由于大气中气溶胶垂直分布复杂,常用的梯度法和小波协方差变换法在边界层高度自动连续识别方面仍具有较大的不确定性。选取在浙江金华两个时间段内观测得到的双波长激光雷达数据,将二维矩阵方法应用于复杂大气垂直分布情况下边界层高度的反演,从距离和时间两个维度优化边界层高度的反演结果。研究结果表明,在气溶胶垂直分布出现多层结构的情况下,小波协方差变换法反演的边界层高度有较大的误差,而二维矩阵方法具有一定的优势。在532 nm波长基于二维矩阵方法和小波协方差变换方法反演的边界层高度与1064 nm波长基于小波协方差变换方法反演的边界层高度的相关系数分别为0.87和0.37。边界层高度与地面温度变化趋势一致,表明边界层高度的反演结果是可靠的。二维矩阵方法为进一步改善无云情况下(3 km以内无云层出现)激光雷达自动连续识别边界层高度可靠性提供了很好的借鉴与参考。

Due to the complex vertical distribution of aerosols in the atmosphere, the commonly used gradient method and wavelet covariance transform (WCT) method still have large uncertainties in automatically and continuously identifying the planetary boundary layer height (PBLH). For this reason, we applied the 2-D matrix method to the retrieval of PBLH in two cases observed by a dual-wavelength lidar over Jinhua city. The 2-D matrix method optimizes PBLH retrieval results from the distance and time dimensions. The results show that in the case of profiles with multi-layer aerosols, the 2-D matrix method is superior to the WCT method in terms of PBLH retrieval. The correlation coefficient between the PBLH retrieved by the 2-D matrix method at 532 nm and that by the WCT method at 1064 nm is 0.87. However, the correlation coefficient is only 0.37 if the 2-D matrix method is replaced by the WCT method at 532 nm. Additionally, the consistent variation trend in PBLH and surface temperature indicates reliable PBLH retrieval results. The 2-D matrix method provides a reference for the automatic and continuous identification of the PBLH by lidar in the absence of low clouds (no clouds appearing below 3 km).