[Scale Effects of Landscape Pattern on Impacts of River Water Quality： A Meta-analysis].

The landscape pattern determines water pollution source and sink processes and plays an important role in regulating river water quality. Due to scale effects, studies on the relationship between landscape pattern and river water quality showed variance at different scales. However, there is still a lack of integrated study on the scale effect of landscape pattern and river water quality dynamics. This study collected 4 041 data from results of previous publications to address the characteristics of landscape pattern and river water quality dynamics at different scales and to identify the key temporal and spatial scales as well as landscape pattern indices for regulating river water quality. The results indicated that, compared to precipitation events, base flow periods, and interannual scales, the high-flow period was the key temporal scale for linking landscape pattern on river water quality. Compared to the watershed scale, the landscape pattern of buffer zones had a greater impact on river water quality. The high-flow period-buffer zone scale was the key spatiotemporal coupling scale for linking landscape pattern and river water quality. Compared to croplands, water bodies, grasslands, and the overall landscape of the watershed, the landscape pattern of forests and urban areas had a greater impact on river water quality. Fragmentation degree was the most important landscape pattern factor regulating river water quality. In river water quality management, it is important to focus on the landscape configuration of buffer zones, increase forest area, reduce patch density of forests and water bodies, and decrease the aggregation degree of urban areas.

[Nitrate Pollution Characteristics and Its Quantitative Source Identification of Major River Systems in China].

Accurate source identification/apportionment is essential for optimizing water NO3--N pollution control strategies. This study conducted a meta-analysis based on data from 167 rivers across China from 2000 to 2022 to analyze the spatial and temporal variation patterns of nitrate pollution in seven major river systems and to quantitatively identify the source composition of riverine nitrate. The average ρ（NO3--N） in the seven major river systems was （4.54±3.99） mg·L-1， with 9.6% of river ρ（NO3--N） exceeding 10 mg·L-1. The riverine ρ（NO3--N） in eastern China were higher than that in western China， and the highest concentration was observed in the Haihe River system. Additionally， tributaries experienced more serious NO3--N pollution than that in the main stream. The ρ（NO3--N） in most river systems in the dry season was higher than that in the wet season， except in the Yellow River system. There was significant nitrification in the Pearl River system， the middle and lower reaches of the Yellow River system， the middle reaches of the Liaohe River system， the Songhua River system， and the Haihe River system， whereas there was significant denitrification in the Yangtze River system， the Huaihe River system， and the lower reaches of the Pearl River system. Based on the dual stable isotopes-based MixSIAR model， the major NO3--N source was sewage/manure （ > 50%） in the Yangtze River system， Haihe River system， Liaohe River system， and Southeast River system. Soil nitrogen was the main NO3--N source in the Songhua River system （56.4%）， and the contribution of fertilizer nitrogen， soil nitrogen， and sewage/manure to NO3--N pollution in the Pearl River system， Huai River system， and Yellow River system was 20%-40%. The contribution rate of sewage/manure to NO3--N in the tributaries was higher than that in the main stream， whereas the contribution rate of soil nitrogen to NO3--N in the main stream was higher than that in the tributaries. The contribution rate of soil nitrogen， fertilizer nitrogen， and atmospheric deposition nitrogen to nitrate nitrogen in the wet season was higher than that in the dry season， whereas the contribution rate of sewage/manure to NO3--N pollution in the dry season was higher than that in the wet season. Therefore， point source pollution such as domestic and production sewage discharge should be controlled in the Haihe River system， the Yangtze River system， the Liaohe River system， the tributaries and the downstream main stream areas of Yellow River system， and the downstream area of the Pearl River system， whereas non-point source pollution caused by the loss of fertilizer and soil nitrogen should be controlled in the Huaihe River system， the Songhua River system， the middle reaches of the main stream area of the Yellow River system， and the middle and upper reaches of the Pearl River system. The results can provide a scientific basis for the effective control of nitrate pollution in the river systems in China.

[Estimation of Cropland Nitrogen Runoff Loss Loads in the Yangtze River Basin Based on the Machine Learning Approaches].

