To investigate the potential impact of PM2.5-bound heavy metals on the water source area, four monitoring sites were established in the Xichuan Reservoir area of the Danjiangkou Reservoir, a core component of the South-to-North Water Diversion Project. PM2.5 samples were collected from June 2022 to April 2023 (n = 112), and concentrations of nine heavy metals (Cr, Mn, Fe, Ni, Cu, Zn, As, Cd, and Pb) were analyzed. The study identified major pollution sources and atmospheric transport pathways and assessed potential health risks. Results indicated that the total average concentration of the nine heavy metals was found to be 1.9062 μg m−3, with Zn and Fe being the most abundant, contributing 72

Groundwater and surface water are key components of the natural hydrological cycle, but the quantitative understanding of how their interactions dynamically respond to extreme climatic events is still limited. This study investigated the riparian wetlands downstream of the Xiaolangdi Reservoir (Henan, China) to elucidate how extreme precipitation (specifically the “7·20” extreme rainfall event) disrupted normal hydrogeochemical evolution and groundwater and river water exchange mechanisms. Integrating hydrochemistry, stable isotopes (δD, δ18O), and a Two-Endmember Mixing Model, we quantitatively evaluated the interaction between the groundwater and river water. The results indicated that while rock weathering primarily governed the HCO3-Na·Ca hydrochemical type, the extreme precipitation event triggered a dramatic dilution effect, sharply decreasing major ion concentrations in the aquifer. Isotopic analysis revealed that the δD and δ18O values of both river water and groundwater were relatively enriched during the flood season, indicating that the floodwater underwent intense pre-recharge evaporation before infiltrating the aquifer. Furthermore, the results demonstrated that extreme precipitation significantly impacted the interaction between groundwater and river water by reversing the exchange direction. Prior to the annual flood season (before extreme precipitation occurred), river water primarily recharged groundwater, whereas groundwater became the dominant source discharging into the river after extreme rainfall. This altered recharge-discharge pattern persisted for a period after the precipitation events had ended. These findings improve understanding of hydrological connectivity in reservoir-regulated riparian wetlands and provide acritical insights for regional water resource management and ecological protection under a changing climate.

Water-soluble organic nitrogen (WSON) affects the formation, hygroscopicity, acidity of organic aerosols, and nitrogen biogeochemical cycles. However, qualitative and quantitative characterizations of WSON remain limited due to its chemical complexity. In the study, 1year field samples of particulate matter 2.5 (PM2.5) were collected from June 2022 to May 2023 to analyze the WSON concentration in PM2.5, and correlation analysis, positive matrix factor (PMF), and potential source contribution function (PSCF) models were employed to elucidate WSON source apportionment and transport pathways. The results revealed that the mean WSON concentrations reached 1.98 +/- 2.64 mu g/m3 with a mean WSON to water-soluble total nitrogen (WSTN) ratio of 21 %. Further, WSON concentration exhibited a seasonal variation trend, with higher values in winter and lower in summer. Five sources were identified as contributors to WSON in PM2.5 within the reservoir area through a comprehensive analysis including correlation analysis, PSCF and concentration weighted trajectory (CWT), and PMF analyses. These sources were agricultural, dust, combustion, traffic, and industrial sources, of which agricultural source emerged as the primary contributor (76.69 %). The atmosphere in the reservoir area were primarily influenced by the transport of northeastern air masses, local agricultural activities, industrial cities along the trajectory, and coastal regions, exerting significant influences on the concentration of WSON in the reservoir area. The findings of this study addressed the research gap concerning organic nitrogen in PM2.5 within the reservoir area, thereby offering a theoretical foundation and data support in controlling nitrogen pollution in the Danjiangkou Reservoir area. (c) 2025 The Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences. Published by Elsevier B.V.

