Automatic detection of boundary layer height from Doppler lidar using K-means algorithm.

Atmospheric boundary layer height is a parameter of first interest for both research and operational issues. The High Resolution Doppler Lidar from NOAA can measure both backscatter and wind profiles. From these data, this work propose to derive a single estimate of boundary layer height that can deal with unavailability of one type of data. This is done thanks to the K-means algorithm. It has been compared with 2 other methods and it is shown to have large success and availability rates