A quantitative understanding of cropland nitrogen (N) runoff loss is critical for developing efficient N pollution control strategies. Using correlation analysis, a structural equation model, variance decomposition, and machine learning methods, this study identified the primary influencing factors of total N (TN) runoff loss from uplands (n=570) and paddy (n=434) fields in the Yangtze River Basin (YRB) and then developed a machine learning-based prediction model to quantify cropland N runoff loss load. The results indicated that runoff depth, soil N content, and fertilizer addition rate were the major influencing factors of TN runoff loss from uplands, whereas TN runoff loss rate from paddy fields was mainly regulated by runoff depth and fertilizer addition rate. Among the four used machine learning methods, the prediction models based on the random forest algorithm presented the highest accuracy (R2=0.65-0.94) for predicting upland and paddy field TN runoff loss rates. The random forest algorithm based model estimated a total cropland TN loss load in the YRB of 0.47 Tg·a-1 (upland:0.25 Tg·a-1; paddy field:0.22 Tg·a-1) in 2013, with 58% of TN runoff loss load derived from the midstream and downstream regions. The models predicted that TN runoff loss loads from croplands in YRB would decrease by 2.4%-9.3% for five scenarios, with higher TN load reductions occurring from scenarios with decreased runoff amounts. To mitigate cropland N nonpoint source pollution in YRB, it is essential to integrate efficient water, fertilizer, and soil nutrient managements as well as to consider the midstream and downstream regions as the high priority area. The machine learning-based modeling method developed in this study overcame the difficulty of identifying the functional relationships between cropland TN loss rate and multiple influencing factors in developing relevant prediction models, providing a reliable method for estimating regional and watershed cropland TN loss load.

[Estimation of Nitrous Oxide Emission from River System Based on Water Discharge and Dissolved Nitrous Oxide Concentration].

Due to increasing active nitrogen pollution loads, river systems have become an important source of nitrous oxide (N2O) in many areas. Due to the lack of monitoring data in many studies as well as the difficulty in estimating intermediate parameters and expressing temporal-spatial variability in current methods, a high level of uncertainty remains in the estimates of riverine N2O emission quantity. Based on the monthly monitoring efforts conducted for 10 sampling sites across the Yonganxi River system in Zhejiang Province from June 2016 to July 2019, the temporal and spatial dynamics of riverine N2O dissolved concentrations ρ(N2O), N2O fluxes, and their influencing factors were addressed. A multiple regression model was then developed for predicating riverine N2O emission flux to estimate annual N2O emission quantity for the entire river system. The results indicated that observed riverine ρ(N2O) (0.03-2.14 µg·L-1) and the N2O fluxes[1.32-82.79 µg·(m2·h)-1] varied by 1-2 orders of magnitude of temporal-spatial variability. The temporal and spatial variability of ρ(N2O) were mainly influenced by the concentrations of nitrate, ammonia, and dissolved organic carbon, whereas the N2O emission fluxes were mainly affected by river water discharges and ρ(N2O). A multiple regression model that incorporates variables of river water discharge and ρ(N2O) could explain 90% of the variability in riverine N2O emission fluxes and has high accuracy. The model estimated N2O emission quantity from the entire Yonganxi River system of 3.67 t·a-1, with 29% from the main stream and 71% from the tributaries. The IPCC default emission factor method might greatly overestimate and underestimate N2O emission quantities for rivers impacted by low and high pressures of human activities, respectively. This study advances our quantitative understanding of N2O emission for the entire river system and provides a reference method for estimating riverine N2O emission with more accuracy.

[Dynamic response of riverine nitrate flux to net anthropogenic nitrogen inputs in a typical river in Zhejiang Province over the 1980-2010 period].

