Estimation of Atmospheric Boundary Layer Depth from Raman Lidar Using 2-D Spatial Filters
Document ID: 98
Master's Thesis
The Pennsylvania State University
The Graduate School
College of Engineering
Abstract
Estimation of the depth of the various layers in the Atmospheric Boundary Layer (ABL) is of interest to the meteorological, air quality and aeronautical community. While previously published literature on this topic concentrated on using wavelet techniques applied to wind profiler or lidar backscatter data, a case is presented for using Raman lidar to describe the ABL depth. The techniques introduced in this thesis complement, as well as offer an alternative to the existing technique of ABL depth estimation using wavelet techniques. The image processing techniques introduced in this thesis demonstrate a capability to detect as well as follow gradients in the ABL time series, to obtain the time series of atmospheric layer boundaries. The algorithm performs edge detection and then uses dynamic programming techniques to perform edge following. The 2-D spatial filters are used for edge detection. These filters utilize information of the current and neighboring profiles to estimate gradients in the current profile. This additional information compared to the 1-D wavelet technique yields a robust gradient detection scheme. The edge following algorithm can also be used with the wavelet technique mentioned in the published literature to obtain a time series of the boundaries between the various layers in the atmosphere. Examples are described showing the traces of multiple layers in the atmosphere with the new technique, which is now available for routine data processing.
| Citation: | S. N. Kizhakkemadam, "Estimation of Atmospheric Boundary Layer Depth from Raman Lidar Using 2-D Spatial Filters", The Pennsylvania State University, Master's Thesis, December 2002, 96 pages |