【作者】 李俊；

【导师】 龚威；

【作者基本信息】 武汉大学 ， 摄影测量与遥感， 2012， 博士

【摘要】 大气气溶胶指的是大气中处于悬浮状态直径小于10μm的各种固体和液体微粒。大气气溶胶在大气中相对含量很少,质量大约仅占整层大气的十亿分之一,但对于气候变化和大气辐射传输均有重要影响。激光雷达(LIDAR)作为大气气溶胶主动遥感观测手段,具有很高的空间和时间分辨率,广泛用于大气探测、环境监测等领域。在大气光学特性的测量过程中,激光雷达可实时监控,提供各大气参数值的高时空分辨率的垂直廓线,如：消光系数,后向散射系数和光学厚度,由消光系数垂直分布廓线可以进一步了解气溶胶的垂直分布情况,为各种研究提供数据支持。激光雷达技术在大气气溶胶探测领域中得到广泛的应用。激光雷达观测由于回波信号与距离的平方呈反比,整体的动态范围很大,要实现从低空到高空大范围覆盖的激光雷达观测是有困难的,具有低空探测盲区无观测信号和高空探测高度需求的矛盾。而且激光雷达是强激光发射和弱信号检测并存的系统,回波信号受到各种因素噪声的干扰,而反演结果的精度和可靠性受到回波信号中各种噪声的影响,需要对接收到的信号进行适当的降噪处理来改善反演结果。本论文即根据上述状况,提出开展双视场激光雷达分层探测,同时获得大气气溶胶在低空和高空的分布信息。该激光雷达将采用不同视场角的两个视场进行接收：小视场角可以降低天空背景噪声,接收高空信号；大视场角可以降低低空盲区,接收低空信号。采集两个通道的信号然后进行拼接,从而解决低空探测盲区和高空探测高度的矛盾,硬件技术上提高激光雷达探测能力。激光雷达信号存在着系统本地和天空背景噪声的干扰,降低了信号的准确度和可靠性,因此需要采用降噪算法进行处理,也是在软件技术上提高激光雷达探测能力的必要方法。由于采用双视场接收技术,和自动化降噪算法,可望达到显著提高大气气溶胶光学特性探测能力的目的,通过对空观测和数据反演处理,并与其它主被动设备相结合,获得武汉上空大气气溶胶光学特性的分布情况,为大气、遥感、气象和环保研究提供支持。本文的主要研究内容和创新点如下：1)双视场激光雷达技术研究。研制双视场激光雷达对大气进行分层探测,同时获得大气气溶胶在低空和高空的回波信号。该激光雷达将采用不同视场角的两个独立光学接收通道进行接收：小视场角接收高空信号,可以降低天空背景噪声；大视场角接收低空信号,可以降低低空盲区,降低几何重叠因子的影响。2)双视场激光雷达信号拼接。双视场激光雷达有两个视场的光学接收系统和多路信号采集系统。解决了光子计数模式和模拟探测模式的拼接问题和近场通道信号和远场通道信号拼接的问题,将多路信号合为一条从低空到高空覆盖的信号。通过对信号拼接,解决低空探测盲区和高空探测高度的矛盾。3)激光雷达信号降噪算法研究。充分利用双视场激光雷达具有不同信噪比的两路回波信号,进行基于实测信号的降噪算法对比试验。对基于经验模态分解的降噪算法进一步发展,解决其分解结果模态混叠的现象,更有效的实现信噪分离。更多还原

【Abstract】 Atmospheric aerosols are small particles suspended in the atmosphere. Atmospheric aerosols’ absolute quality is tiny, but they can directly affect climate change and radiative transfer by scattering and absorption of solar and terrestrial radiation, and also indirectly affect the radiation by affecting cloud formation as cloud condensation nuclei. As an active remote sensing instrument, lidar provides a high spatial resolution vertical profile of aerosol optical properties, such as aerosol extinction coefficients, backscattering coefficients and aerosol optical depth which can be applied in atmospheric study and environment monitoring. So lidar becomes more and more popular in the atmospheric research domain.The lidar equation shows that the power of the received signal rapidly falls with an increase in range, and the dynamic range is large. It is difficult to achieve the cover from lower to higher atmosphere. And the backscattering lidar signal and inversed results are often limited by various noises. Performing a proper pre-processing will improve the quality of the results.This thesis mainly focuses on the detection of atmospheric aerosol of the whole troposphere by a dual field of view lidar. The lidar has two independent receiving channels to solve the problem of the dynamic range, one for the lower atmosphere (especially the planetary boundary layer) by a larger field of view, and the other for higher altitudes by a smaller field of view. Combining these two channels can effectively increase the detection ability of the lidar by hardware approach. Lidar signals are contaminated by local and sky background noise. So we can increase the detection ability of the lidar by software approach, by processing a proper noise reduction algorithm. The detection ability of the atmospheric aerosol optic properties can be effectively increased by dual field of view technology and automatic denosing algorithm. The optic properties of atmospheric aerosol over Wuhan can provide data support to atmosphere, remote sensing, meteorology and environment protection..Firstly, the optic properties of the whole troposphere aerosols are obtained by a dual field of view lidar. The lidar has two independent receiving field of view to solve the problem of the dynamic range, the larger field of view for the lower atmosphere (especially the planetary boundary layer) by lowering the overlap area, and the smaller field of view for higher altitudes by reducing the effect of sky background noise.Secondly, there are two optical receiving system and three data acquisition system in the dual field of view lidar. After the combination of analog detection signal and photon counting signal, near-range channel signal and far-range channel signal, the obtained signals are combined to one signal which covers from atmospheric boundary layer to high atmosphere. The contradiction of lower blind detection zone and higher atmospheric detectiong ability, are resolved by data combination.Thirdly, the lidar noise reduction algorithm are studied by employing the two channel signals with different signal to noise ratio. The denosing algorithm based on empirical mode decomposition (EMD) is applied to lidar signals. It is found that the signal and noise are extracted to one intrinsic mode function (IMF) by the EMD-based denoising, so the threshold method is applied. The results show that the improved method can effectively distinguish the signal and noise.更多还原