In order to obtain the visibility more conveniently and quickly and realize the long-term observation of the visibility, the visibility extraction method of the video image based on the theory of the dark channel prior is analyzed and studied. Through the brightness analysis on the target images observed at different times of the day, the corresponding atmospheric transmittance and visibility can be obtained. Furthermore, the visibility extraction results are compared with the measured lidar visibility data under the same conditions. The results show that the extracted visibility results are in good agreement with the measured data, and the correlation coefficient reaches 0.96616.

Atmospheric transmittance is an important factor affecting infrared radiation transmission. The calculation of atmospheric transmittance of infrared band based on the actual meteorological parameters of marine atmospheric is still weak, which inevitably causes the calculation error. In the paper, the actual atmospheric parameter profile in coastal areas of China is constructed and embedded into MODTRAN to restore the influence of clouds cover on infrared radiation transmission in real atmospheric environment. It is found that the change of the atmospheric temperature, humidity and other parameters directly affects the calculation results of atmospheric transmittance, and the change of cloud thickness has also obvious influence on atmospheric transmittance, and its influence degree is closely related to the cloud parameters and the relative geometric position between detector and cloud layer. It is shown that to restore the real atmospheric environment as much as possible is the basis of studying infrared radiation transmission, and the error of calculation results caused by atmospheric environment cannot be ignored.

Utilizing long-path differential optical absorption spectroscopy technique, air pollutants from ship emission in typical shipping channel area of the downstream of Huangpu River were measured with hightemporal resolution. It is found that SO2 concentration can increase by 2 to 4 times instantaneously and reach more than 10×10-9 due to the significant impacts of ship plume. However, since its emission sources are more complicated, the variation of NO2 concentration is relatively gentle, and under the influence of vehicle emission nearby, the diurnal variation of NO2 concentrations shows obvious bimodal characteristic. Affected by ship flow, SO2 concentration is relatively high in daytime. The analysis of the impact of human activities demonstrats that the NO2 concentration dropped by more than 30% during the Spring Festival holidays compared with that before and after, and even decreased to 50% after the first-level response to major public health emergencies of COVID-19. It is worth noting that the response of SO2 concentration to the control measures of COVID-19 and holiday effect is not so obvious as that of NO2 due to the differences of ship activity. However, the typical high emission value of SO2 concentration has shown a downward trend year by year.

Based on the method of multi-axis differential optical absorption spectroscopy (MAX-DOAS) retrieval of NO2 vertical column density (VCD), the corresponding ground-based MAX-DOAS system was constructed and the variation characteristics of NO2 VCD were observed. In the retrieval, the spectrum in the zenith direction was selected as the reference spectrum, and the NO2 slant column density (SCD) was retrieved by nonlinear least square method. Combining the slant column density of different observing directions, the NO2 differential slant column density (dSCD) was evaluated. Then the air mass factor (AMF) was calculated by the geometric approximation method, and finally the NO2 VCD was obtained. A one-year field experiment was carried out in Huaibei area from June 2019 to May 2020. The results showed that the monthly mean value of NO2 VCD presents an inverted “U” type change during the observation period, reaching the highestvalue of 2.13×1016 molecules·cm-2 in December and the lowest value of 5.23×1015 molecules·cm-2 in August. Moreover, the daily mean values of MAX-DOAS observations were compared with OMI (0 <FeC≤ 0.1 and 0 <FeC≤ 0.3) measurements, and it is found that the two results had a good agreement (R2 = 0.88, R2 = 0.90), which indicates that MAX-DOAS can not only accurately retrieve NO2 VCD, but also be used to verify the satellitedata.

Atmospheric PM2.5 concentration is a kind of data with strong time series characteristics, so currently the prediction of PM2.5 concentration is mostly based on RNN, LSTM and other sequence models. However, RNN, LSTM and the other similar models use the same weight to calculate the input data at different times, which is not in line with the brain-like design, resulting in the low accuracy of PM2.5 concentration prediction. In view of the above problems, a PM2.5 prediction method (AT-RNN and AT-LSTM) based on Adam attention mechanism is proposed. This method firstly looks for the optimal parameters of RNN or LSTM through Adam algorithm, and introduces attention mechanism in Encoder stage to assign attention weight to input with time series characteristics, and then carries out Decoder analysis and prediction. Through the experiment, the prediction effects of BP, RNN, LSTM and AT-RNN and AT-LSTM on PM2.5 concentration in Hefei city were compared. The results show that the prediction method based on Adam attention model is more accurate than other methods, which proves the effectiveness of this method in pollutant prediction.

