A method, system and device for analyzing and estimating pollution load of a water area in a lake region

Abstract

The application discloses a kind of lake area water pollution load analysis estimation method, system and device, and the application constructs a time-space dynamic transport attenuation coefficient model.The model will attenuation coefficient be decomposed into two parts: one is based on land use, slope, soil and other multi-source geographic data, through multi-factor weighted model to calculate a spatial basic attenuation coefficient, to represent the spatial heterogeneity of attenuation ability;Two is to build a time dynamic adjustment function through rainfall, vegetation index, temperature and other time series data, to reflect the seasonal variation of attenuation ability.First, estimate the load generated by pollution source, then use geographic information system to track the pollution transport path, and along the path according to time-space dynamic transport attenuation coefficient, through the series attenuation model to calculate the total path attenuation coefficient.Finally, the load generated and path attenuation coefficient are combined to calculate the pollution load into lake.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of water environment science and environmental information technology, and particularly relates to a lake area water pollution load analysis and estimation method, system and device. BACKGROUND

[0002] Lakes are an important part of surface water resources, and have multiple important ecological functions such as water supply, climate regulation, and maintenance of biodiversity. However, with the rapid development of social economy, a large amount of nutrients such as nitrogen and phosphorus and other pollutants generated by human activities are discharged into lakes, leading to increasingly serious eutrophication of lake water bodies, frequent blue-green algae blooms, and serious threats to the health of lake ecosystems and drinking water safety.

[0003] Accurate estimation of pollution load into lakes is a prerequisite and scientific basis for effectively carrying out lake water environment protection and management. Pollution load into lakes mainly includes point source pollution and non-point source pollution. Point source pollution mainly refers to the centralized discharge of industrial wastewater and urban domestic sewage, which has a fixed location and relatively stable discharge amount, and is relatively easy to monitor and quantify. Non-point source pollution is derived from a wide range of sources, such as loss of fertilizers and pesticides in farmland, scattered discharge of rural domestic sewage, atmospheric deposition, and urban surface runoff, etc., which has randomness, universality and hysteresis, and is the focus and difficulty in current lake pollution load estimation.

[0004] At present, the methods for estimating pollution load in a watershed at home and abroad mainly include the following two types:

[0005] The first type is the output coefficient model. This method is based on the assumption that different land use types (such as farmland, forest land, and construction land, etc.) have different unit area pollutant output intensities. By dividing the study area into different land use units, the area is multiplied by the corresponding output coefficient value and added up to obtain the total pollution load. For example, the famous Johnes model is a representative of this type of method. The advantages of this method are simple structure, small amount of data required, and easy to operate. However, its fundamental defect is that it greatly simplifies the complex pollution generation-transportation-transformation process, assigns a static and uniform coefficient value to the same type of land use, and completely ignores the following key factors:

[0006] Spatial heterogeneity: Within the same land use type, there are great differences in topography (such as slope and aspect), soil properties (such as permeability and texture), and vegetation coverage, etc. These differences directly affect the loss and migration ability of pollutants.

[0007] Attenuation effect during transport: Pollutants will undergo a long transport process (e.g., surface runoff, ditch, underground seepage, etc.) from the source to the final entry into the lake. During the transport process, pollutants will be significantly reduced and attenuated due to the effects of vegetation absorption, soil adsorption, microbial transformation (such as denitrification), physical sedimentation, etc. The traditional export coefficient model fails to effectively quantify this process, resulting in overestimation of the estimated results and large deviations from the actual lake inflow. For example, when estimating agricultural non-point source pollution in Poyang Lake, researchers usually estimate the total amount of production and then assume a "natural reduction of 50%" or "aquaculture reduction of 25%" coefficient based on literature or experience. Although this method considers attenuation, the coefficient is static and non-spatial, and cannot reflect the real and complex geographical processes.

[0008] Special nature of the lake area: There are usually a large number of lake buffer zones around the lake area, such as wetlands, mudflats, and aquatic vegetation belts. These areas are the last and most critical ecological barrier for intercepting and purifying pollutants entering the lake. Traditional methods cannot accurately depict the purification capacity of this barrier.

