Research On Cloud Parameter Inversion Algorithm Based On Lidar Data

Clouds are one of the important meteorological elements that influence the Earth’s climate in various aspects such as water cycle and large-scale circulation,and have important research significance.In the field of cloud detection,Li DAR has been widely used because of its strong detection continuity,long range and high resolution.However,due to the influence of background noise and aerosol signal,the traditional lidar cloud inversion method is prone to miss and misjudgment,and only for a specific cloud parameter,which is difficult to achieve a large range of cloud structure detection.In this paper,starting from the characteristics of Li DAR echo signal,we study the influence brought by background noise and aerosol signal on Li DAR cloud detection,combine the layer peak to layer bottom ratio threshold and improve the background noise threshold to realize the accurate inversion of cloud parameters,and on this basis,based on the vertical height and relative humidity characteristics of clouds,we propose a Li DAR cloud detection algorithm based on cluster analysis,and the detailed research is arranged as follows.1 、 Introduce the basic principle of lidar cloud detection,from the system noise,atmospheric background noise,aerosol signal three perspectives on the impact of various factors in lidar cloud detection has been more detailed,and for the background noise,aerosol signal,thin cloud signal proposed a combination of layer peak layer bottom ratio threshold,improve the background noise threshold of the differential zero-crossing method improvement scheme,combined with effective signal pre-processing technique,the two-threshold cloud parameter inversion algorithm based on the differential zero-crossing method is realized,which effectively reduces the misjudgment and omission of the traditional algorithm.2.In order to achieve effective Li DAR cloud hierarchy detection,this paper deeply investigates the physical properties of clouds that affect the cloud structure,and proposes a scheme to discriminate the Li DAR cloud hierarchy structure through cluster analysis according to the vertical distribution height of clouds and the relative humidity of clouds,and applies the scheme to experimental tests based on the ISODATA algorithm.The experimental results show that the method has good stability and detection continuity,and achieves 93.62% correct detection rate for low-level clouds,92.78% correct detection rate for mid-level clouds,and93.03% correct detection rate for high-level clouds.3.The algorithm proposed in this paper was used to process the Li DAR data from the Atmospheric Radiance Program at the Southern Great Plains site in the United States for the whole year of 2021,and invert the cloud hierarchy at this site,and analyze and discuss the spatial and temporal distribution characteristics of the vertical distribution range and occurrence frequency of different clouds at this site according to the experimental results,and compare the algorithm inversion results with the reference standard values provided by this site.The results show that the algorithm has a good agreement with the reference value in the height of cloud bottom and cloud top,and the distribution range and occurrence frequency of clouds are also consistent with the characteristics of the reference value at this location.