Nitrate (NO3-) is a key component of atmospheric particulate matter. However, studies quantifying NO3- formation and sources in water source areas remain scarce. This study employs NO3- concentration, dual isotopic analyses (δ15N, δ18O), and SIAR and PSCF models to elucidate NO3- formation mechanisms and apportion sources in the Danjiangkou Reservoir area. The result showed that the concentration of NO3- (0.52 μg m-3-2.39 μg m-3), δ15N-NO3- (-2.3‰-4.2‰) and δ18O-NO3- (64.2‰-96.6‰) values varied seasonally in total suspended particulate matter (TSP). The contribution of the NO2 + ·OH pathway was the highest in summer (84%), while the main contributions came from the N2O5 + H2O and NO3 + VOCs pathways in winter (75%). The SIAR model revealed that the contributions of coal combustion and traffic emissions were 34% and 23%, respectively, indicating that fossil fuel release contributed most. Moreover, the contribution of biomass combustion sources was 31%, and soil emission sources had the least (12%). Considering liquefied petroleum gas, gasoline, and diesel vehicles as traffic emission sources, their contribution to NO3- in TSP increased to 50%. The PSCF analysis indicated that external input areas were predominantly from central Henan Province, northern Hubei Province, and central Shaanxi Province. This study quantitatively analyzed the sources of NO3- in atmospheric particulate matter in water sources for the first time, indicating that controlling the emissions of nitrogen oxides from gasoline cars, heavy-duty diesel trucks, and ships could help alleviate haze pollution and enhance air quality in the Danjiangkou reservoir area.

Dissolved organic nitrogen (DON) deposition was the substantial component of dissolved total nitrogen (DTN) deposition in the world's nitrogen deposition hot spots areas. However, the information on the importance for DON deposition and its sources was still scarce, which limited the comprehensive assessment of the ecological threat from nitrogen deposition. Six sampling sites around the Danjiangkou Reservoir were set up to collect the dry and wet deposition samples from October 2017 to September 2021. The results showed that dry and wet DTN deposition averaged 34.72 kg ha(-1) yr(-1) and 22.27 kg ha(-1) yr(-1), respectively. Dry NH4+-N, NO3--N and DON deposition averaged 14.28 kg ha(-1) yr(-1), 5.91 kg ha(-1) yr(-1) and 14.53 kg ha(-1) yr(-1), respectively. Wet NH4+-N, NO3--N and DON deposition averaged 11.14 kg ha(-1) yr(-1), 3.89 kg ha(-1) yr(-1)and 7.24 kg ha(-1) yr(-1), respectively. The contributions of DON to DTN were 41.85% (in dry deposition) and 32.50% (in wet deposition), respectively. Dry DON deposition varied between 26.44 kg ha(-1) yr(-1) and 9.11 kg ha(-1) yr(-1), and significantly differed among six sampling sites (P < 0.05). The different intensity of agricultural activities disturbance at the sampling sites was the important reason for the spatial variations of DON deposition. DON deposition was significantly correlated with ammonium nitrogen (NH4+-N) deposition (P < 0.05). According to the results of positive matrix factorization (PMF) model, agriculture source contributed significantly to the DON deposition, the contributions at six sampling sites ranged from 45.8% to 73.7% in dry deposition, and from 56.8% to 81.6% in wet deposition. In summary, our findings found that agricultural activities were the important factors influencing the spatial patterns of DON deposition around Danjiangkou Reservoir and provided new evidence for the anthropogenic source of DON deposition in China.

Ammonium nitrogen (NHx) in atmospheric wet deposition is a primary external nitrogen source for reservoirtype water sources, and identifying its sources is crucial for controlling nitrogen pollution in water. This study aimed to examine the NHx characteristics and sources in wet deposition in the Xichuan reservoir area of the Danjiangkou Reservoir, the South-to-North Water Diversion Project (SNWDP) water source, from September 2019 to August 2020. Major inorganic nitrogen concentrations in precipitation and ammonia-nitrogen isotope (delta 15N-NH4+) values were measured, and the sources were analyzed using the Bayesian mixing model. The results showed that NH4+ was the main inorganic nitrogen form in wet deposition, with a concentration ranged of 0.16-5.26 mg L-1. The delta 15N-NH4+ values ranged from -14.7%o to +6.3%o, with significant isotopic effects from agricultural sources and climatic conditions, showing higher values in summer and lower values in winter. The release of ammonia (NH3) from agricultural sources was the primary NHx source (58.0%-76.6%), with fertilizer application being the most significant contributor (32.1%-46.6%). Seasonally, the relative contribution of agricultural sources to NH4+ in wet deposition was higher in autumn and winter. Spatially, agricultural activities significantly impacted atmospheric nitrogen accumulation across the reservoir area. This study quantified NHx sources in wet deposition, providing isotopic evidence and scientific references for nitrogen control measures and atmospheric nitrogen cycling studies.