[0009] The second type is the mechanism process model. This type of model, such as SWAT (Soil and Water Assessment Tool), HSPF (Hydrological Simulation Program-Fortran), and AnnAGNPS (Annual Agricultural Non-Point Source Pollution Model), is based on detailed physical, chemical, and biological processes to continuously and dynamically simulate hydrological cycles (rainfall, evaporation, infiltration, runoff), sediment erosion, and nutrient form transformation and migration in the watershed. This type of model has a solid theoretical foundation and can provide high spatial and temporal resolution simulation results, making it a powerful tool for watershed water environment research. However, its widespread application, especially in practical management-oriented work, also faces significant challenges:

[0010] Strong data dependence: The construction and calibration of the model require a large amount of long-term, high-quality input data, including but not limited to daily or smaller time scale meteorological data (rainfall, temperature, wind speed, humidity, solar radiation), high-precision digital elevation models (DEM), detailed soil physical and chemical property databases (layered texture, organic matter, bulk density, permeability coefficient, etc.), detailed land management measures parameters (crop rotation, irrigation system, fertilizer application time and amount, etc.), hydrological monitoring data (river section flow, sediment, water quality concentration), etc. In many areas, especially in developing regions with weak data infrastructure, it is extremely difficult and even impossible to obtain these data comprehensively.

[0011] Model complexity and uncertainty: The model has a large number of internal parameters, and the parameter calibration process is complex and requires professional knowledge. The model may be sensitive to small errors in input data, resulting in uncertainty in the results. For managers without a professional background, such models are like a "black box", and there is a high threshold for their operation and interpretation of results.

[0012] In summary, there is a clear gap in the existing method system: on the one hand, the output coefficient model leads to distorted results due to its oversimplification, making it difficult to meet the demand for precise pollution control; on the other hand, the complex mechanism model is difficult to be widely applied in areas that need rapid assessment or have limited data due to its strict requirements for data and complexity of operation. In particular, for the specific geographical unit of "lake area", the existing methods generally fail to fully and finely quantify its core geographical feature, i.e. the significant and spatio-temporal dynamic attenuation effect of pollutants on the "last transport path" before they enter the main body of the lake from the buffer zone around the lake. SUMMARY

[0013] To solve the above technical problems, the present application discloses a lake area pollution load analysis and estimation method, characterized in that the method comprises the following steps:

[0014] Step 1: Research area definition and multi-source heterogeneous data preparation.

[0015] 1. Research area definition: Based on the digital elevation model (DEM), the hydrological analysis module of the GIS platform is used to automatically or with human assistance to accurately extract the catchment boundary of the target lake through a series of standard processes such as filling, flow direction calculation, flow accumulation, etc. The land area surrounded by the boundary is the research area for this estimation.

[0016] 2. Data preparation, processing and spatialization: Collect and preprocess the following multi-source heterogeneous data within the research area systematically, including format conversion, coordinate system unification, spatial registration, cropping and rasterization, to form a standardized spatial database.

[0017] Geospatial basic data:

[0018] Digital Elevation Model (DEM): used to extract terrain factors such as slope and slope direction, and as the basis for hydrological path analysis.

[0019] Land use / cover (LULC) status map: key attenuation capacity characterization data, which needs to include fine classification of farmland, forest land, grassland, construction land, water area, wetland, bare land, etc.

[0020] Soil type distribution map: provides information on soil texture, permeability, organic matter content, etc., for assessing the adsorption and filtration capacity of soil for pollutants.

[0021] Drainage map: including rivers, ditches, reservoirs, ponds, etc. for analyzing hydrological connectivity.

[0022] Pollution source inventory data:

[0023] Point source: location (latitude and longitude), design treatment capacity, discharge standard, actual operation load rate, monthly or annual discharge flow and pollutant concentration monitoring data of industrial enterprises and urban sewage treatment plants.

[0024] Agricultural non-point source: planting area, average fertilizer rate (N, P), irrigation method of various crops (such as rice, corn, vegetables) in administrative units (such as towns, villages); breeding scale, stocking rate and manure treatment method of main livestock and poultry types (pigs, cattle, poultry).

[0025] Urban and rural life non-point source: data of permanent population and floating population, urbanization rate, per capita domestic sewage and garbage generation and discharge coefficient of each administrative unit.

[0026] Time series remote sensing and meteorological data:

[0027] Meteorological data: long-term (such as more than 10 years) monthly (or daily) rainfall and average temperature data of one or more representative meteorological stations in the study area.