Affected by the COVID-19 epidemic, the activity level of Wuhu residents has been reduced to the lowest level during the Spring Festival in 2020 (from January 24 to February 8). However, there were three different air pollution events occurred in Wuhu city, which provided an opportunity for exploring the cause of air pollution in Wuhu. The pollution characteristics and causes of the three pollution events are analyzed based on the monitoring data of ground meteorological elements, lidar monitoring data and HYSPLIT backward trajectory model. The results show that the first pollution event from 8:00 on January 27 to 18:00 on January 28 is a 10-hour transient pollution process induced by the high humidity and static atmospheric environment with the local basic emissions unchanged. The second pollution event from 17:00 on January 29 to 13:00 on February 1 is caused by a long-distance and long-term transportation of polluted air masses from Hebei, Henan and Shandong provinces, which intensifies the degree of pollution, and finally induces a two-day light pollution from January 30 to 31 under the increased emissions from key enterprises and unfavorable atmospheric diffusion conditions. And the third pollutionevent from 19:00 on February 2 to 07:00 on February 4 is a day light pollution caused by the short-term input of pollution groups from Jiangsu and Zhejiang provinces under the unfavorable diffusion conditions. Compared with the Spring Festival in 2019, the air quality in 2020 has improved significantly, the number of days of moderate and above pollution has decreased by 100%, and the pollution duration of light pollution has decreased by 48 hours. It can be seen that local source emission reduction is the fundamental to improve the air quality of Wuhu, and regional joint prevention and control is an effective means to deal with the polluted weather.

Ultraviolet polarization imaging technology has unique advantages in trace detection and space exploration. However, the current research mainly focuses on the ultraviolet polarization imaging technology and the system, and the research on the calibration method of ultraviolet polarization imaging system is relatively lack. In order to solve the calibration problem of uv-polarization imaging system, a systematic calibration experimental method is designed by using the self-developed uv-polarization time-sharing imaging system. By analyzing the design principle of uv-polarization imaging system, the calibration experiment is carried out, and the radiation calibration and polarization calibration of uv-polarization imaging system are realized. The experimental results show that after calibration, the uniformity error of the system is reduced to less than 1% and the measurement error of polarization degree is less than 2%, which meets the requirements of system calibration accuracy. The effectiveness of the calibration method is verified, which is of signicance for the calibration of ultraviolet polarization imaging system.

Turbidity monitoring can provide a scientific basis for the prevention, control and early warning of lake water quality. In order to improve the dynamic monitoring capability of lake turbidity, satellite remote sensing monitoring and buoy detection station monitoring are combined to study the temporal and spatial changes of Chaohu Lake turbidity in 2019. The temporal and spatial variation characteristics of Chaohu Lake′s turbidity are studiedthrough the statistical analysis of high-frequency continuous measured turbidity data at the buoy detection station, while the turbidity is quantitatively retrieved by constructing the optimal band combination model for remote sensing monitoring. The results showed that (1) the overall turbidity dynamics of Chaohu Lake is obvious, and the variation is significant within a short period of time, (2) the turbidity is sensitive to the red band and near-infrared band, (3) the overall turbidity of Chaohu Lake during the cyanobacteria outbreak period is relatively high, and diurnal turbidity changes significantly. This work provides a new idea for water quality monitoring, and at the same time promotes the application of air-ground joint methods in water quality monitoring.

In the development of space-borne trace gas different optical absorption spectrometer, it is necessary to test the performance indexes such as signal-to-noise ratio, spectral resolution and spectral calibration accuracy. Therefore, a special two-dimensional turntable needs to be designed to test different field of view at different angles. According to the test requirements of spectrometer parameters, the design scheme of the turntable is proposed, andthe structure of the turntable is designed, then the three-dimensional model is established, and the main parts of the turntable are selected. The azimuth axis of the designed turntable can rotate ±180°, which meets the requirements of 114° main field of view test, the pitching axis can rotate ±20° to meet the requirements of solar field of view, and the mechanical positioning accuracy of the turntable is ±0.01°. The turntable provides two installation methods, where vertical installation is for solar field test and horizontal installation for main field of view test. The structure of the turntable supporting frame is simulated under different working conditions by using finite element software, and the rationality of the frame design is verified.