AUTOMATIC METHOD FOR DETERMINING THE MIXING DEPTH FROM LIDAR OBSERVATIONS

Lidar observations of the atmosphere reveal layers having different concentrations of aerosols. A computer method has been developed to analyze digital records of lidar backscatter in a manner that reveals the presence and heights of significant aerosol layers and the depth of the mixed layer. Boundaries between layers are represented by vertical gradients in the backscatter data. To compute the mixing depth, the backscatter gradients for a given time period (e.g. 10 min) are treated in a two-dimensional height-time array. Positive gradients and negative gradients are organized into groups that are contiguous in space and time. Then an algorithm selects one group as representing the mixing depth, on the basis of the height of each group and the strength of the gradients. Mixing depths determined automatically from the lidar records agreed with subjective estimates of mixing depth (based on data from lidar, radiosondes, and helicopter temperature soundings) 91% of the time for data collected in St. Louis, Missouri during the RAPS program of July and August, 1976. A slightly altered version of the automatic method, applied to lidar observations from the October 1977 AMBIENS experiment in Indiana, was 97% accurate. Real-time application of the method is believed to be feasible. © 1979.