Child pages
  • User Question - AIRS Level 2 data
Skip to end of metadata
Go to start of metadata


Hello! I am a graduate student and I am doing research about the climate of the Tibet plateau. I need the first guess 
O3 data of AIRS Level 2 data. I want to know the meaning of the Data Fields named 'O3CDInit' and why the 'O3CDInit' in my data is '-9999'.


  • No labels


  1. The data field “O3CDInit” is left over from an older version of the data and is no longer used. It is just 
    filled up with all missing values of -9999. Now the first guess of O3 is from the UARS climatology. It is provided for each of 17 latitude zones for
    each month and for the 100 pressure layers in the Level 2 Support Pressure array. The latitude zones are
    at 10° intervals from 80S to 80N.
  2. The first guess ozone in AIRS v6 is the McPeters, Labow, and Logan (2007) climatology, interpolated in 
    latitude but not in time. We might have written this out on a spot-by-spot basis for the user since
    we had the available O3CDInit variable, which used to hold the output from the regression which we
    no longer do, but we didn't get this into the output product in the v6 delivery. This oversight has
    been corrected in the prototype Version 7. If it is documented as "regression ozone" or "ozone output from initial retrieval", -9999 is
    correct as we don't calculate that regression in v6.
    Here is the citation for the climatology:
    Ozone climatological profiles for satellite retrieval algorithms
    Article in Journal of Geophysical Research Atmospheres 112(D5) · March 2007
    DOI: 10.1029/2005JD006823
    I hope the user is able to use these values for his data month to linearly interpolate in latitude 
    to find the first guess profile.
  3. Here is the additional relevant material we can collect for your work: 
    Inside the retrieval everything is defined on the 100 pressure levels (0.016 to 1100). 
    Internally temperature is on LEVELS, ozone output is provided on both layers and levels,
    but the internal format is on LAYERS, with column density reported at a level being the
    LAYER ABOVE value. A perturbation trapezoid function in the physical retrieval is defined on the 100 pressure
    levels by defining the shape using a few of the fixed levels. If your set is [1 26 35 39 44 49 56 63 80]
    then the second trapezoid starts at level 1 with value zero, rises to value 0.5 at level 26, stays at
    0.5 through level 35, then falls to zero at level 39. Each trapezoid is like that, defined by
    four consecutive numbers in the set. There should be an extra number added to the end of the set
    which is the surface pressure point, for some reason that was not chosen to be written out in this array. ____ / \ / \ ab cd (very bad graphics for levels a,b,c,d) The first and last trapezoid are special because they are defined by only three points and are
    not really trapezoids. The first function has value 0.5 at level 1, constant value 0.5 to
    level 26, then falling to value 0 at level 35. The bottom one has value zero at level 63, rising
    to value 0.5 at level 80, then constant value 0.5 to the surface. Temperature trapezoids work the same way, with two exceptions. The bottom temperature function
    continues to increase to value 1.0 at the surface instead of to 0.5, so it looks more like a
    triangle. Ozone perturbations represent a MULTIPLICATIVE perturbation (or linear in log(O3)),
    while temperature perturbations represent a LINEAR perturbation in temperature. In all trapezoid cases, if the surface is too high elevation, some trapezoids are omitted
    (pressure level set will change).