Based on long-term records of river water quality and discharge and nitrogen sources as well as the LOADEST model, annual riverine NO3(-)-N flux and net anthropogenic nitrogen input (NANI) were both estimated for a typical river catchment (2 474 km2) in Zhejiang Province over the 1980-2010 period. Historical trends in both riverine NO3(-) -N flux and NANI and their dynamic relationships were then fully addressed. Finally, the contributions of annual NANI, retained nitrogen pools, and natural background sources to riverine NO3(-)-N flux were indentified. Results indicated that both riverine NO3(-) -N flux and NANI showed parabolic changing trends with peak value of 5.74 kg x (hm2 x a) for flux and 77.5 kg x (hm2 x a)(-1) for NANI both occurring around 1998. In 1980-2010, net increase of riverine NO3(-) -N flux and NANI was -42% and -77%, respectively. Chemical nitrogen fertilizer application and atmospheric nitrogen deposition, which accounted for -48% and -40% of NANI, respectively, were the major sources of NANI. Although interannual change of riverine NO3(-) -N flux was significantly related to NANI (R2 = 0. 27 * *) as well as the chemical nitrogen fertilizer application amount (R2 = 0.32 * *), it showed higher dependence on the river water discharge (R2 = 0.79 * *) or precipitation (R2 = 0.63 * *), implying that annual riverine NO3(-) -N was not only originated from current year's NANI, but also derived from retained N pools that were ultimately derived from NANI in previous years. A regression model developed by incorporating both NANI and water discharge could account for 94% of the variability of annual NO3(-) -N flux. This model predicted that NO3(-) -N flux could have been reduced by -21% and -30% if the annual NANI and water discharge had been cut by 30%, respectively. Annual NANI, retained nitrogen pools, and natural background sources contributed to -53%, -24%, and -23% of the riverine NO3(-) -N flux, respectively, suggesting that -77% of flux was derived from anthropogenic nitrogen sources. Although observed long-term interannual change of riverine NO3(-) -N flux was dependent on the combined influences of NANI and hydroclimate, a more immediate reduction of riverine NO3(-) -N flux may result from interception strategies than from cutting nitrogen source inputs due to the contribution of retained nitrogen pools.

[Bivariate statistical model for calculating phosphorus input loads to the river from point and nonpoint sources].

Based on the hydrological difference between the point source (PS) and nonpoint source (NPS) pollution processes and the major influencing mechanism of in-stream retention processes, a bivariate statistical model was developed for relating river phosphorus load to river water flow rate and temperature. Using the calibrated and validated four model coefficients from in-stream monitoring data, monthly phosphorus input loads to the river from PS and NPS can be easily determined by the model. Compared to current hydrologica methods, this model takes the in-stream retention process and the upstream inflow term into consideration; thus it improves the knowledge on phosphorus pollution processes and can meet the requirements of both the district-based and watershed-based wate quality management patterns. Using this model, total phosphorus (TP) input load to the Changle River in Zhejiang Province was calculated. Results indicated that annual total TP input load was (54.6 +/- 11.9) t x a(-1) in 2004-2009, with upstream water inflow, PS and NPS contributing to 5% +/- 1%, 12% +/- 3% and 83% +/- 3%, respectively. The cumulative NPS TP input load during the high flow periods (i. e. , June, July, August and September) in summer accounted for 50% +/- 9% of the annual amount, increasing the alga blooming risk in downstream water bodies. Annual in-stream TP retention load was (4.5 +/- 0.1) t x a(-1) and occupied 9% +/- 2% of the total input load. The cumulative in-stream TP retention load during the summer periods (i. e. , June-September) accounted for 55% +/- 2% of the annual amount, indicating that in-stream retention function plays an important role in seasonal TP transport and transformation processes. This bivariate statistical model only requires commonly available in-stream monitoring data (i. e. , river phosphorus load, water flow rate and temperature) with no requirement of special software knowledge; thus it offers researchers an managers with a cost-effective tool for quantifying TP pollution processes in both district and watershed scales.

[Change traits of phosphorous consumption structure in China and their effects on environmental phosphorous loads].

