References:

1
NO2 and Aerosol Absorption Needed for
Atmospheric Correction from New Ground-Based
Direct-Sun Spectrometers Comparison with
OMI/AURA Retrievals   J. Herman2, A. Cede 1,2, E.
Spinei 4, N. Abuhassan 5,2, B. Bojkov 6,2, G.
Mount 3 1 UMCP ESSIC, 2 GSFC, 3 WSU, 4 CALTEC, 5
SSAI, 6 UMBC GEST
WHY DIRECT SUN? All remote sensing techniques
measure slant column amounts of atmospheric trace
gases (here NO2). The slant column gives the
amount of NO2 along the average path photons
take to get through the atmosphere into the
instrument, called the air mass factor (AMF).
Dividing the slant columns by the AMF gives the
vertical column, which is the quantity we are
interested in. For scattered light measurements
(e.g. from satellite), the air mass factor
depends strongly on the vertical NO2
distribution, surface albedo and aerosols, and is
the largest uncertainty source in the retrieval.
For direct-sun measurements, the air mass factor
is easily computed and only marginally depends on
atmospheric parameters or surface reflectivity,
thus significantly reducing the uncertainty in
the retrieval. MFDOAS (WSU) The Multi-Function
Differential Optical Absorption Spectrometer
(MFDOAS), from Washington State University, is a
CCD spectrograph for direct sun and sky radiance
measurements. NO2 columns are derived using the
405-430nm window using cross sections for 295K
based on Harder et al. 1997. MFDOAS is a large
(2 meters high and 100 kg) spectrometer capable
of very high percision and good accuracy. As
such, it is the reference instrument for the
small (4 kg) Pandora spectrometer
systems. PANDORA (GSFC) The Pandora-1
spectrometer (PANDORA) measures direct-sun
irradiances from 270 to 500nm at
0.5nm resolution. The outdoor head sensor is
mounted on a small tracking system that holds a
single strand fiber optic cable, which collects
the light passed through a collimator (1.6 FWHM
field of view) and a filterwheel. The other
end of the fiber is inside a building and
connects to a 75mm focal length spectrometer
using a 1024x1 pixel CMOS detector, stabilized to
room temperature (20C). The whole system weighs
less than 15kg. A measurement sequence
consists of averaging 20 seconds of data at an
optimized integration time for different settings
in the filterwheel open hole, low-pass filter
390nm cut off, low-pass filter 320nm cut off, and
opaque blank for the dark counts. The measured
raw counts are corrected for dark signal and
stray light and converted into count rates.
Pandora-2 is a more capable version of
Pandora-1for measuring both direct sun and sky
radiances using a 2048x16 backthinned CCD to
obtain trace gas amounts and their altitude
profiles, aerosol optical depth and absorption,
and ozone profiles. Both are low cost (less than
10K for hardware). CLEO (GSFC) CLEO is designed
as a low-cost (6K) spectrometer version (300 nm
to 900 nm with 1nm resolution) of our
commercially available 7-filter (305 nm to 443
nm) GSFC modified shadowband radiometer MSBR.
The algorithms for determining aerosol optical
depth and single scattering albedo (absorption)
are mature Cede et al., 2006 Krotkov et al.,
2004. The biggest differences between CLEO and
MSBR are the spectral range and resolution, and
CLEOs much higher signal-to-noise ratio (SNR gt
10001). The spectrometer and a golf ball are
shown in the inset. The key reasonfor developing
CLEO is to be able to obtain the absorption
optical properties of aerosols as a continuous
spectrum from the UV to the visible (see example
in next column from data obtained in Sant Cruz,
Bolivia). The resulting absorption spectra can be
used to distinguish black carbon aerosols from
absorbing hydrocarbons and sulfate aerosols.