[0028] Remote sensing vegetation index: long time series (such as MODIS, Landsat) remote sensing images covering the study area, and monthly or seasonal time series grid data of normalized vegetation index (NDVI) are extracted from them to dynamically represent vegetation growth conditions and coverage.

[0029] Engineering facilities and management data:

[0030] Urban drainage network map: the scope of combined rainwater and sewage system and separated rainwater and sewage system should be clearly marked, and the location of main pump stations and overflow outlets (CSO) should be provided.

[0031] Farmland water conservancy facilities distribution map: including the distribution of main irrigation and drainage trunks, branch trunks, as well as the spatial range of ecological ditches, drip irrigation / sprinkler irrigation and other water-saving and efficient agricultural areas.

[0032] Step two: spatial estimation of pollution source load based on inventory method ).

[0033] According to the type of pollution source, inventory analysis method or pollution coefficient method is used to estimate the total amount of pollutants generated in each grid in the study area at a specific time step (such as year, season, month).

[0034] 1. Point source load: the discharge of point source is allocated to the grid unit where it is located according to its geographical coordinates. .

[0035] 2. Agricultural non-point source load: Based on the spatial distribution of land use types (such as paddy fields, dry land), the administrative statistical data are spatially decomposed. For example, the N fertilizer load generated by rice planting in a certain town can be allocated according to the area proportion of the paddy field grid within the town.

[0036] 3. Urban and rural life non-point source load: Combine population data with land use types (construction land) or night light remote sensing data and other auxiliary information to allocate life pollution load more finely to the grid where the residential area is located.

[0037] 4. Atmospheric deposition load: Based on the dry and wet deposition rate of regional monitoring or literature research, multiply the area of the lake surface and the land grid to calculate.

[0038] Finally, all source data are overlaid in GIS to form a comprehensive "pollution source load spatial distribution map" covering the entire study area. ).

[0039] Step three: Build a time-space dynamic transport and attenuation coefficient model ).

[0040] This step is the core innovation of the present application, which aims to quantify the proportion of pollutants reduced by the environment during the process from the source to the water body. The value range of the transport and attenuation coefficient , , represents that the pollutant is not reduced at all and is transported to the downstream (such as the direct outlet of the hardened ground); represents that the pollutant is completely reduced and has no contribution to the downstream (such as the source being completely purified by the high-efficiency wetland). The built by the present application is a two-dimensional variable of space and time, which is determined by the static spatial basic attenuation ability and the dynamic time adjustment factor.

[0041] ;

[0042] wherein, is the geographical coordinates of the grid unit, is the time dimension (such as month or season).

[0043] 3.1 Build spatial basic attenuation coefficient ).

[0044] It aims to finely depict the spatial heterogeneity of the inherent reduction ability of pollutants determined by relatively stable geographical environmental elements. It is a comprehensive index calculated for each grid unit, which is obtained by a multi-factor weighted linear combination model:

[0045] ;

[0046] wherein:

[0047] : the row base attenuation coefficient value of the grid cell.

[0048] : the number of key geographical factors that affect the attenuation effect. The present application at least considers the following four types of factors:

[0049] Land use / cover type (LULC): Different LULC has a huge difference in the interception, absorption, and conversion ability of pollutants. The root system and litter layer of vegetation (especially forest land, wetland, and grassland) can effectively intercept surface runoff, adsorb pollutants, and promote nutrient absorption, thus having strong attenuation ability; on the contrary, hard or semi-hard surfaces such as construction land and bare land have fast runoff speed and weak purification ability.

[0050] Terrain factor (taking slope as an example): extracted from DEM. Slope directly affects the flow rate and erosion force of surface runoff. The steeper the slope, the faster the runoff flow rate, the shorter the contact time between pollutants and surface media (soil and vegetation), and the fewer opportunities for adsorption, sedimentation, absorption, and other purification processes, thus the weaker the attenuation ability.

[0051] Soil properties (Soil Type): Soil is an important carrier and reactor for pollutant migration. The permeability, organic matter content, and clay content of soil jointly determine its adsorption and filtration ability for pollutants. For example, loam soil with moderate texture and rich organic matter generally has better comprehensive purification ability than sandy soil with fast permeability and poor adsorption, or clay soil prone to surface runoff.

[0052] Hydrological connectivity: represents the degree of connection convenience between the plot and the water system network. The present application can use the "distance to the nearest water system" or more complex indicators such as DEM-based water flow path and cumulative flow to represent. If a plot is directly adjacent to a river, its risk of pollutant entering the river is high, and the attenuation path is short; if it needs to pass through multiple levels of ponds, ditches, and wetlands for filtration, the attenuation efficiency is high.

