An Automated Algorithm for Detection of Hydrometeor Returns in Micropulse Lidar Data

Abstract

A cloud detection algorithm for a low power micropulse lidar is presented that attempts to identify all of the significant power returns from the vertical column above the lidar at all times. The main feature of the algorithm is construction of lidar power return profiles during periods of clear sky against which cloudy-sky power returns are compared. This algorithm supplements algorithms designed to detect cloud-base height in that the tops of optically thin clouds are identified and it provides an alternative approach to algorithms that identify significant power returns by analysis of changes in the slope of the backscattered powers with height. The cloud-base heights produced by the current algorithm during nonprecipitating periods are comparable with the results of a cloud-base height algorithm applied to the same data. Although an objective validation of algorithm performance on high, thin cirrus is lacking because of no truth data, the current algorithm produces few false positive and false negative classifications as determined through manual comparison of the original photoelectron count data to the final cloud mask image.