Substance flow analysis was used to construct a model to analyze change traits of China's phosphorous (P) consumption structure from 1980 to 2008 and their influences on environmental phosphorous loads, then the correlation between several socioeconomic factors and phosphorous consumption pollution was investigated. It is found that phosphorous nutrient inputs of urban life and rural life on a per capita level climbed to 1.20 kg x a(-1) and 0.99 kg x a(-1) from 0.83 kg x a(-1) and 0.75 kg x a(-1) respectively, but phosphorous recycling ratios of urban life fell to 15.6% from 62.6%. P inputs of animal husbandry and planting also kept increasing, but the recycling ratio of the former decreased from 67.5% to 40.5%, meanwhile much P input of the latter was left in agricultural soil. Correlation coefficients were all above 0.90, indicating that population, urbanization level, development levels of planting and animal husbandry were important incentives for P consumption pollution in China. Environmental Kuznets curve showed that China still stayed in the early development stage, promoting economic growth at an expense of environmental quality. This study demonstrates that China's P consumption system is being transformed into a linear and open structure, and that P nutrient loss and environmental P loads increase continually.

[Inversion formula of one-dimensional water quality equation for the export loads of nonpoint sources pollution in headwater area].

An inversion formula for the export loads of nonpoint sources pollution in headwater area was established based on one-dimensional water quality equation, and it was used to calculate the pollution loads for tributaries in the headwater catchment of Laohutan Reservoir, in Huzhou City, Zhejiang Province of China. Monte Carlo method was adopted to determine the sensitivity about each input parameter in the inversion formula. Because each sensitive parameter can be measured directly in the inversion formula, so that this approach can decreased calculation error, which is often caused by the parameter estimation. Furthermore, the inversion formula can be adopted to calculate pollution loading on any time scale. Monthly nonpoint sources pollution export loads in 2007 were calculated by the model in the research catchment. Results showed that pollution loads in stream were significantly positive related with flow rates (r > 0. 90, p < 0.01), and the flow rate was the most sensitive factor in the model, followed by the nutrient concentration and background concentration at the stream end. While, comprehensive degradation coefficient and flow velocity contributed very little influence to the model uncertainty.

[Uncertainty analysis of water environmental capacity in the nonpoint source polluted river].

Based on the one-dimension model for river water environmental capacity (WEC) and the statistical analysis of the measured hydrological and water quality variables, a uncertainty analysis method for the WEC in nonpoint source polluted river was developed, which included the sensitivity analysis for input parameters of the model and the probability distributions analysis for the WEC using Monte Carlo simulation approach. The method, which described the uncertainty derived from the adopted information of the river system and the randomicity from the occurrence of nonpoint source pollution, could provide different WEC combined with reliabilities for different hydrological seasons. As a case study, the total nitrogen (TN) WEC in the Changle River located in southeast China was calculated using the method. Results indicated that the TN WEC with 90% of reliability were 487.9, 949.8 and 1392.8 kg x d(-1) in dry season, average season and flood season, respectively, and the dilution effect of river water flow accounted for the main content of WEC. In order to satisfy water quality target of the river, about 1258.3-3591.2 kg x d(-1) of current TN quantity that entered into the river should be reduced in watershed, and the largest reducing quantity of TN was occurred during flood season. The uncertainty method, which reflected hydrology and water quality variations in the nonpoint source polluted river, provided a more reliable and efficient method for the WEC calculation.

[Water environmental capacity calculation model for the rivers in drinking water source conservation area].

Based on the one-dimension model for water environmental capacity (WEC) in river, a new model for the WEC estimation in river-reservoir system was developed in drinking water source conservation area (DWSCA). In the new model, the concept was introduced that the water quality target of the rivers in DWSCA was determined by the water quality demand of reservoir for drinking water source. It implied that the WEC of the reservoir could be used as the water quality control target at the reach-end of the upstream rivers in DWSCA so that the problems for WEC estimation might be avoided that the differences of the standards for a water quality control target between in river and in reservoir, such as the criterions differences for total phosphorus (TP)/total nitrogen (TN) between in reservoir and in river according to the National Surface Water Quality Standard of China (GB 3838-2002), and the difference of designed hydrology conditions for WEC estimation between in reservoir and in river. The new model described the quantitative relationship between the WEC of drinking water source and of the river, and it factually expressed the continuity and interplay of these low water areas. As a case study, WEC for the rivers in DWSCA of Laohutan reservoir located in southeast China was estimated using the new model. Results indicated that the WEC for TN and TP was 65.05 t x a(-1) and 5.05 t x a(-1) in the rivers of the DWSCA, respectively. According to the WEC of Laohutan reservoir and current TN and TP quantity that entered into the rivers, about 33.86 t x a(-1) of current TN quantity should be reduced in the DWSCA, while there was 2.23 t x a(-1) of residual WEC of TP in the rivers. The modeling method was also widely applicable for the continuous water bodies with different water quality targets, especially for the situation of higher water quality control target in downstream water body than that in upstream.

