To deeply understand the sunshine duration characteristics in the ecological conservation area of West Beijing (ECAWB), China, based on the data of daily total cloud cover, low cloud cover, precipitation days, relative humidity and daily sunshine duration of Mentougou and Zhaitang National Meteorological Stations in Beijing from 1975 to 2019, the spatial and temporal variation of sunshine duration in ECAWB, as well as the correlations with the variousmeteorological elements, are analyzed by using climate tendency rate and Mann-Kendall test methods. The results show that from 1975 to 2019, the annual, monthly, and daily sunshine duration in ECAWB were long in western mountainous area and short in southeastern plains, with significant correlations in the daily sunshine duration between the two meteorological stations (with about 80% of the difference in the range of-2 h to 2 h). The peak of monthly average sunshine duration was observed in April and May, and the valley appeared in November and December, with a most significant decreasing trend observed in June, September and October. The seasonal distribution of sunshine duration was generally uniform, despite more sunshine duration observed in spring and less in winter. In the meantime, the annual sunshine duration during 2002-2010 were consistently below the mean value, with a greater decreasing trend observed in southeastern plains (-51.6 h·(10a)-1) than in western mountainous area(-39.5 h·(10a)-1). Anomalously higher annual and seasonal sunshine duration were mainly observed from 1980s to 1990s, while anomalously lower annual and seasonal sunshine duration mainly occurred after 2000. The sunshine duration increased significantly during 1982-1990, and then decreased considerably from 1991 to 2019. The comprehensive analysis shows that the sunshine duration in this area were negatively correlated with total cloud cover, low cloud cover, daily average humidity, and daily precipitation, while significantly positively correlated with daily wind speed as well as visibility. These findings could be critical reference for the construction and climate evaluation in ECAWB.

Based on the monitoring data of O3 concentration from 2014 to 2020 and the meteorological data of automatic meteorological station in the same period, the daily and monthly variation rules of O3 concentration on the surface of Hefei, China are studied, and the influence of temperature, humidity, wind speed and other meteorological factors on O3 concentration is analyzed. The results show that O3 concentration in Hefei area has typical unimodal diurnal variation characteristics, which usually reaches the peak at about 15:00 and drops to the lowest daily value at about 07:00. The variation of monthly average of the maximum 8 h moving average of O3 concentration (O3-8h) showsan “M” type, which generally reaches the highest level in June and August, and falls to the lowest level in January and December, the monthly maximum value of O3-8h was 2.8~3.7 times of the minimum value, and the average value was 3.1 times. Affected by meteorological factors, the annual variation rule of O3 concentration is basically the same as that of temperature, but has no obvious relationship with the variation trend of humidity.

Atmospheric aerosols can affect the climate system of the earth by scattering or absorbing long-wave and short-wave radiation. As one of the aerosol tracers, 2-Methylglyceric acid has been observed frequently in atmospheric observations and nucleation experiments. Sulfuric acid and methanesulfonic acid, as very important aerosol precursors, have also received extensive attention and been widely studied. Therefore, 2-methylglyceric acid-sulfuric acid/methanesulfonic acid clusters were simulated based on the theory level of DF-MP2-F12/VDZ-F12 combined with M06-2X/6-311G(3df,3pd), and their physicochemical properties in the atmosphere were analyzed. The results show that 2-methylglyceric acid-sulfuric acid cluster and 2-methylglyceric acid-methanesulfonic acid cluster have the same Gibbs free energy temperature dependence. 2-Methylglyceric acid-sulfuric acid/methanesulfonic acid clusters will preferentially evaporate 2-methylglyceric acid molecules rather than sulfuric acid/methanesulfonic acid molecules, and the evaporation rate of 2-methylglyceric acid molecules increases rapidly with the increase of the cluster size. In addition, the Rayleigh scattering intensity and polarizability of 2-methylglycericacid-sulfuric acid/methanesulfonic acid clusters are calculated, which is helpful to understand the influence of the clusters on atmospheric radiation.

As remote sensing monitoring of mobile source exhaust can be affected by the complex external environment, it is difficult to establish a correaltion model between vehicle driving conditions and pollution emissions through traditional statistical methods. For this reason, the research on the analysis of influencing factors and emission prediction based on remote sensing monitoring of mobile sources is carried out. Firstly, Spearman correlation is usedto exclude the factors that have no correlation with CO, HC and NO, the main components in emission of mobile source pollution. Secondly, Lasso algorithm is used to choose the principal influencing factors. And after principal components analysis and theselection of algorithm and architecture, the Back-Propagation (BP) neural network model is established as the optimal algorithm. Finally, the validity of the model for predicting the main components of emission of mobile source pollution is verified on the test set. The results of model prediction show that the prediction models based on feature selection and BP has high prediction accuracy, which can reduce the cost of mobile source pollution emission detection and provide theoretical basis for policy making.

