an observationally-constrained global dust aerosol optical depth (aod) david a. ridley 1, colette l....

Download An Observationally-Constrained Global Dust Aerosol Optical Depth (AOD) DAVID A. RIDLEY 1, COLETTE L. HEALD 1, JASPER F. KOK 2, CHUN ZHAO 3 1. CIVIL AND

If you can't read please download the document

Upload: patience-joseph

Post on 18-Jan-2018

219 views

Category:

Documents


0 download

DESCRIPTION

PM 2.5 from satellite-retrieved AOD Dust has a significant impact on air quality globally van Donkelaar et al., EHP (2010)

TRANSCRIPT

An Observationally-Constrained Global Dust Aerosol Optical Depth (AOD) DAVID A. RIDLEY 1, COLETTE L. HEALD 1, JASPER F. KOK 2, CHUN ZHAO 3 1. CIVIL AND ENVIRONMENTAL ENGINEERING, MASSACHUSETTS INSTITUTE OF TECHNOLOGY 2. DEPARTMENT OF ATMOSPHERIC AND OCEANIC SCIENCES, UCLA 3. ATMOSPHERIC SCIENCES AND GLOBAL CHANGE DIVISION, PACIFIC NORTHWEST NATIONAL LAB AMERICAN GEOPHYSICAL UNION, FALL MEETING, 2015 Radiative impact of dust Dust accounts for a quarter of the total global AOD (model-based estimate) Heald et al., ACP (2014) AOD (23%)SW TOA DRE (28%) LW TOA DRE (74%) PM 2.5 from satellite-retrieved AOD Dust has a significant impact on air quality globally van Donkelaar et al., EHP (2010) AEROCOM dust AOD uncertainty Large spread (40%) in AEROCOM model dust AOD estimates AEROCOM mean of (AEROCOM ensemble median of 0.023) Large spread in AEROCOM model dust AOD estimate (0.028 0.011) Dust AOD Derived from Huneeus et al. (2010) KDE probability distribution mean Methodology to retrieve Dust AOD 3 satellite retrievals, 4 models, 5 years of daily data Satellite AOD AERONET AOD AOD bias correction Model AOD Non-dust AOD Dust AOD Daily, Gridded Seasonal, Gridded PDF Seasonal, Regional PDF Seasonal, Global PDF Model Observation s GEOS-Chem CESM WRF-Chem MERRAERO MISR MODIS Aqua MODIS Terra Satellite AOD bias correction with AERONET Co-located daily AERONET AOD (550nm) used to create seasonal bias correction for satellite AOD Bias assessed globally using GEOS-Chem AOD spatial covariance Average bias correction of 0.0%, -5.1% and +6.0% for MODIS Aqua, MODIS Terra, and MISR AERONET AOD / Satellite AOD Methodology to retrieve Dust AOD Model dust AOD only used to scale from regional to global dust AOD 3 satellite retrievals, 4 models, 5 years of daily data Satellite AOD AERONET AOD AOD bias correction Model dust AOD Satellite AOD Model non-dust AOD Model AOD Non-dust AOD Dust AOD Daily, Gridded Seasonal, Gridded PDF Seasonal, Regional PDF Seasonal, Global PDF Model Observation s GEOS-Chem CESM WRF-Chem MERRAERO MISR MODIS Aqua MODIS Terra Regional Dust AOD Dust AOD PDF derived in the key dust-influenced regions 11 key dust regions account for > 75% of global dust in models For each 2 x 2.5 grid box: Dust AOD sat = (AOD sat ) AOD non-dust Where is the bias correction GEOS-Chem, CESM, WRF-Chem and MERRAERO ( ) Regional Dust AOD Regional, seasonal dust AOD ensemble with uncertainty Methodology to retrieve Dust AOD Model dust AOD only used to scale from regional to global dust AOD 3 satellite retrievals, 4 models, 5 years of daily data Satellite AOD AERONET AOD AOD bias correction Model dust AOD Satellite AOD Model non-dust AOD Dust AOD Model AOD Non-dust AOD Dust AOD Daily, Gridded Seasonal, Gridded PDF Seasonal, Regional PDF Seasonal, Global PDF Model Observation s GEOS-Chem CESM WRF-Chem MERRAERO MISR MODIS Aqua MODIS Terra Global Dust AOD ensemble Multiple satellite-model combination estimates of dust AOD Satellite-retrieved dust AOD uncertainty Satellite dust AOD distribution better constrained (7%) Satellite dust AOD greater than 13 out of 14 AEROCOM models (but 5 within 1 s.d.) Two models in this study within 1 s.d. of satellite estimate Observational estimate of dust AOD greater than most models (0.035 0.008) Dust AOD Derived from Huneeus et al. (2010) KDE probability distribution mean Comparison With Model Dust AOD Fractional differences are more useful to assess models Satellite Model MISR TERRA AQUA GEOS CESM WRF MERRA DJF MAM JJA SON Fractional dust AOD MISR TERRA AQUA GEOS CESM WRF MERRA Global seasonal dust AOD Relative regional distribution of dust AOD Slight high bias in African dust AOD at the expense of Asian dust AOD Africa Asia Middle East +10% -5% +2% -5% +3% +4% -3% -1% +1% -2% +1% CESM MERRAERO GEOS-Chem WRF-Chem Observational estimate Atlantic dust transport Satellite dust AOD suggests model dust export from Africa is weak Models are biased % low relative to satellite estimate in all seasons except JJA (23%) Dust lifetime likely too short in all models, especially in African outflow Africa Atlantic Ridley et al. (JGR, 2012) Satellite dust AOD over Asia higher than models Model dust AOD 30-60% lower, especially in winter (and CESM) Limited observations close to Gobi and Taklamakan deserts AERONET suggests more dust AOD than inferred from sites further downwind Lanzhou City Beijing Gobi Taklamakan AERONET AOD Year Month AERONET Angstrom Exp. Month Year Models appear to underestimate dust in Asia in winter Fractional dust AOD A new benchmark for model dust emissions Observationally-constrained dust AOD estimate developed Global dust AOD estimated at % higher than AEROCOM ensemble median (0.023) Systematically higher than most models Will act as a benchmark for model dust AOD on a seasonal and regional basis An Observationally-Constrained Global Dust Aerosol Optical Depth (AOD) DAVID A. RIDLEY 1, COLETTE L. HEALD 1, JASPER F. KOK 2, CHUN ZHAO 3 1. CIVIL AND ENVIRONMENTAL ENGINEERING, MASSACHUSETTS INSTITUTE OF TECHNOLOGY 2. DEPARTMENT OF ATMOSPHERIC AND OCEANIC SCIENCES, UCLA 3. ATMOSPHERIC SCIENCES AND GLOBAL CHANGE DIVISION, PACIFIC NORTHWEST NATIONAL LAB AMERICAN GEOPHYSICAL UNION, FALL MEETING, 2015 Dust AOD seasonality Satellite sampling frequency represents the clear-sky dust AOD (more so for MODIS) Fractional dust AOD MISR TERRA AQUA GEOS CESM WRF MERRA Fractional dust AOD Middle East Africa MISR TERRA AQUA GEOS CESM WRF MERRA