Currently, many sensing structure models can only sense the refractive index of single variable samples to be measured. In order to achieve high-throughput detection of different samples to be measured and reduce the interference of environmental factors, a variable period subwavelength dielectric grating multilayer composite structure based on wavelength modulation is proposed. Take the double period as an example for analysis. The variable period grating layer is composed of two dielectric gratings, A and B, with different grating periods. The transmission characteristics are analyzed by the finite element method. The TE polarized incident light is incident on the surface of the dielectric grating in a manner perpendicular to the surface of the grating layer. When the phase matching conditions are met in the dielectric grating areas A and B, the variable period subwavelength dielectric grating will form GMR, Providing two double discrete resonance defect peaks with a single narrow band. Because the F-P-like cavity contains a periodic photonic crystal, the photonic band gap will be generated when the light wave propagates to the photonic crystal, providing a wide band continuous state. Under the condition that the phase matching condition is satisfied, the double discrete state resonance defect peak formed in the variable period subwavelength waveguide structure is coupled with the continuous state formed in the F-P-like cavity composed of the periodic photonic crystal multilayer dielectric film, and the double Fano resonance is realized. Then, by exploring the influence of waveguide layer thickness d(w) and photonic crystal cycle number N on the sensing characteristics, we choose d(w)=97 nm and N=3 to maximize the FOM value. Finally, the variable period dielectric grating layer is composed of two materials with different dielectric refractive indices, two sensing and detection units can be set in the grating groove part of the dielectric grating areas A and B, and a dual Fano resonance all-dielectric sensing model based on wavelength modulation is established. Different sensing and detection areas are set, and it is found that the dual Fano spectral curve can change with n(s1) and n(s2) in different sensing and detection areas; the dynamic detection of the refractive index of the sample to be measured is indirectly realized. Therefore, the multivariate detection of different refractive index intervals of the sample to be measured can be realized in the same sensing structure model. The results show that the FOM values of FR1 and FR2 in the sensor detection unit A are 631.53 and 463.7 RIU-1, respectively; in the sensor detection unit B, the FOM is 480.67 and 834.04 RIU-1 respectively. The sensor structure model designed has realized high reflectivity, high FOM value and wide detection range of the sensor structure through structural parameter optimization, which provides a theoretical reference for the dual Fano resonance and has certain research value for the multivariate detection of the sample's refractive index to be measured.

为了解丹江口水库淅川库区大气干沉降中氨氮的沉降特征和主要来源,于2019年9月-2020年8月对库区周边设置的5个大气监测点进行干沉降样品采集,测定并分析了样品中氨氮浓度及其氮同位素,基于贝叶斯混合模型量化了氨氮的主要来源.结果表明,库区大气干沉降中氨氮月均浓度为0.96mg·L-1,全年大气干沉降中氨氮沉降通量为11.77kg·hm-2,δ15N-NH+值月均值为-9.20‰.干沉降中氨氮沉降通量、浓度和δ15N-NH4+值的季节差异显著,均表现出夏季最高、春季和秋季次之、冬季最低的变化规律,主要与夏季高温和氮肥施用有关.运用稳定同位素模型(SIAR)分析发现,库区干沉降中氨氮的主要污染源为农业源,贡献率为60％,其中化肥释放源和畜禽排放源贡献率分别为36％和24％.夏季农业源贡献率最高,其中64％来源于化肥释放,进一步证实了夏季高温以及氮肥的大量施用是影响库区干沉降中氨氮的主要因素.研究结果为探索库区水体氮污染控制途径提供理论基础和数据支撑.

