DYNAMO: DYnamic Inputs of Natural Conditions for Air Quality MOdels - - PowerPoint PPT Presentation

DYNAMO: DYnamic Inputs of Natural Conditions for Air Quality MOdels

AQAST Year 3 Tiger Team Daniel Cohan, Loretta Mickley, Richard McNider, Arastoo Pour Biazar, Bryan Duncan

Key Additional Participants in DYNAMO

Air Quality Management partners:

• EPA: Jesse Bash, Pat Dolwick, Chris Misenis

– 2011 CMAQ CONUS simulation

• Texas Commission on Environmental Quality:

Mark Estes

• California Air Resources Board: Jeremy Avise

Students: Ben Lash (Rice), Erin Chavez Figueroa (Rice), and Lulu Shen (Harvard) Postdoc: Dr. Rui Zhang (Rice)

DYNAMO Objectives

• Stratospheric ozone: Satellite-based daily varying

columns, to replace weekly averages – Impacts on tropospheric photochemistry

• Clouds & Radiation: GOES-based clouds for photolysis

rates and photosynthetically active radiation (PAR) – Impacts on biogenic VOC, ozone, and PM

• Soil NO: Implement & extend BDSNP scheme

– Impacts on NO2 columns, ozone, and PM

• Provide data on EPA’s RSIG
• Review paper: Satellite-based inputs of natural

conditions for regulatory modeling

Total ozone from OMI

DU

Daily stratospheric ozone columns

• Motivation:

– Synoptic variability in stratospheric ozone (Hudson et al., 2003) – Models interpolate from weekly or monthly averages – Impact on tropospheric UV & photochemistry??

• Satellite observations:

– Large daily variability in

• zone columns in winter

and spring

Images prepared by Lulu Shen and Loretta Mickley using data from Wargan et al., 2014

30-day moving avg daily values

OMI-MLS ozone over Chicago, 2005

GEOS-Chem simulation for 2005 Difference from 30-day moving average for:

• Strat ozone column
• JO3 photolysis
• MDA8 ozone
• OH

Conclusion: Little impact on peak summer ozone

• Impacts in other

seasons??

Impacts of daily varying stratospheric

• zone columns on tropospheric oxidants

Daily variability: Summer over Chicago

Image prepared by Lulu Shen and Loretta Mickley

Satellite-based clouds: Impact on PAR

Satellite PAR (Pinker UMD) WRF PAR

Pinker (UMD) Satellite PAR << WRF PAR for July 2007

Initial evaluation of UAH satellite PAR data for September 2013 showed (Satellite PAR < WRF PAR), but less difference and more +/- (See Rui Zhang (Rice) poster)

Satellite-based radiation better matches SURFRAD monitors

Solar Insolation from:

• Satellite (Pinker UMD)
• SURFRAD observations
• WRF simulation

OBS(W/m2) SIM(W/m2) R RMSE(W/m2) NMB (%) NME(%) PAR(WRF) 90 120 0.90 74 36.8 47.8 PAR(UAH) 90 117 0.92 70 32.3 44.4

UAH satellite-based data slightly outperform WRF for Sept. 2013 (results from preliminary check before screening & refining data)

UMD satellite data far

• utperform WRF, correct
• ver-prediction bias

Impact of UAH satellite PAR on emissions

(Rui Zhang et al. poster, for Sept. 2013 simulation)

Isoprene Mono- terpenes

Base Emissions from MEGAN Satellite (UAH) minus WRF

Improved Soil NO Emissions Scheme

• Berkeley Dalhousie Soil NO Parameterization (BDSNP)

– Introduced by Hudman et al. 2012; In GEOS-Chem

• Ben Lash (Rice) implemented in CMAQ (inline biogenics)

– Providing to EPA (J. Bash) for upcoming CMAQ release – Also shared with UMD and other interested parties

• Base-level emissions factors for each land cover

category based on Steinkamp and Lawrence 2011

• Fertilizer and N deposition add N to soil reservoir
• Meteorology influences emissions

– Soil moisture and T from land surface model, instead of rainfall and air T – Pulse of emissions when rain follows dry period

BDSNP >> Y&L for Soil NO Emissions

(Change in NO emissions per cell for Aug. 2005)

BDSNP ~ 2x Y&L soil NO. Also, very different spatial & temporal patterns and responses to meteorology -- Can’t just scale Y&L.

∆24-hr ozone due to BDSNP vs. Y&L

(shown for August 2005 CMAQ, BDSNP – Y&L)

Over-predictions of Lamsal NO2 columns

(shown for August 2005)

Spatial pattern does not correspond to soil emissions, so other factors likely drive CMAQ

• ver-prediction of satellite-observed NO2 in this case. Needs further investigation.

Potential Extensions of BDSNP Soil NO

• EPIC dynamic fertilizer to replace Potter et al. 2010
• More evaluations vs. ambient & satellite NO2
• Offline version of BDSNP for direct creation of soil

NO emissions using WRF or other meteorology data – Requires assumptions about N-deposition

• Add soil emissions of HONO (Su et al. 2011)
• Ultimate goal: More mechanistic model to simulate

soil emissions of all N compounds (NO, NH3, HONO, N2O, etc.)

– Could a mechanistic model simulate responses of these emissions to agricultural control strategies??

Preliminary Findings and Recommendations

• Daily-varying stratospheric ozone columns

– Finding: Little impact on summertime MDA8 ozone – Recommendation: Investigate impacts on winter & spring photochemistry

• Satellite-based clouds, solar radiation, J-values & PAR

– Finding: Satellites can correct WRF over-predictions of PAR

• Need further refinement and evaluation of UAH data

– Recommendation: Investigate joint influences of satellite-based clouds on BVOC emissions and photochemistry

• Soil NO scheme

– Finding: BDSNP much higher than Y&L, and with very different spatio-temporal patterns, for soil NO emissions – Recommendations: (1) Include in next CMAQ release; (2) More evaluation with observations; (3) Extend to EPIC dynamic fertilizer inputs, offline inventory creation, and multi-pollutant emissions