The Atmospheric Space-borne Lidar Cloud And Aerosol Classification Algorithms

With the rapid development of our society and the acceleration of industrialization and urbanization process,the problem of atmospheric environment is increasingly important.LIDAR play a significant role in acquiring the vertical profiles of clouds and aerosols as well as in studying the vertical distribution characteristics of atmospheric clouds and aerosols and their impacts on global climate change.Based on the optical and geographic characteristics of aerosols and clouds observed by CALIOP,in this study,a quick and effective cloud/aerosol detection and classification retrieval algorithm was developed.1.The CALIOP Calibration Algorithms.Accurate calibration of the LIDAR signals is essential for layer detection and the subsequent retrieval of cloud and aerosol optical properties.The 532 nm parallel channel can be calibrated using the traditional high-altitude molecular normalization technique.The nighttime calibration coefficients computed for the 532 nm parallel channel function as the main system calibration,all other channels are calibrated relative to this measurement.We can compute the 532 nm perpendicular calibration coefficients by the PGR data and 532 nm parallel.Based on the optical properties of the cirrus cloud,the 532 nm total calibration coefficients are used to calibrate the 1064 nm channel.2.CALIOP's Layer Detection.The layer detection is the foundation of the retrieve of the layer optical properties,since the profile signals of CALIOP contains a variety of layer feature.In consideration of the low SNR and the top-down detection mode,SIBYL(Selective Iterated Boundary Location)is developed to detect the layer of CLIOP.Based on the threshold methods,SIBYL makes multiple passes through a specified scene,constructing profiles of attenuated scattering ratios at o series of increasingly coarse spatial resolution(333m,1km,5km,20 km and 80km),which can retrieve the tenuous features – e.g.,thin cirrus clouds and faint aerosol layers.3.CALIOP Scene Classification Algorithms.Correct discrimination between clouds and aerosol observed by the space-born LIDAR is critical for accurate retrievals of cloud and aerosol optical properties and the correct interpretation of measurements.A quick and effective cloud/aerosol classification retrieval algorithm was developed by introducing into the support vector machines(SVM)and decision tree methods.The algorithm includes 3 parts: 1)the cloud and aerosol discrimination with the classification confidence functions of 5-D probability density function 2)Ice-water algorithm with SVM,we chose SVM as the basis of the ROI and water cloud classification.Then,by combining the probability density functions based on the ?532,?,?,and Z values of the cloud layers and the cloud-top temperature T,we constructed classification confidence functions for ROI,HOI and water clouds to correct the feature layers misclassified by SVM,as well as removed a small portion of the HOI ice clouds from the water clouds.3)the aerosol subtype classification with decision tree classification.Our retrieval results showed a good agreement with the CALIOP VFM products.For the cloud and aerosol discrimination results,the consistency ratios between our retrieves and VFM products for aerosol and cloud are 98.51% and 88.43%,respectively.In addition,the consistency ratios in the day are higher than those in the night.For the cloud phase retrieval results,93.44%,88.04% and 51.11% of water cloud,randomly oriented ice(ROI)and horizontally oriented ice(HOI)classifications,respectively,are the same with the VFM data.The consistency ratio of HOI is low due largely to the confusion between HOI and ROI.For the aerosol subtype classification,most aerosol subtypes could be well recognized by our algorithm.However,the consistency ratios of the mixed subtypes(e.g.polluted continental and polluted dust)between retrieval results and VFM products are relatively low.4.Validation and application of CALIOP Algorithms.Moreover,the cloud/aerosol,cloud phase and aerosol subtype classifications were also compared with the VFM products under three typical air conditions,i.e.haze,dust and clean.Under the haze condition,our results for most of the smoke aerosols and polluted aerosols(polluted dust and polluted continental)agree quite well with the corresponding results from VFM.Under the duststorm condition,our algorithm can effectively discriminate the most of dust and polluted dust aerosols.Under the sunny condition,our results for the few existing cloud and aerosol layers are quite consistent with the VFM results.5.CALIOP Separation and Analyze of dust.Based on the 532 nm parallel and perpendicular channel,we separate the dust and the local aerosol and acquire the each optical properties,which can provide a means to study the interaction of the dust and the local aerosol.This paper is the important improvement of the cloud and aerosol classification algorithms,which can simplify the processing and improve efficiency with satisfactory accuracy.In the future work,we will build day/night and seasonal training sample sets,and consider more ice cloud phases and aerosol properties in the cloud/aerosol classification retrieval algorithm.