引入MODIS温湿度廓线的星载激光雷达边界层高度提取算法研究

It is an important remote sensing way to retrieve boundary layer height (BLH) over a large-scale region from satellite-borne lidar due to ground monitoring stations are sparse. However, the local mutation of the spaceborne lidar signal in the non-target layer (optical thick layer and abnormal noise layer) may be much larger than that in the boundary layer, which will result in serious error in locating boundary layer height for traditional algorithm. Since temperature and humidity profiles are less disturbed by these non-target layers, these profiles can provide constraint conditions to refine boundary layer height from space-borne lidar. Based on this consideration, this project will conduct research on the synergy of MODIS atmospheric profiles and aerosol profiles from space-borne lidar with applications in regional boundary layer height retrieval. The project consists of three main parts:.1) This project will establish a method selection theory to estimate the range of boundary layer height. Then, the range can remove the influence of a non-target layer at a far distance..2) This project will explore a relationship between color/polarization ratio and boundary layer height. Then, the relationship can weaken the influence of the optical thick layer near a boundary layer top on locating boundary layer height..3) This project will develop an anti-noise method by reconstructing the signal tendency. This method can overcome the problem that the traditional method is sensitive to the local noise fluctuation..This project will provide a new idea for monitoring of regional boundary layer height and propose a new method to promote the application of spaceborne lidar in retrieving boundary layer height.

基于星载激光雷达遥感提取大区域边界层高度，是解决地面监测站点稀疏和以“点”代“面”问题的重要手段。然而，传统星载激光雷达边界层高度提取算法容易受到非目标层（光学厚层和异常噪声层）局部波动的干扰，导致区域边界层高度难以有效提取。由于温湿度廓线受非目标层的干扰较少，可为星载激光雷达精细化边界层高度定位提供约束条件。因此，本项目拟协同低分辨率MODIS温湿度廓线和高分辨率星载激光雷达气溶胶廓线来提取边界层高度。包括：1）建立不同大气条件下边界层高度范围预估算法的选择理论、剔除远场非目标层突变干扰；2）探索色比-极化比与边界层高度规律关系、削弱边界层顶附近光学厚层的影响；3）发展信号趋势重构的抗噪算法、克服传统算法对局部噪声波动敏感的问题。由此，为中国区域边界层高度精细化监测提供新思路，为推动星载激光雷达在大区域边界层高度遥感提取应用提供新算法。

边界层高度决定了大气边界层内湍流混合、垂直扩散和对流传递的垂直范围，是预测天气、气候和空气质量的关键因子。因此，准确地探测边界层高度对大气环境预测和全球气候变化的研究是十分必要的，边界层高度产品可服务于气象预报和空气污染预测，具有重要的科学意义、广泛的应用价值。本项目为获取大区域高精度边界层高度，本项目从站点边界层高度反演理论探索出发到实现区域反演，进行由点到面的激光雷达边界层高度反演算法综合研究，主要开展了以下几个方面工作：地面激光雷达连续探测下的边界层高度提取方法研究、多层次干扰下的星载激光雷达观测站点边界层高度抗噪算法研究、顾及气象条件的星载激光雷达观测站点边界层高度提取方法研究、多源数据和机器学习驱动下的区域边界层高度测算模型研究。研究成果包括：1）提出了考虑地面激光雷达连续观测特征的边缘检测算法，避免了当前算法受单一阈值限制易受大气条件变化的检测不稳定问题；2）提出了具有较强抗噪能力的随机采样拟合算法，突破了当前算法仅考虑衰减散射系数的梯度变化而易受光学厚层突变干扰的问题；3）提出构建站点激光雷达和气象数据驱动下的边界层高度装袋树模型，提高了模型泛化测算能力并反演出高精度站点边界层高度；4）构建了中国区域星载激光雷达边界层高度模型，克服了星载激光雷达边界层高度反演算法仅考虑局部突变而对噪声敏感和易受光学厚层干扰的问题，实现了中国区域边界层高度测算。研究成果为推动星载激光雷达在大区域边界层高度遥感提取应用提供新算法，为中国区域边界层高度精细化监测提供新思路，对发展具有我国星载激光雷达技术与方法，加强边界层高度的监测、预报和研究，具有重要的科学意义和应用价值。