河南某县产业集聚区污水处理厂收集该区生产废水及部分城区生活污水,进水以有机类工业废水为主,占比在70％以上,且进水水质波动大、氮磷等污染物浓度高.污水处理厂设计近期规模2万m3/d,远期规模4万m3/d.针对进水特点,生化段设计采用改良A2O+MBR工艺,并设置双缺氧池及两级硝化液回流以强化脱氮除磷.出水指标执行《地表水环境质量标准》(GB 3838—2002)Ⅳ类标准,其中出水TN参考《河南省辖海河流域水污染物排放标准》(DB 41/777—2013),TN≤15 mg/L.冬季低温期3个月的运行结果表明,该污水处理厂运行稳定,出水COD、NH3-N、TN、TP的去除率分别为94.26％、98.77％、87.76％、95.19％,出水水质全部优于设计出水标准,该工艺对水质冲击有较好的承受能力.本工程运营成本为1.47元/m3.

为探究丹江口水库淅川库区大气干沉降中水溶性有机氮(WSON)特征及其来源,采用降水降尘自动采样器收集了2021年1-12月的大气干沉降样品,分析了其中WSON、水溶性总氮(WSTN)以及水溶性无机离子的浓度,探讨了 WSON的浓度水平及其来源.结果表明,采样期间,库区大气干沉降中WSON月均浓度为0.15～1.14 mg·L-1,年平均WSON值为0.52 mg·L-1,占WSTN的31％.干沉降中WSON浓度季节差异性显著,呈现出冬季最高、春季和夏季次之,秋季最低的变化规律.WSON与二次离子(NH4+、SO42-和NO3-)、Cl-、K+、Mg2+、Ca2+呈显著相关,表明WSON可能来自二次转化、生物质燃烧和扬尘.后向轨迹分析表明,库区主要受长距离传输的影响,冬季和春季的气团主要来自西北部,夏季和秋季的气团主要来自西南部.PMF分析结果表明,干沉降中WSON主要来源于二次源、农业源、扬尘源、工业源和燃烧源,且以二次源为主,贡献率为55％.研究结果为探索库区水体氮污染控制途径提供了理论基础与数据支撑.

Ammonia and ammonium (NHx=NH3+NH4+) deposition from agricultural sources is one of the important nitrogen sources received by water bodies in reservoir-type water sources. The identification of deposition source is an important basis for the prevention and control of nitrogen pollution in water bodies. For evaluating the source of ammonia-nitrogen in the wet deposition of reservoir-type water sources and contribution rates of different ammonia (NH3) sources, wet-deposition samples were collected from five monitoring sites in the Xichuan area of the Danjiangkou Reservoir from September 2019 to August 2020. The ammonium (NH4+) concentration, nitrate (NO3-) concentration, and δ15N-NH4+ in the samples were measured. The results showed that the concentrations of NH4+ and NO3- in the wet deposition had significant seasonal changes, which were higher in winter and autumn than in spring and summer. Rainfall and temperature were important factors affecting nitrogen concentration in the wet deposition. The values of δ15N-NH4+ in the wet deposition was in the range of -14.7‰ to +6.3‰. The δ15N-NH4+ value significantly positively correlated with the precipitation and temperature, and this value was higher in summer than in winter. The results of the Bayesian isotope mixing model showed that the agricultural source, especially NH3 released from fertilizer, was the principal NH3 source released to the wet deposition in the Xichuan reservoir region. The contribution of agricultural sources to NH4+ in the wet deposition in Xichuan reservoir area was highest in summer (61.2%) and lowest in winter (59.6%). Spatial differences were not significant, but the greatest impact was found in areas with less interference from anthropogenic activities. This study quantifies ammonia nitrogen sources in atmospheric wet deposition of reservoir type water source and provides a reference for controlling nitrogen input from outside the atmosphere.

