6_RemoteSensing_mount by xiuliliaofz


									Ground and Satellite Observations of
    Atmospheric Trace Gases

          George H. Mount
 Laboratory for Atmospheric Research

            13 April 2007
• Aura/OMI - Ozone Monitoring
  Instrument on the Aura satellite
• Aura ground truth/validation of data
• MFDOAS instrument for urban airshed
  trace gas measurements and satellite
• NASA INTEX B results, PNNL, spring
• WSU involvement began in 1997 at the inception
  of the OMI project - Dutch instrument on NASA
  bird, small US team
• double channel spectrograph covering 270 - 510
  nm at 0.7nm resolution
• launched in July 2004
• measures column O3, NO2, BrO, OClO, CH2O,
  SO2, aerosol indices
• data products:
  –   atmospheric column of above gases
  –   trop NO2
  –   trop O3 - not yet routinely available
  –   aerosol data
 trace gas measurements: observing the
Earth’s backscattered uv/visible radiation


              Sunlight passes through the atmosphere, reflects
              off clouds and the surface, and is scattered back
              into the instrument field of view. Molecular spectral
              absorption is proportional to the concentration of
              the gas doing the absorbing along the path.
     Observing Principle for OMI

                              2-dimensional CCD

                   ~ 580 pixels
                                             ~ 780 pixels

flight direction
» 7 km/sec
                                   viewing angle
                                   ± 57 deg

                                  12 km/24 km (binned & co-added)

                                       13 km
                     2600 km           ))



   1                TOMS

single pixel size for four satellite instruments
                                           OMI single pixel

OMI pixel 12 km x 13 km superposed onto the Seattle airshed (zoom mode)
            New WSU MFDOAS Instrument
•   WSU was funded 3 years ago to develop a new ground based
    instrument that would mimic the satellite measurements from the
    ground & support validation of the Aura satellite data from the ground
•   new instrument uses the molecular spectrum of the sky and direct sun
    as light sources for measurement of urban air pollution
•   scans the sky at low elevation angles where the tropospheric air mass
    is significantly enhanced
•   completing development at WSU as we speak
•   ground based campaign at NASA Goddard Space Flight Center 7-21
•   ground based campaign at NASA Jet Propulsion Laboratory 1-15 July
•   was fielded in prototype form during the NASA INTEX at Pacific
    Northwest National Laboratory in central Washington spring 2006
                          MF-DOAS: geometry
                          for sky-viewing mode

stolen from Platt group
MFDOAS instrument
data during
PNNL, Richland,
OMI data
during INTEX
over PNNL, WA
spring 2006

comparison with
MFDOAS shows a
20% bias between
with OMI
underestimating NO2
column. Other
validations show a
Things to consider in using satellite data:

• basic limitations:
    • can only observe when clear - problem in Pacific NW in winter
    • tied to equator crossing time (1345h for Aura) for a polar orbiter
    • cannot get more than a couple of orbits of data before the urban area rotates
          out of the FOV - orbital period ~ 90 minutes
    • cannot observe at night for most instruments
    • accuracy of the trop result depends on removing the strat overburden (if exists)

• trace gases:
     • ozone and NO2 total columns are done well from space
     • problem is the stratospheric overburden --> must separate strat/trop
         • this is not an easy problem - cloud slicing, use of other instruments onboard
         • RT codes have advanced this a lot in the last 5 years (spherical codes)
     • CO can be done fairly easily in the IR part of the spectrum
     • SO2 and HCHO are very hard due to low levels in the trop - few ppbv sensitivity
     • aerosols are measured by using uv radiances and work well

• footprint size and model grid size
    • this has followed the technology with steady evolution to smaller footprints
    • 2-D detectors --> large advances in footprint and simultaneous large swath angle
    • OMI: 114° = 2600 km swath with 13 km pixels and a spectrum at central pixel
    • scanning mirrors are a thing of the past for swath coverage (perp. to orbital track)
• temporal resolution
    • orbital period of ~ 90 min --> get a picture each period until airshed
         rotates out of the swath - larger swath angles are better --> information
         on multiple orbits
    • may take several days to build up an image due to spatial scanning swath
    • satellite moves at 7 km/sec --> ground track always moves at that speed
         • integration time to get good s/n - sets footprint in velocity vector
                  direction (NS)
         • short int. times produce low s/n, especially in the Huggins O 3 bands
    •time to get complete global coverage - e,g, GOME 3d
to Farren’s talk
Spectroscopic Technique for Air Pollution Measurements
• measure the absorbed spectrum of sunlight reflected from the Earth’s surface
• many tropospheric trace gases have complicated molecular spectra allowing
    identification and quantification of concentration, e.g.

                                                                         cross section

                                                           [Harder, Brault, Johnston, and Mount, 1997]
• need a telescope to look down on the Earth
    • to collect light
    • to image the Earth surface onto the spectrograph over
        a wide field of view (e.g. for OMI: 114° = 2600 km) at high
        spatial resolution (e.g. for OMI: 12 km x 24 km pixel size in “global mode”)
• spectrograph to sort out the molecular spectra
• imaging detector which allows simultaneous detection ofa large spectral region for
          each spatial resolution element over a wide swath of geography with high
          spatial resolution. Older systems use a scan mirror to change the FOV.
• basic physics is very simple
    • solar light traverses an atmospheric path: Sun --> Earth surface, then
         reflecting Earth surface/clouds --> satellite sensor
         • measurement of spectrum of incoming light - solar irradiance spectrum
         • measurement of spectrum of light into sensor (reflected solar
              spectrum from Earth = Earth radiance spectrum)

• ratio of Earth radiance spectrum to solar spectrum
    • elimination of spectrum of solar spectrum (to first order)
    • reveals absorption spectrum of atmospheric molecules of interest
    • absorbance depth is proportional to the abundance of that molecule
         along the absorption path
         • note: this is NOT a tropospheric concentration
    • technique only useful if there is “differential” absorption from
        the molecule doing the absorbing - a smooth continuum spectrum with no
        spectral structure only depresses the entire intensity of the spectrum
Current Satellite Data - TOMS, GOME, TES, MOPITT, Sciamachy

• spatial footprint
    • currently operational satellite doing air pollution work
        • Sciamachy footprint: ~ 30 km x 60 km - scanning mirror for swath
        • GOME: ~ 80 km x 340 km - scanning mirror for swath
        • TOMS: ~ 50 km x 200 km - scanning mirror for swath
        • MOPITT: 22 km x 22 km - scanning mirror for swath - CO
        • TES: 5 km x 9 km: ozone

• temporal resolution at time of overpass (typically about 1:30PM)
    • get a picture each orbit of 90 min duration when airshed underneath
    • with a large swath, get several orbits of data sequentially over airshed
    • with small swath, may take several days to build a picture (GOME -3d)
    • time of day is restricted: depends on equatorial time transit
        • e.g. for Aura, it is 1345 h --> obsv at same time of day each day
data from PNNL
during INTEX
                            DOAS Theory

• Measure wavelength dependent light intensity (I[l]) as light passes
  through the air mass

• Initial intensity (Io [l]) decreases in the airmass due to

    q   absorption by the trace gases,
    q   scattering by molecules and aerosol particles

• trace gases can be detected in the ratio of I [l] to Io[l] as a function of
  wavelength due to their unique absorption features
            DOAS Theory: Beer-Lambert Law

• Theory

• Reality
        Air Mass Factors: Geometrical Approach

Total Slant
   Recent Developments in DOAS: an Overview. Ulrich Platt. Institut für Umweltphysik, Universität Heidelberg

To top