1308 IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 45, NO. 5, MAY 2007
A Method to Determine the Spatial Resolution
Required to Observe Air Quality From Space
Christopher P. Loughner, David J. Lary, Lynn C. Sparling, Ronald C. Cohen, Phil DeCola, and W. R. Stockwell
Abstract—Satellite observations have the potential to provide
an accurate picture of atmospheric chemistry and air quality on
a variety of spatial and temporal scales. A key consideration in
the design of new instruments is the spatial resolution required
to effectively monitor air quality from space. In this paper, vario-
grams have been used to address this issue by calculating the
horizontal length scales of ozone within the boundary layer and
free troposphere using both in situ aircraft data from five different
NASA aircraft campaigns and simulations with an air-quality
model. For both the observations and the model, the smallest scale
features were found in the boundary layer, with a characteristic
scale of about 50 km which increased to greater than 150 km
above the boundary layer. The length scale changes with altitude.
It is shown that similar length scales are derived based on a to-
tally independent approach using constituent lifetimes and typical
wind speeds. To date, the spaceborne observations of tropospheric
constituents have been from several instruments including TOMS,
GOME, MOPITT, TES, and OMI which, in general, have different
weighting functions that need to be considered, and none really
measures at the surface. A further complication is that most
satellite measurements (such as those of OMI and GOME) are of
the vertically integrated column. In this paper, the length scales
in the column measurements were also of the order of 50 km. To
adequately resolve the 50-km features, a horizontal resolution of
at least 10 km would be desirable.
Index Terms—Air-quality observations, length scales,
variograms.
I. INTRODUCTION
I T IS CONCEIVABLE that many different criteria could beused to determine the spatial resolution needed to observe
air quality from space, for example, the ability to resolve the
finest structures in the trace gas fields (for aerosols or NO
2
,
this would be on the level of individual streets in a city) or
the ability to resolve the typical structures in the air-quality
fields. Since observing air quality at the scale of a few meters,
Manuscript received March 16, 2006; revised September 1, 2006.
C. P. Loughner is with the Department of Atmospheric and Oceanic Science,
University of Maryland, College Park, MD 20742 USA.
D. J. Lary is with the Atmospheric Chemistry Division, NASA Goddard
Space Flight Center, Greenbelt, MD 20771 USA, and also with the Goddard
Earth Science and Technology Center (GEST), University of Maryland
Baltimore County, Baltimore, MD 21250 USA (e-mail: David.Lary@
umbc.edu).
L. C. Sparling is with the Department of Physics, University of Maryland
Baltimore County, Baltimore, MD 21250 USA.
R. C. Cohen is with the Departments of Chemistry and Earth and Planetary
Science, Lawrence Berkeley National Laboratory, and Berkeley Atmospheric
Sciences Center, University of California, Berkeley, CA 94720 USA.
P. DeCola is with the National Aeronautics and Space Administration,
Washington, DC MD 21228 USA.
W. R. Stockwell is with the Department of Chemistry, Howard University,
Washington, DC 20059 USA.
Digital Object Identifier 10.1109/TGRS.2007.893732
necessary for street level monitoring, is likely to be beyond the
reach of space observing systems, for a while, we would like
to suggest a framework for objectively determining what the
spatial scales of typical air-quality features are. To this end, this
paper presents a variogramatic analysis which helps address
the question, “what horizontal resolution is needed to observe
air quality from space?” Since it is unlikely that there is a
universal length scale applicable for all air-quality observations,
it is useful to have a methodology for characterizing the spatial
length scales needed for a given constituent under different
conditions. We use this variogramatic approach for both in situ
aircraft data from the five different NASA aircraft campaigns
and regional model simulations at a range of resolutions.
A clear advantage of observing air quality from space is the
global coverage which provides a broader context for source
regions and allows transport away from the source regions to
be monitored. However, what is not clear is the satellite spatial
resolution required to quantitatively assess the relative impacts
of individual sources.
As an example of applying a variogramatic approach for
determining the required spatial resolutions for observing pol-
lutants, we investigate the spatial distribution of tropospheric
ozone in the polluted urban area of Los Angeles (LA) for early
November. LA is an ideal choice for our study because its
emissions are characterized, and it is a highly polluted urban
area that is afflicted by both high levels of ozone and particulate
pollution. Although ozone mixing ratios reach maximum levels
during the summertime while concentrations of particulates
maximize during the late fall and early winter, this paper is
directed toward the use of satellites to observe air quality
throughout the year. Also, one of the most important future
applications of these observations will be to improve year-
round air-quality forecasting. Accurate air-quality forecasts
must include ozone at all concentrations and not just summer-
time peak values. As shall be seen, in the LA region, the length
scales are rather short, at around 60–80 km in the atmospheric
boundary layer. Tropospheric-ozone data from several global
tropospheric experiment (GTE) missions suggest length scales
from 50 to 150 km. In the eastern U.S., on the other hand, spa-
tial correlations across ozone monitoring sites suggest longer
scales of around 500–1000 km [1]–[3]. In contrast, for urban
air-quality applications, air-quality models are typically used at
resolutions down to 4 km to account for source variability.
The natural question to address is “what resolution do we
need for satellite-borne observations of air quality?” Even
though we do not conclusively answer this question, we can
say that if the satellite measurements are to be useful, then
they should be capable of capturing the spatial scales typically
0196-2892/$25.00 © 2007 IEEE