These properties are needed for atmospheric
corrections, especially in coastal regions. The
current operating version shown in the picture is
a prototype. The final version is much smaller
with the band motor located inside the black
tube. OMI (KNMI, Holland) The Ozone Monitoring
Instrument (OMI), one of four instruments on
board of the AURA satellite, retrieves many
atmospheric parameters from space Levelt et al.,
2005. Here we present Level 2 of total NO2
columns Bucsela et al., 2006, which are derived
using the 405-465nm window with 0.5 nm resolution
and the cross sections from Vandaele et al.
1998. Two CALIBRATION Both MFDOAS and PANDORA
measure relative slant columns, i.e. the NO2
slant column amount in the measured spectrum
minus the NO2 slant column amount in the
reference spectrum. The obtain absolute slant
columns, the slant column amount in the reference
spectrum has to be known. Here we used as a
reference spectrum (for all campaigns at
Thessaloniki, Goddard, and Table Mountain) a
measurement taken around noon on July 7, 2007 at
Table Mountain, California. We have developed two
techniques for field calibration of Pandora 1)
Bootstrap for polluted regions and 2) Modified
Langley for very clean sites. These techniques
permit absolute calibration to better than half
the stratospheric value of NO2 or 0.1
DU. COMPARISON WITH OMI The figures show the
retrieved total vertical column amounts of MFDOAS
(green), PANDORA (blue), FTUVS (black), and OMI
(red) during the 3 campaigns at Thessaloniki,
Goddard Space Flight Center (GSFC) and JPL-Table
Mountain. 1 Dobson Unit (DU) corresponds to a
column density of 2.687x 1016 particles per cm2.
JPL Table Mountain Facility, CA, USA 34.38N,
117.68W
Table Mountain, MFDOAS and PAN-1 data at running
1min averages calibrated with bootstrap method
(July 2007). Also shown are the OMI overpass data
(both total column and the estimated
stratospheric value). There is no PAN-1 data on
July 12. FTUVS is a JPL Fourier Transform
spectrometer permanently located at TMF
Santa Cruz, Bolivia 17 45S 63 14'W
Absorption optical depth (left) and SSA (right)
of aerosols from a campaign in Bolivia using our
MSBR. Notice the deviation from the AERONET
extrapolation. CLEO will obtain the spectrally
resolved values from 300 to 900 nm. CLEO is
currently deployed at GSFC.
SHADOWBAND
SHADOWBAND

• CONCLUSION
• The combination of the two ground-based
  direct-sun instruments PANDORA and MFDOAS
  provides very accurate retrievals of total
  vertical NO2 columns.
• The ground-based NO2 data cover the entire day
  at high time resolution. Therefore, they can be
  used for an accurate atmospheric correction, to
  monitor air pollution. and to validate
  tropospheric chemistry models.
• The ground-based measurements have been compared
  to OMI retrievals at a remote, a suburban, and an
  urban site. The agreement is very good. Most
  situations with large differences between ground
  and satellite data explained by the satellite
  view not being directly over the ground station.
• OMIs relatively small footprint is capable to
  pick up regional plumes with enhanced NO2 (see
  e.g. May 25, 2007, at GSFC).
• The spatial variability of NO2 at a scale smaller
  than one OMI pixel is still unknown. The enormous
  temporal variability of NO2 at a polluted site
  like GSFC (see figures above) suggests that the
  spatial variability must also be very large.
• Weekly or monthly measurement campaigns with
  ground-instruments at one single location do not
  necessarily provide enough information to
  validate satellites. Even at rural, unpolluted
  sites the statistics are to small (one or two
  satellite overpasses a day) to draw any
  conclusion. Therefore monitoring records at least
  longer than 1 year are needed.
• Goddard and Thessaloniki are relatively flat, but
  are near NO2 sources, giving rise to high
  temporal and spatial variability in NO2 amounts.
• Future location of these instruments near coastal
  areas will validate or correct the satellite
  measurements needed to obtain an accurate
  atmospheric correction for NO2 absorption in the
  blue and UV wavelengths.
• Because of the high temporal variation of NO2
  and aerosols, ground-based measurements are
  needed to provide corrections to the OMI 130 pm
  retrievals of NO2 and aerosols.
