Detecting cloud vertical structures from photon path lengths distribution

Photon path length distribution is sensitive to three-dimensional cloud structures. The first two moments of photon path length distribution have enough sensitivity to distinguish multi-layer clouds from single-layer clouds. A simple uniform approach, based on the photon diffusion theory, is developed to diagnose cloud vertical layer information, using retrieved photon path length information from advanced rotating shadowband spectrometer (ARSS). This approach is then verified through thousands of cloud cases simulated in Monte Carlo model, which shows it can nearly distinguish all of the multi-layer clouds from single-layer clouds except for those multi-layer clouds having very thin upper layer that is close to the lower layer. The approach is also compared with cloud radar measurements at the ARM Southern Great Plains (SGP) site. The consistent results of identifying cloud type between radar measurements and the photon path length approach show the feasibility of our diagnose method. More importantly, some case studies show that the radar failed to detect some of the multi-layer clouds that can be detected by the photon path length distribution approach. In the analysis of the entire year measurements of 2000 at the SGP site, we find that about 27.67% of single-layer clouds detected by cloud radar could actually be multi-layer clouds even when we use a relatively strict threshold. For those cases, the additional cloud layers are either outside the field of view of the cloud radar or too thin to be detected by the cloud radar. Those “missed” clouds, however, could have significant impacts on broadband heating rate profiles.