Atmospheric nitrogen deposition (AND) may lead to water acidification and eutrophication. In the five months after December 2019, China took strict isolation and COVID-19 prevention measures, thereby causing lockdowns for approximately 1.4 billion people. The Danjiangkou Reservoir refers to the water source in the middle route of South-to-North Water Diversion Project in China, where the AND has increased significantly; thus, the human activities during the COVID-19 period is a unique case to study the influence of AND to water quality. This work monitored the AND distribution around the Danjiangkou Reservoir, including agricultural, urban, traffic, yard, and forest areas. After lockdown, the DTN, DON, and Urea-N were 1.99 kg · hm−2 · month−1, 0.80 kg · hm−2 · month−1, and 0.15 kg · hm−2 · month−1, respectively. The detected values for DTN, DON, and Urea-N in the lockdown period decreased by 9.6%, 30.4%, and 28.97%, respectively, compared to 2019. The reduction in human activities is the reason for the decrease. The urban travel intensity in Nanyang city reduced from 6 to 1 during the lockdown period; the 3 million population which should normally travel out from city were in isolation at home before May. The fertilization action to wheat and orange were also delayed.

The separation of metal ions with different valences is an intractable challenge in the treatment of rare earth (RE) ore leaching solutions, and nanofiltration (NF) is one of the most promising approaches to separate ions with different valences. Therefore, the separation behaviors of three commercial NF membranes (NF5, NF270 and NF90) were investigated in this study. The results showed that NF technology could be used for the separation of cations with different valences from nitrate and chloride; the NF270 membrane could be used for the separation of Na+/Ca2+ due to its appropriate pore size and negative charge. The NF5 membrane could be used for Ca2+/Nd3+ separation due to its large pore size and suitable zeta potential. Increasing the operating pressure and cross-flow velocity was beneficial to increase the separation factor of Ca2+/Nd3+. Increasing the solution pH and salt concentration facilitated the adsorption of metal cations on the membrane surface, resulting in the separation factor of Ca2+/Nd3+ increasing first and then decreasing. This work may provide new strategies to recycle valuable metals from RE wastewater or other complex mediums.

Diatoms constitute an important part of the phytoplankton community in lakes and reservoirs and play a significant role in regulating ecological balance. Danjiangkou Reservoir is the water source area of the middle route of China's South-to-North Water Diversion project. In order to explore the spatial and temporal distribution and know the governing factors of the diatom community, 18srRNA sequencing was carried out from seven sampling sites of the reservoir. At the same time, the concentration of nutrients present in the collected sample water was also determined. The results showed that a total of 51 genera and 96 species were thriving the community of diatoms in Danjiangkou Reservoir. Discostella was dominant in summer and autumn, accounting for 98.84% and 62.71% of the diatom abundance, respectively. Aulacoseira was dominant in spring and winter, accounting for 60.62% and 60.90%, respectively. Discostella and Aulacoseira showed significant differences in seasonal variation (p < 0.05). The colinear network of diatoms changed significantly with the seasons, mainly consisting of Aulacoseira, Discostella, and Stephanodiscus. RDA redundancy analysis showed that water temperature (WT), total nitrogen (TN), NH4+-N, pH, and electrical conductivity (Cond) were the main environmental factors driving the changes in diatom community structure.

Nitrogen deposition is an important means of exogenous nitrogen input in reservoir water. Agricultural activities around the reservoir lead to a sharp increase in the concentration of ammonia in the atmosphere, which poses a threat to the reservoir water body. Clarifying the contribution of agricultural ammonia release to atmospheric NHx (gaseous NH3 and particulate NH4+), in the reservoir area can provide a theoretical foundation for local reactive nitrogen control. We collected atmospheric NH3 and NH4+ samples during the agricultural periods and analyzed the isotopic characteristics of atmospheric NHx and the contribution rates of different ammonia sources in the Xichuan area of the Danjiangkou Reservoir. The results showed that the initial δ15N values of NH3 (-30.0‰ to -7.2‰) and particulate NH4+(-33‰ to +4.9‰ for finer and coarser particles, respectively) are different, and their contribution ratios from dissimilar ammonia sources are also different, among which NH4+ is more susceptible to meteorological factors. However, since the atmospheric NHx in the Xichuan area is mainly gaseous NH3, the final sources of atmospheric ammonia nitrogen source depend on gaseous NH3. Agricultural sources (59%-74%) were the main NH3 sources in this area. Among them, the fertilizer use emission was dominant; it had the highest contribution rate in summer during the agricultural period and a more prominent impact in areas with less human interference. Reasonable regulation of the application of high-ammonia releasing fertilizer, especially during the agricultural period in summer, is an effective way to reduce the threat of atmospheric ammonia to water health.

