Photon Pathlength Distributions from High Resolution Measurements of the Oxygen A-band

点击翻译按钮获取中文摘要

Solar radiation penetrating a cloudy atmosphere undergoes many scattering events. The total pathlength of photons reaching the ground may therefore be longer than the line-of-sight path. Of the photons received at a given instant, some will usually have experienced more scattering events than others. One of the ways to describe the effects of the propagation medium is by the probability distribution of photon pathlength. From radiative transfer theory, it is possible in principle to infer the pathlength distribution from measurements of the solar spectrum in the vicinity of strong oxygen absorption lines. In effect, the process is one of inverting the spectroscopic measurements to determine (or "retrieve") the unknown pathlength distribution. The accuracy of the computed distribution depends on the accuracy and resolution of the spectroscopy and on the particular inversion procedure. This grant supports the development of a high-resolution, sensitive spectrometer to measure the solar spectrum near the oxygen A-band (760 nm wavelength) and the water vapor band near 820 nm. Preliminary experiments have shown that the instrument is sensitive enough to require only a few seconds of signal integration. Hence the pathlength distributions can be measured with high time resolution, enabling the observation of rapidly varying pathlength distributions as clouds move overhead. Monte Carlo simulations have shown that a parametric inversion procedure is able to recover pathlength distributions with acceptable accuracy. Results will be analyzed to determine if multiple scattering can account for anomalous cloud absorption, which refers to the observation that the solar irradiance emerging from cloud base is less than predicted from scattering calculations. The longer the pathlength, the greater the extinction of the radiation by scattering and absorption. If the pathlengths are longer than assumed, then the calculations will underestimate the actual absorption. Another objective of the project is to develop ways of introducing pathlength distributions into climate and general circulation models as a more accurate way of characterizing the effects of clouds on radiative transfer.