Comparison of aerosol lidar retrieval methods for boundary layer height detection using ceilometer aerosol backscatter data
- 1Department of Earth and Atmospheric Sciences, University of Houston, Houston, TX, USA
- 2Tropospheric Composition Program, Earth Science Division, NASA Headquarters, Washington, DC, USA
- 3School of Natural Sciences, Saint Edward's University, Austin, TX, USA
- 4Department of Physics, University of Oxford, Oxford, UK
- 5Joint Center for Earth Systems Technology, University of Maryland Baltimore Country, Baltimore, MD, USA
Abstract. Three algorithms for estimating the boundary layer heights are assessed: an aerosol gradient method, a cluster analysis method, and a Haar wavelet method. Over 40 daytime clear-sky radiosonde profiles are used to compare aerosol backscatter boundary layer heights retrieved by a Vaisala CL31 ceilometer. Overall good agreement between radiosonde- and aerosol-derived boundary layer heights was found for all methods. The cluster method was found to be particularly sensitive to noise in ceilometer signals and lofted aerosol layers (48.8 % of comparisons), while the gradient method showed limitations in low-aerosol-backscatter conditions. The Haar wavelet method was demonstrated to be the most robust, only showing limitations in 22.5 % of all observations. Occasional differences between thermodynamically and aerosol-derived boundary layer heights were observed.