Abstract
Cloud absorption inferred from the difference between the net fluxes measured by stacked aircraft below and above clouds is strongly affected by the uncertainties due to cloud horizontal inhomogeneity. the simplest way to get rid of these uncertainties is to perform grand averages over flight legs; if flight legs are long enough, grand averaging may lead to a reliable estimate of cloud absorption. However, the amount of information on βtrueβ cloud absorption returned from such an expensive measurement program will be very limitedβoften one number per flight leg.
This paper contains a discussion on how to enhance the harvest of true absorption data using two related methods: (a) subtraction and (b) conditional sampling. Both methods assume that, simultaneously with broadband measurements, some narrow nonβabsorbingβband net flux measurements are also available. Both methods are related to AckermanβCox type corrections, where subtracting fluxes in a transparent spectral band from those in an absorbing band partially removes the radiative effects of horizontal inhomogeneity and allows the recovery of spatially resolved cloud absorption. the output of the two methods is different: while the subtraction method provides a contiguous record of recovered cloud absorption, the conditional sampling method yields a discrete set of data points where the vertical net flux divergence reliably estimates true cloud absorption.