[Estimation and allocation of water environmental capacity in nonpoint source polluted river].

Based on the investigation of the application and emission quantities (QAE) of total nitrogen (TN) and total phosphorus (TP) for nonpoint sources in river catchment' s area, included fertilizer applications, livestock and living pollutants emissions, the quantities of TN and TP entered the river were computed by means of export coefficient model in Changle River, southeast China. Self-purification capacities of TN and TP in the reach were also estimated in terms of input-output balance analysis method. According to the provisions of water function planning in the river, the water environment residual capacity (WERC) or the demand for reducing the application and emission (DRAE) of nitrogen and phosphorus in the corresponding catchment were monthly estimated, and WERC and DRAE were respectively allocated among the pollution sources. Results indicated that about 28.8% of TN loads and 51.2% of TP loads could be self-purified respectively in the reach, i. e., purification of 775.9 t a(-1) for TN and 30.9 t a(-1) for TP. Seasonal variations of the self-purification for the pollutants not only resulted from riverine hydrological and ecological conditions, but also affected by the pollution loading. According to the demand of the water quality protection in the reach, about 1581.0 t a(-1) QAE of TN had to reduce in Changle catchment. The maximum demand for the reducing QAE of TN was the fertilizer application (1047.4 t a(-1)), and the highest ratio for the reducing QAE of TN was livestock-poultry breeding (32.4%). There was about 2335.7 t a(-1) WERC for TP in the reach. The largest DRAE of nitrogen was during mid-water season and the least WERC of TP was during higher-water season.

Spatial and temporal variations of water quality in Cao-E River of eastern China.

Evaluation and analysis of water quality variations were performed with integrated consideration of water quality parameters, hydrological-meteorologic and anthropogenic factors in Cao-E River, Zhejiang Province of China. Cao-E River system has been polluted and the water quality of some reaches are inferior to Grade V according to National Surface Water Quality Standard of China (GB2002). However, mainly polluted indices of each tributary and mainstream are different. Total nitrogen (TN) and total phosphorus (TP) in the water are the main polluted indices for mainstream that varies from 1.52 to 45.85 mg/L and 0.02 to 4.02 mg/L, respectively. TN is the main polluted indices for Sub-watershed I, II, IV and V (0.76 to 18.27 mg/L). BOD5 (0.36 to 289.5 mg/L), CODMn (0.47 to 78.86 mg/L), TN (0.74 to 31.09 mg/L) and TP (0 to 3.75 mg/L) are the main polluted indices for Sub-watershed III. There are tow pollution types along the river including nonpoint source pollution and point source pollution types. Remarkably temporal variations with a few spatial variations occur in nonpoint pollution type reaches (including mainstream, Sub-watershed I and II) that mainly drained by arable field and/or dispersive rural dwelling district, and the maximum pollutant concentration appears in flooding seasons. It implied that the runoff increases the pollutant concentration of the water in the nonpoint pollution type reaches. On the other hand, remarkably spatial variations occur in the point pollution type reaches (include Sub-watershed III, IV and V) and the maximum pollutant concentration appears in urban reaches. The runoff always decreases the pollutant concentration of the river water in the seriously polluted reaches that drained by industrial point sewage. But for the point pollution reaches resulted from centralized town domestic sewage pipeline and from frequent shipping and digging sands, rainfall always increased the concentration of pollutant (TN) in the river water too. Pollution controls were respectively suggested for these tow types according to different pollution causes.