微脉冲激光雷达水平探测气溶胶两种反演算法对比与误差分析

大气气溶胶对人类健康、环境和气候系统具有重要的影响。微脉冲激光雷达(MPL)是一种新型的探测大气气溶胶水平分布的有效工具，而消光系数的反演和误差分析是其数据处理的重要内容。为了探测近地面大气气溶胶的水平分布情况，采用分段斜率法和Fernald算法对合肥地区实测MPL水平数据进行了消光系数反演，并将反演结果进行了对比和误差分析。误差分析表明，分段斜率法和Fernald算法的误差分别主要来源于其理论模型和多个假设条件。虽然分段斜率法和Fernald算法应用于大气水平消光系数的反演都仍存在一定的问题，精确度有待提高，但都能在一定程度上反映出气溶胶粒子的时空分布特征，与前向散射能见度仪的测量结果相关性都能达到95%以上，具有一定的可行性。且相对来说，Fernald法更适用于大气非均匀分布情况下消光系数的反演。

Two Data Inversion Algorithms of Aerosol Horizontal Distribution Detected by MPL and Error Analysis

Atmospheric aerosols have important impacts on human health, the environment and the climate system. Micro Pulse Lidar (MPL) is a new effective tool for detecting atmosphere aerosol horizontal distribution. And the extinction coefficient inversion and error analysis are important aspects of data processing. In order to detect the horizontal distribution of atmospheric aerosol near the ground, slope and Fernald algorithms were both used to invert horizontal MPL data and then the results were compared. The error analysis showed that the error of the slope algorithm and Fernald algorithm were mainly from theoretical model and some assumptions respectively. Though there still some problems exist in those two horizontal extinction coefficient inversions, they can present the spatial and temporal distribution of aerosol particles accurately, and the correlations with the forward-scattering visibility sensor are both high with the value of 95%. Furthermore relatively speaking, Fernald algorithm is more suitable for the inversion of horizontal extinction coefficient.