• The shadowband aerosol absorption measurements
  show a clear deviation from the extrapolations of
  AERONET as applied to the blue and UV
  wavelengths.
• Accurate measurements of absorbing aerosol
  optical properties are needed to complete the
  atmospheric correction for retrieval of water
  leaving radiances in current and future
  satellites. This has been demonstrated using a
  commercial filter shadowband instrument and will
  be extended using CLEO, a new low-cost (6K)
  spectrometer version of the shadowband
  instrument.
• The use of very accurate inexpensive instruments
  means that it is practical to widely deploy the
  Pandora and CLEO systems at sites of interest
  within cities, coastal regions, and onboard ships.

• References
• - Bucsela, E.J., E.A. Celarier, M.O. Wenig, J.F.
  Gleason, J.P. Veefkind, K.F. Boersma, and E.
  Brinksma, Algorithm for NO2 vertical column
  retrieval from the Ozone Monitoring Instrument,
  IEEE Trans. Geosc. Rem. Sens., 44 (5),
  1245-1258, 2006.
• Burrows, J. P., A. Richter, A. Dehn, B. Deters,
  S. Himmelmann, S. Voigt, and J. Orphal,
  Atmospheric remote- sensing reference data from
  GOME - 2. Temperature-dependent absorption cross
  sections of O3 in the 231-794 nm range, J.
  Quant. Spectrosc. Radiat. Transfer, 61 (4),
  509-517, 1999.
• Harder, J. W., J. W. Brault, P. V. Johnston, G.
  H. Mount, Temperature dependent NO 2 cross
  sections at high spectral resolution, J.
  Geophys. Res., 102(D3), 3861-3880,
  10.1029/96JD03086, 1997.
• - Hermans, C., A.C. Vandaele, B. Coquart, A.
  Jenovrier, and M.F. Merienne, Absorption bands of
  O2 and its collision induced absorption bands in
  the 3000-7500cm-1 wavenumber region, in
  Proceedings of the International Radiation
  Symposium, St. Petersburg, Russia, 24-29 July
  2000, edited by W.L. Smith and Y.M Timofeyev, A.
  Deepak, Hampton , Va, 639-642, 2001.
• - Kurucz, R.L., Synthetic infrared spectra, in
  Infrared Solar Physics, IAU Symp. 154, edited by
  D.M. Rabin and J.T. Jefferies, Kluwer, Acad.,
  Norwell, MA, 1992.
• - Levelt, P. F., E. Hilsenrath, G. W.
  Leppelmeier, G. B. J. van den Oord, P. K.
  Bhartia, J. Tamminen, J. F. de Haan and J. P.
  Veefkind, Science objectives of the Ozone
  Monitoring Instrument, IEEE Trans Geo. Rem.
  Sens., 2005.
• Newnham, D. A., J. Ballard, Visible absorption
  cross sections and integrated absorption
  intensities of molecular oxygen (O2 and O4), J.
  Geophys. Res., 103 (D22), 28801-28816,
  10.1029/98JD02799, 1998.
• Nizkorodov, S. A., S. P. Sander, and L. R.
  Brown, Temperature and Pressure Dependence of
  High-Resolution Air- Broadened Absorption Cross
  Sections of NO2 (415-525 nm), J. Phys. Chem. A,
  108 (22), 4864 -4872, 2004.
• - Rothman, L.S., The HITRAN data base, J. Quant.
  Spectrosc. Radiat. Transfer, 48, 5, 6, 1992.
• - Vandaele, A.C., C. Hermans, P.C.Simon, M.
  Carleer, R. Colin, S. Fally, M.F. Mérienne, A.
  Jenouvrier, and B. Coquart, Measurements of the
  NO2 absorption cross-section from 42 000 cm1 to
  10 000 cm1 (238-1000 nm) at 220 K and 294 K, J.
  Quant. Spectrosc. Radiat. Transfer, 59 (3-5),
  171-184, 1998.