The performance of the OMPS/LP retrieval algorithm is assessed by conducting a series of numerical experiments and
evaluating the quality of the primary (ozone profile) and secondary products (aerosol profiles, NO2, cloud height, surface
reflectance) as well as height registration under a set of realistic atmospheric conditions selected randomly. The study
considers a number of orbits corresponding to Winter/Summer solstice and Spring/Autumn Equinox. It is shown that the
quality of the OMPS/LP retrieval products (accuracy, precision, vertical resolution, height registration) varies along the
orbit, as the single scattering angle transitions from backscatter to forward scatter and zenith angles vary from sunrise to
sunset. Instrument effects (straylight, gain consolidation, instrument noise) are also investigated. It is shown that ozone
profiles can be retrieved with an accuracy of 5% or better from the tropopause up to 50 km, a precision of about 3-5%
from 18 to 50 km, and a vertical resolution of 1.5-2 km. Stratospheric aerosol extinction profile can be retrieved with an
accuracy/precision of about 30%. The scene-based tangent height registration algorithm is shown to yield height
information with an RMS error of 250-300m.
|