When laser propagates in the atmosphere, the refractive index fluctuation caused by atmospheric turbulence will lead to optical turbulence effects such as wavefront aberration, beam wander, and scintillation, which severely restrict the development of remote sensing imaging systems and laser communication. Because of the influence of atmospheric optical turbulence on many fields, it′s significant to measure optical turbulence profile. To obtainthe temporal and spatial distribution of optical turbulence and evaluate the impact of optical turbulence on optical imaging and laser transmission systems require accurate measurement of optical turbulence. From the perspective of optical turbulence characteristic parameters, the current retrieval methods and research progresses of measuring atmospheric turbulence profile distribution at home and abroad are introduced, and the measurement principles, advantages and disadvantages of the different technologies are summarized. Finally, the new lidar system, named differential wavefront lidar, is presented, and its measurement principle is also introduced. The technology has the advantages of high spatial resolution and no focal shift. The preliminary simulation results of echo signal show that this technology can effectively measure the atmospheric optical turbulence profile.

Using the data of Aerosol Robotic Network (AERONET) in Beijing site from January 2016 to December 2018, the seasonal characteristics of aerosol optical depth (AOD), Angstrm exponent α, sizedistribution in Beijing were analyzed. Meanwhile, the characteristics of contaminant and lidar ratio were analyzed under typical pollution conditions. The results show that the seasonal variation of AOD in Beijing is obvious, its value is higher in spring and summer than autumn and winter. Especially in summer, the value (0.83) is significantly higher. The variation of Angstrm exponent α shows the similar regularity with AOD, with the lowest in spring (α=0.95) and the largest in summer(α=1.23), indicating that the main pollutant is dust in spring and fine particle in summer regarding to the typical temperate monsoonal climate. The diagram of AOD and α shows that the distribution of different pollutants have different characteristics, thresholding method can be used to classify the pollutant. Moreover, lidar data in two typical pollution cases was analyzed and the extinction coefficients were obtained by using different lidar ratios, and it shows that the solar photometerdata can be used to optimize the inversion parameters.

A heavy air pollution process in Beijing was observed from March 29 to April 10, 2018, and its pollution characteristics and optical characteristics were analyzed by using air quality monitoring data and the observation data of AERONET Beijing station. Combined with HYSPLIT airflow backward track mode, the temporal and spatial transport characteristics of atmospheric particles during this heavy pollution were comprehensively studied. The results show that, the heavy pollution process was dominated by fine mode aerosols with weak absorption, and the PM2.5 concentration and aerosol optical characteristics were significantly affected by the meteorological element during the pollution process. The results also show that the results from AERONET are in good agreement with those from the ground air monitoring stations, which indicates that AERONET results can be adapted to the study of aerosol optical characteristics in heavy air pollution process.

In order to get a better understanding of the correlation between respiratory diseases outpatients and atmospheric PM2.5, SO2 concentrations, multifractal detrended cross-correlation analysis(MF-DCCA) was used to study the sequence of respiratory diseases outpatients and PM2.5, SO2 concentrations in Yongding District. The results show that the correlation between respiratory diseases outpatients and atmospheric PM2.5, SO2concentrations has the characteristics of long-term persistence and multifractal. Then the dynamic sources of their correlation multifractal features are analyzed. Through random rearrangement and phase randomization procedure, the results show that the long-term persistence effect is the main driving force at different time scales. Further study found that the correlation between respiratory system diseases and atmospheric PM2.5, SO2 concentrations sequence in the four seasons has long-term multifractal characteristics, and the multiple fractal features in summer are stronger than those in other seasons.

When the air flow passes through the sweep plane of the wind turbine, the wake effect will be produced at the downstream of the wind turbine, which has different effects on the generation efficiency and fatigue load of the wind turbine. Based on coherent Doppler wind lidar, an all-weather wind field detection experiment was carried out in an offshore wind farm in Jiangsu Province. Due to the double Gaussian distribution of wind speed on the vertical section of the wind turbine wake, the traditional single Gaussian fitting algorithm has a large calculation error and can′t reflect the wind speed change law of real flow field near the wind turbine. An improved single-double Gaussian model fitting algorithm is proposed and used to analyze the wake width, velocity deficit and wake length of the target wind turbine. The feasibility and accuracy of the single-double Gaussian fitting algorithm for the fitting of the wake lateral wind speed are verified.

Cloud detection is the prerequisite of remote sensing image processing and application. It is a widely challenge that the accuracy of cloud detection from remote sensing image is easily influenced by thin clouds and cloud-like ground targets. Therefore, a hierarchical support vector machine (H-SVM) cloud detection algorithm combining grayscale, texture, and frequency features of remote sensing image is proposed in this work. Firstly, a simple linear iterative clustering algorithm is used to segment the remote sensing image into pixel blocks. Secondly, a H-SVM classifier is designed to perform cloud detection on the segmented pixel blocks, where the first layer of the H-SVM preliminarily divides the pixel blocks into “cloud” and “landscape categories”, and the second layer containing two classifiers further classifies the classification results of the first layer and then merges the classified results into three categories of “thick cloud”, “thincloud”, and “land features”. Finally, the classification results are processed using expansion algorithm to get the final cloud detection results. RGB band remote sensing images of GF-1 WFV are selected for verification experiments. It is shown that the method proposed in this study has an average accuracy of 95.4% for the cloud detection in the experimental images, which indicates that the method can be used for cloud detection of remote sensing images in multiple scenarios, and serve the productionand application of remote sensing products.

Aiming at the problem of information extraction of suspected illegal buildings in cities, a new algorithm for extracting suspected illegal building information based on multi-scale segmentation method is proposed. Firstly, the remote sensing image is segmented by setting multi-scale segmentation parameters, then is preprocessed by orthorectification, radiometric calibration, and atmospheric correction. Finally, the influence of each parameter on the scale segmentation effect is analyzed through multi-scale segmentation experiment. The results show that when the segmentation scale is 150, the shape scale factor is 0.7, and the compactness scale factor is 0.3, the image segmentation effect is the best.

Temperature control plays an important role in LF refining process, which is the key to control the quality of molten steel composition, continuous casting quality and stable process. A non-contact molten steel temperature monitoring system is developed. The system can obtain the real-time and high-clear heat image of molten steel in LF furnace through a dual optical path area array CCD detector, and then processes the image information by computer, so as to accurately identify the molten steel and calculate the real-time temperature data according to the temperature measurement model. The system can also be used as industrial TV, which is convenient for steelworkers to know the conditions in the LF furnace in time. Different from the previous contact temperature measurement methods, the new system can measure the molten steel temperature in LF furnace in real time without affecting the smelting progress. It is convinient and safe to operate, which isa technology worthy of promotion and also has a guiding role in other temperature measurement application fields.