Molecular line absorption in a scattering atmosphere. Part III: Pathlength characteristics and effects of spatially heterogeneous clouds

Abstract

The paper analyzes the influence of horizontal variability of clouds on sunlight reflected in a narrow portion of the solar spectrum and how this influence affects the ability to estimate cloud properties from measurements of reflection. This paper is part of a series that examines the use of high-resolution measurements of absorption lines of the oxygen A band located between 0.763 and 0.773 μm as a way of deriving information on cloud structure presently unobtainable from other passive measurements. The effects of spatial heterogeneity on reflectances and pathlength distributions are examined for marine stratocumulus cloud fields derived from Landsat data. The results showed that for the marine stratocumulus fields studied, the radiance errors were on the order of 3%-20%, and the spectral radiance ratio (in-band to out-of-band) errors were on the order of 1%-5%. When these errors are too large, the retrieved quantities possess too much error to be useful. The errors due to horizontal transport of photons and subpixel variability are studied independently as a function of spatial scale. Using both the radiance and radiance ratios, a technique was developed that allows the plane-parallel forward model typically used in retrieval schemes to be able to predict when its retrievals are so influenced by spatial heterogeneity that the results are invalid. This ability would represent a significant step forward in the current abilities of passive retrievals, which cannot determine the effect of cloud spatial variability on their retrievals. Lastly, a method was demonstrated that used domain-averaged measurements of absorption formed by reflection along with plane-parallel theory to estimates the distribution of optical depth throughout the domain. The results for the four simulated cloud fields showed this technique to have significant promise in quickly classifying the level of cloud heterogeneity over a large area.