Surface Reflectivity from OMI: Effects of Snow on OMI NO2Gray O’Byrne1, Randall Martin1,2, Aaron van Donkelaar1, Joanna Joiner3, Edward A. Celarier3

1Dalhousie University 2Harvard-Smithsonian Center for Astrophysics 3NASA Goddard Space Flight Center

Locating Cloud and Aerosol filtered OMI scenes

We use the MODIS/Aqua cloud mask to determine the presence of clouds within the OMI field of view. Using MODIS to screen for clouds ensures that an OMI scene is cloud free even when surface reflectivity is unknown. We account for horizontal displacement of the clouds during the time between the MODIS and OMI overpass (~12 minutes). We repeat this analysis for scenes with AOD>0.2 using the MODIS AOD product.

Snow-Free Surface Reflectivity

0.1

0

0.2

We filter the Lambertian-Equivalent Reflectivity (LER) retrieved from OMI [Joiner and Vasilikov, 2006] to exclude clouds and aerosols as determined by MODIS. Deserts are more reflective then vegetation and ocean. White space indicates persistent cloud.

Surface Reflectivity of Snow Covered Lands (LER)

This is the surface reflectivity we measure for scenes that are both cloud free (as determined by MODIS) and snow covered, as flagged in the OMI product according to the NISE dataset. Surface reflectivity over snow varies considerably from the value of 0.6 that is typically used in current OMI retrievals.

Annual Mean

Surface Reflectivity

(LER)at 354nm

We apply this database to evaluate previous climatologies of surface reflectivity. Except for regions of seasonal snow cover, agreement is best with a climatology from OMI [Kleipool et al., 2008] which selects the surface reflectivity from a histogram of observed LER. Three other climatologies of surface reflectivity from TOMS

[Herman & Celarier, 1997] , GOME [Koelemeijer et al., 2003] and OMI [Kleipool et al., 2008], based on minimum observed LER, are less consistent with our cloud- and aerosol-filtered dataset.

Effect on OMI NO2 – Challenges with Clouds and Snow

Here we compare the OMI NO2 product [Bucsela et al., 2006] over the cities of Calgary and Edmonton (in the highly reflective region in South-Central Canada) for three different snow-on-ground categories. The current fixed surface reflectivity for snow used in OMI cloud and NO2 retrievals leads to trends of

increasing NO2 with snow and cloud. Errors in the a priori surface

reflectance will introduce errors in the OMI (O2-O2) cloud fraction retrieval, so we make the distinction between reported and real cloud fractions.

The top panel shows the bias between the OMI NO2 columns for cloudless scenes

(determined by MODIS) retrieved with our snow-covered surface reflectivity versus columns calculated with the surface reflectivity (and cloud data) used in the current OMI NO2 product. The error in the

a priori surface reflectance leads to non-zero cloud fractions being reported for these cloudless scenes. In practice scenes with high cloud fractions are often masked to ensure good sensitivity to the boundary layer. The bottom panel shows the bias once scenes with OMI cloud fractions greater then 0.3 are removed. Most positive biases are removed, but negative biases remain unchanged.

This research was supported by NASA and by the Canadian Foundation for Climate and Atmospheric Science. The authors would like to acknowledge Jim Gleason and Terry O’Byrne for their helpful comments.References:•Bucsela, E. J., E. A. Celarier, M. O. Wenig, J. F. Gleason, J. P. Veefkind, K. F. Boersma, and E. J. Brinksma (2006), Algorithm for NO2 vertical column retrieval from the ozone monitoring instrument, IEEE Trans. Geosci. Remote Sens., 44(5), 1245-1258.•Herman, J. R. and E. A. Celarier (1997), Earth surface reflectivity climatology at 340-380 nm from TOMS data , Journal of Geophysical Research-Atmospheres, 102(D23), 28003-28011. •Joiner, J. and A. P. Vasilkov (2006), First results from the OMI rotational Raman scattering cloud pressure algorithm , IEEE Trans. Geosci. Remote Sens., 44(5), 1272-1282. •Kleipool, Q. L., M. R. Dobber, J. F. de Haan, and P. F. Levelt (2008), Earth surface reflectance climatology from 3 years of OMI data , Journal of Geophysical Research-Atmospheres, 113(D18). •Koelemeijer, R. B. A., J. F. de Haan, and P. Stammes (2003), A database of spectral surface reflectivity in the range 335-772 nm derived from 5.5 years of GOME observations, Journal of Geophysical Research-Atmospheres, 108(D2), doi: 10.1029/2002JD002429 ER.

Retrievals of NO2 from the OMI/Aura satellite instrument are being widely applied to improve understanding of air quality and NOx emissions. OMI NO2 retrievals depend on information about surface reflectivity. We use observations from the MODIS/Aqua satellite instrument, which flies 12 minutes ahead of OMI, to exclude clouds and aerosol from OMI observations and determine surface reflectivity for clear-sky conditions. The resultant dataset is used to evaluate surface reflectivity inferred from other techniques, and to assess the implications for OMI NO2 retrievals over snow.

Summary

0.5

0

1

MODIS Cloud Mask

Potential

Transport

Cloud Free OMI Scenes!

Clouds in Red

OMI Reported Cloud Fraction

≥ 5cm of snow

0 > snow < 5cm

no snow

Mea

n T

rop

. NO

2 (m

ole

c/cm

2 )

Winter OMI NO2 over Calgary & Edmonton

Previous Reflectivity Climatologies

Mean Difference

Standard Deviation

OMI 0.0002 0.011

OMI Mininum -0.002 0.033

GOME Mininum 0.012 0.026

TOMS Mininum -0.008 0.022

corrected

correctedoriginal Bias NO Relative 2

With CloudFractionThreshold (f < 0.3)

-0.5 0 1.0

OMI NO2 for Snow-Covered Scenes

Is Minimum Best?