportation

7

Model-ready Emissions

SO2 (10-10 moles/(m2s))

EDGAR

H-CMAQ

H-CMAQ emission in July 2008 compared with EDGAR emission, with both units transferred to moles/(m2s)

NOx (10-10 moles/(m2s)) CO (10-7 moles/(m2s))

8

H-CMAQ Meteorology Field

Parameters Russia US China Benchmark*

T2 Bias (K) -0.51 -0.64 -1.35 ±0.5Gross Error(K) 1.95 2.27 2.55 ≤2IOA 0.95 0.94 0.92 ≥0.8

Wind Speed Bias (m/s) 0.26 -0.33 0.23 ±0.5RMSE (m/s) 1.57 1.83 1.66 ≤2

IOA 0.77 0.72 0.72 ≥0.6Wind Direction

Bias (deg) 0.99 0.05 2.6 ±10Gross Error (deg) 37.01 36.84 48.82 ≤30

Water VaporMixing Ratio

Bias (g/kg) -0.25 0.71 -0.36 ±1Gross Error (g/kg) 1.16 1.62 1.91 ≤2IOA 0.91 0.93 0.94 ≥0.6

Observation Sites WRF Evaluation Statistics (July, 2008)

WRF performance is evaluated against NCDC data Three countries (Russia, United States, and China) are selected to Three countries (Russia, United States, and China) are selected to

show the hemispheric scale WRF performanceshow the hemispheric scale WRF performance Higher latitude region has a generally better predictions from WRF, Higher latitude region has a generally better predictions from WRF,

while the overall meteorology filed is reliable for H-CMAQwhile the overall meteorology filed is reliable for H-CMAQ

*Bench mark values suggested by Emory et al. (2001)

9

Global Scale Models Performance in CONUS

GEOS-Chem -Projection distortion along the latitude -Hard to capture peak values

H-CMAQ-Projection distortionalong the diagonal

StatisticsDaily Max O3 July 2008

Both GEOS-Chem and H-CMAQ generally reproduce ozone distribution over CONUS

ppbv

ppbv

10

Boundary Conditions Profile BC derived from H-CMAQ is close to dynamic downscaled BC fr

om GEOS-Chem

East Boundary

0

2

4

6

8

10

0 20 40 60 80 100

ozone concentration (ppbv)

Hei

ght (

km)

North Boundary

0

2

4

6

8

10

0 20 40 60 80 100

ozone concentration (ppbv)

Hei

ght (

km)

West Boundary

0

2

4

6

8

10

0 20 40 60 80 100

ozone concentration (ppbv)

Hei

ght (

km)

South Boundary

02468

10

0 20 40 60 80 100

ozone concentration (ppbv)

Hei

ght (

km)

11

Modeling with Different IC/BCs

Driven by IC/BC from GEOS-Chem

Driven by IC/BC from H-CMAQ

StatisticsDaily Max O3 July.2008

ppbv

ppbv

12

Summary Preliminary Conclusions:

Hemispheric-CMAQ is capable of reproducing the air quality generally aHemispheric-CMAQ is capable of reproducing the air quality generally agree well with observation over CONUS domain, and IC/BC derived from gree well with observation over CONUS domain, and IC/BC derived from Hemispheric-CMAQ is close to that from GEOS-ChemHemispheric-CMAQ is close to that from GEOS-Chem

Regional CMAQ simulation over CONUS domain demonstrate the capabiRegional CMAQ simulation over CONUS domain demonstrate the capability of H-CMAQ to provide IC/BC for regional modelinglity of H-CMAQ to provide IC/BC for regional modeling

Next steps:

Stability check: H-CMAQ derived IC/BC performance in different seasons,Stability check: H-CMAQ derived IC/BC performance in different seasons, at different regional areas at different regional areas

Inter-continental transport application: can H-CMAQ better track the polInter-continental transport application: can H-CMAQ better track the pollutants transportation ?lutants transportation ?

Update H-CMAQ emission with local emission inventory: NEI, EMEP, BRAUpdate H-CMAQ emission with local emission inventory: NEI, EMEP, BRAVO, CAC, INTEX-B …VO, CAC, INTEX-B …

13

Acknowledgement

The authors acknowledge GEOS-Chem team for their The authors acknowledge GEOS-Chem team for their user support.user support.

14

Thanks !