针对稀土矿浸出液中不同价态硝酸盐的分离需求,采用正交试验研究了NF5纳滤膜对NaNO3、Ca(NO3)2、Nd(NO3)3和Ce(NO3)4混合盐的分离效果,并考察了盐溶液浓度、操作压力、进料流速和溶液pH等因素对不同价态硝酸盐分离因子(SF)的影响,确定了分离硝酸盐中不同价态阳离子的最佳条件.结果表明,最佳条件下硝酸盐的分离效果为:SFNa/Ca=18.23,SFNa/Nd=35.20,SFNa/Ce=33.26,SFCa/Nd=4.55,SFCa/Ce=4.45,SFNd/Ce=1.10,说明 NF5 纳滤膜对一价与高价离子(SFNa/Ca、SFNa/Nd和SFNa/Ce)具有优异的分离效果,对二价与大于二价离子(SFCa/Nd和SFCa/Ce)的分离具有一定效果,而对三价和四价离子(SFNd/Ce)的分离效果不佳.

Organic nitrogen (ON) is an important part of atmospheric nitrogen deposition, but the content and distribution of components other than urea and amino acids are the blind area of current research. The deposition of organic amines (OA) in strategic water sources poses a great public health risk to unspecified populations. In order to further reveal the composition of about 50% soluble organic nitrogen, besides urea and amino acids, five functional sampling points (such as industrial area, agricultural area, urban area, tourism area and forest area) were set in the reservoir area to detect dissolved total nitrogen (DTN), dissolved organic nitrogen (DON) and OA components. The results show that the total nitrogen concentration was 6.42–10.82 mg/m3 and the DON concentration was 2.77–4.99 mg/m3. Ten kinds of OA were detected: dimethylamine (DMA), diethylamine (DEA), propylamine (PA), butylamine (BA), pyrrolidine (PYR), dibutylamine (DBA), N-methylaniline (NMA), 2-ethylaniline (2-ELA), benzylamine (BMA), and 4-ethylaniline (4-ELA). The average concentrations were 7.64, 26.35, 14.51, 14.10, 18.55, 7.92, 10.56, 12.84, 13.46 and 21.00 ng/m3, respectively. The total concentration of ten OA accounted for 2.28–9.81% of DON in the current month, of which the content of DEA was the highest, reaching 0.71%, the content of 4-ELA, PYR, PA and BA was 0.4–0.56%, and the content of DMA, DBA and NMA was 0.2–0.36%. The sources of OA in the reservoir area have significant seasonal differences. The content is the highest in spring, followed by autumn, and lower in summer and winter. The rainfall in spring and autumn is small, the source of road dust is relatively high, and the rainfall in summer is large. After the particles in the air are washed by rain, the concentration of OA in the sample is the lowest. On account of spring and autumn being the time of frequent agricultural activities, the concentration of OA is significantly higher than that in winter and summer.