NO2 is one of the main atmospheric pollutants, which plays an important role in atmospheric photochemical process. It is of great significance to study the temporal and spatial variation law of NO2 concentration and predict the variation trend of NO2 concentration. The BP neural network based on particle swarm optimization (PSO) was proposed to predict atmospheric NO2 concentration. Based on the air pollution data and meteorologicaldata of Hefei area, China, from January 1, 2017 to December 31, 2019 and combined with the stepwise regression method, the influencing factors with high correlation with NO2 concentration were selected as the input samples. The PSO-BP neural networkprediction model was constructed, and then the optimal solution of the initial weight and threshold value of the BP neural network were found by using PSO algorithm. By comparing the prediction results of the traditional BP neural network, BP neural network improved by genetic algorithm and BP neural network improved by PSO, it was found that PSO-BP model can accurately predict the dynamic change of NO2 concentration with high prediction accuracy and simple model, which is expected to be widely usedin air pollutant concentration prediction in the future.

As the operating conditions of motor vehicles are complex and changeable, the concentration range of exhaust components is very large. Due to the fixed structure of optical gas cell and the detection limit of the system for weak photoelectric signal, the concentration range and detection accuracy of the gases to be measured are greatly limited when using conventional optical methods to detect the concentration of pollutants in exhaust gas. It is found that based on Lambert Beer′s law, adding an exponential factor correction to the gas concentration variable to be measured can realize the large range detection of CO and CO2 without reducing the measurement accuracy. The calibration experiment of the portable vehicle exhaust detecter with standard gas shows that the fitting degrees of CO and CO2 obtained by the traditional linear correction method are 0.988 and 0.998 respectively, while those of CO and CO2 obtained after adding nonlinear correction factor are 0.999 and 0.999 respectively. Further field comparison shows that the measurement results of the modified instrument are in good agreement with those of similar advanced instruments. The experimental correlation of diesel vehicle is 0.93 and that of gasoline vehicle is 0.95, which verifies the necessity and practicability of the proposed nonlinear correction method.

Imaging differential absorption spectroscopy technology (IDOAS) can obtain the spatial distribution of pollutants. It has been successfully applied to multiple platforms such as ground-based scanning, airborne and spaceborne, providing strong support for environmental monitoring and management. Among them, the ground-based IDOAS is mainly used in the detection of a certain pollution source. The principle of the imaging system based on the “push-broom” mode is analyzed, and the detection technology is applied to the detection of pollutant distribution in urban atmospheric boundary layer. In order to use the DOAS method to invert various trace gas components more efficiently and analyze the temporal and spatial distribution characteristics of polluted gases more accurately, the source-level analysis and optimization of QDoas software have been carried out. On the Windows platform, C++ and QT are used to reorganize the QDoas code. By re-extracting, integrating, rewriting, and optimizing the code, a faster and more convenient inversion function module is realized. In order to test the inversion performance of the module, taking the common pollutants NO2 and SO2 in the atmosphere as examples, an on-site observation experiment was carried out in Tongling Fuxin Iron and Steel Plant, China, on November 6, 2019. The two-dimensional distribution image of the polluted gas concentration in the target area is successfully obtained after the data inversion with the new compiled software, which confirms the applicability of the developed software in the actual atmospheric environment monitoring.

As an important detection method, active fluorescence detection uses fluorescence lifetime as the characteristic parameter of fluorescence detection, which can solve the problem that the fluorescence intensity is easily affected by external environmental factors to a certain extent. Based on the principle of the gated-detection method for measuring fluorescence life, the nonlinear least square regress combined with fluorescence lifetime decay curve isused to fit the fluorescence lifetime decay function to extract the average fluorescence lifetime parameters, and the two-dimensional spatial distribution of fluorescence substances is drawn from the fluorescence life map. Furthermore, a method of oil types recognition using fluorescence lifetime parameter as feature vector is proposed and experimentally verified. The experimental results show that the probability of the pixel point fluorescence lifetime falling into the confidence interval in the excitation region is more than 68% by using the fluorescence average life as a parameter, and the recognition probability is over 77% by using the support vector machine for oil types recognition. It seems that it is feasible and has a good recognition rate to identify oil species by using fluorescence lifetime parameter, and at the same time, less training samples are required for the method combined with support vector machine. Therefore, the oil recognition method based on fluorescence lifetime decay curvecombined with support vector machine will provide another reference for oil types recognition research in the field of environmental pollution.

Sea fog in polar regions poses a challenge to the research on polar science and sea ice. However, due to the lack of relevant cloud monitoring data in the polar region, the research on sea fog in polarregions is still relatively scare. Based on the CALIOP sensor′s ability to observe cloud information in the vertical direction, the MODIS medium resolution imaging spectrometer with plesiochronous observation is used to analyze cloud information in the Arctic region. Firstly, the deep neural network model is applied to invert the cloud top height. Then, according to the inverted cloud top height, whether it is sea fog can be ascertained. Furthermore, the influence of different wavebands on the inversion results is also analyzed. The results show that the average absolute error of the cloud top height inverted by the deep neural network is 1774.280 m lower than that of the traditional method, indicating that using deep neural network model can invert cloud top height better and more accurately, which can improve the detection accuracy of sea fog.