Excessive concentration of organic pollutants in water is harmful, which causes not only serious environmental pollution but also endangers human health. Chemical oxygen demand (COD) can be used to characterize the pollution degree of organic pollutants in water. A quantitative prediction model of COD concentration based on generative adversarial networks (GANs) algorithm is proposed, which combines ultraviolet (UV) and Near Infrared (NIR) spectra with data-level fusion (DLDF) and feature level data fusion (FLDF). In this paper, firstly, COD standard samples are prepared according to a certain concentration gradient, and the ultraviolet spectrum (190 similar to 310 nm) and near-infrared spectrum (830 similar to 2100 nm) of the standard sample are collected respectively. The first derivative and Savitzky-Golay (S-G) smoothing pretreatment of the obtained ultraviolet and near-infrared spectrum data are carried out to eliminate the baseline drift of the spectrum and the interference noise. Then, the data fusion of data level and featural level are carried out directly basing on the pretreated ultraviolet and near-infrared spectra, and the COD concentration prediction model is constructed by GANs algorithm. The model is evaluated by using the square of the correlation coefficient of the evaluation parameters (R-2) , the mean square root error of the predicted value and the real concentration value (RMSEP) and the prediction deviation. The results show that neither FLDF model nor DLDF model is not ideal. The analysis shows that the model contribution of the ultraviolet spectrum is concealed in the near-infrared band due to the unbalanced data in the ultraviolet and near-infrared bands, which makes the spectral fusion meaningless. In order to avoid the problem of fusion failure, the normalizat-ion method is proposed to deal with the mixed spectrum in the text. The effects of standard normal variation (SNV), maximum and minimum normalization (MMN) and vector normalization (VN) on the modeling are discussed. Then the normalized ultraviolet and near-infrared spectral data are fused again under the given sub-interval number, the input X of GAN model is taken as the input X, and the real measured COD value is taken as the output Y. The prediction models of COD concentration are established after different normalization methods. The modeling results show that different normalization methods have a great influence on the hybrid spectral data fusion model, and the prediction accuracy of the data-level fusion model and the feature-level fusion model is significantly improved before it is normalized, among which the prediction model with the maximum and minimum normalization is the most obvious. Finally, in order to verify the accuracy of the multi-spectral data fusion GANs Prediction model, the GANs prediction model of the full wavelength ultraviolet band of a single spectral source and the GANs prediction model of the full wavelength near-infrared band of a single spectral source are established. The experimental results show that the correlation coefficient of the characteristic level spectral fusion model basing on the ultraviolet and near-infrared spectra is 0. 994 7, the prediction mean square root error is 0. 976, the prediction model error is reduced by 52. 9% comparing with the data level fusion, and the predicted recovery rate is 98. 4%similar to 103. 1%, which is much better than the other groups. The generalization ability of the model is strong and the prediction accuracy is high. Compared with the monitoring model of single spectral source, the data fusion of mixed spectra can reflect more the chemical information of water samples, and reveals the pollutant degree of a water body more comprehensively, reflects the difference of pollutants in a water body from different levels, provides some technical support for on-line monitoring of COD concentration in water.

为定量反演丹江口水库水质指标含量,明晰水质指标的时空分布特征、迁移转化规律,以南水北调中线工程水源地丹江口水库为研究对象,根据哨兵2号卫星(Sentinel-2)遥感影像不同波段组合的反射率,结合2016年2月的采样点总氮(TN)与氨氮(NH3-N)水质监测数据建立BP神经网络模型,反演2016—2020年TN与NH3-N含量,以此分析库区TN与NH3-N含量的时空变化特征,并探析其变化的影响因素.结果表明:构建的BP神经网络模型中TN和NH3-N的拟合精度均较高,R2分别为0.863和0.877,适用于丹江口水库TN和NH3-N遥感反演研究.2016—2020年丹江口水库水质整体呈向好趋势,NH3-N含量保持Ⅰ类水质标准,而TN含量在Ⅲ类与Ⅳ类水质标准之间.研究表明,利用Sentinel-2影像波段所建立的BP神经网络模型,适用于TN与NH3-N含量的遥感反演,以此分析不同季节适合的反演模型,可以为大型湖泊水生态环境改善及水质监管提供技术支撑.

为了研究曹妃甸地区的大气污染状况,了解其变化特征及其影响因素,采用多轴差分吸收光谱技术(MAX-DOAS),于2019年12月至2020年8月对曹妃甸地区NO2、O3、SO2三类主要污染物进行了监测分析,初步分析结果表明:曹妃甸地区NO2浓度在春季最高,均值为2.86×1016molec/cm2,O3浓度最高值出现在夏季.在日变化上,O3与NO2之间存在明显的负相关性,而O3浓度的峰值时间较太阳光照强度峰值存在滞后性;SO2的浓度分布呈"U"型,峰值出现在早晨与傍晚.8月期间,曹妃甸地区主要盛行东南风,风速多处于2～4m/s,此风向期间O3与SO2浓度较高,NO2主要来自于局地污染源排放.四个离轴观测角的差分斜柱浓度(DSCDs)结果显示,污染物主要分布于对流层底部,其中NO2主要集中于0～0.5km高度内.