[0053] : the row specific quantified value of the th factor on the grid cell (e.g., slope of 5 degrees, LULC type code of 3, etc.).

[0054] : the A function to normalize the original values of the factors. Since the dimensions and value ranges of the factors are different, they need to be converted into dimensionless evaluation values (usually between 0 and 1). For example, for slope, the evaluation value is inversely proportional to the original value; for vegetation coverage, the evaluation value is directly proportional to the original value.

[0055] : the weight of the th factor, indicating the importance of the factor's contribution to the total attenuation capacity, and . The determination of the weight is the key to the scientificity of the model, and the present application provides various determination methods: the analytic hierarchy process (AHP) can be used to organize experts in the fields of hydrology, soil, ecology, etc. to compare and judge each factor pairwise, build a judgment matrix, and calculate the subjective weight vector. This method can systematically and quantitatively incorporate expert knowledge; in addition to the AHP method, those skilled in the art can also use objective weighting methods based on the variability of the data itself, such as entropy weight method, principal component analysis method, or a combination of subjective and objective methods (such as AHP-entropy weight method).

[0056] 3.2 Construction of time dynamic adjustment function (f ).

[0057] It is designed to reflect the time dynamics of the pollutant attenuation capacity caused by the seasonal changes of environmental driving factors (mainly climate and vegetation). It is a dimensionless adjustment coefficient that changes with time (month or season), and it can also be a comprehensive function itself:

[0058] ;

[0059] wherein:

[0060] : the comprehensive time adjustment function value at moment. Its value greater than 1 indicates that the environmental conditions are conducive to purification in this period, and the attenuation capacity is enhanced; less than 1 indicates that the attenuation capacity is weakened.

[0061] : adjustment sub-function driven by rainfall. Strong rainfall events in the wet season (such as summer) will cause surface runoff to surge and flow rate to increase, and the runoff scouring force will be strong, the residence time of pollutants in buffer media such as vegetation and soil will be dramatically shortened, and even a "short circuit flow" may be formed due to surface saturation, resulting in a significant decrease in overall attenuation capacity. Therefore, the function value has a clear negative correlation with rainfall or runoff. One possible mathematical form is: , wherein is the rainfall at t, and are the minimum and maximum rainfall in the study period.

[0062] The adjustment function driven by vegetation growth status, NDVI is an internationally recognized remote sensing index characterizing vegetation growth vitality and coverage. During their peak growing season (usually spring and summer), plants exhibit the strongest photosynthetic and metabolic activities, reaching their peak capacity for absorbing and assimilating nutrients such as nitrogen and phosphorus, thus maximizing their ability to reduce pollutants. This function value shows a significant positive correlation with the NDVI value.

[0063] Temperature-driven adjustment function. Temperature is a key environmental factor affecting the rate of microbial life activities, especially the activity of denitrifying bacteria (which remove nitrogen) and polyphosphate-accumulating bacteria (which remove phosphorus) in soil and water. Within a suitable range (e.g., 5-35℃), increased temperature significantly enhances the transformation efficiency of microorganisms, thereby improving their nitrogen and phosphorus removal capacity.

[0064] These represent the relative importance weights of seasonal driving factors such as rainfall, vegetation, and temperature, which can also be determined through methods such as expert scoring or multiple regression analysis.

[0065] By following the steps above, a unique, dynamically changing transport attenuation coefficient value can be calculated for each grid cell in the study area at each time step (e.g., each month). This forms a series of spatiotemporally dynamic attenuation coefficient raster layers.

[0066] Step 4: GIS-based transport path analysis and lake inflow load integration calculation.

[0067] 1. Delineation of Transport Paths and Accumulation of Attenuation Effects: For each raster cell generating non-point source pollution within the study area, GIS hydrological analysis tools (based on flow direction raster calculated from DEM) can be used to trace the shortest or most probable path of pollutants flowing to the lake within the surface runoff network. Pollutants are not attenuated by only one raster cell at their point of origin, but rather, along their flow path, they pass through multiple raster cells with different attenuation effects. The value of the raster cell. Total path attenuation coefficient ( This represents the cumulative effect of attenuation across all grid cells along the path. This invention employs a physically clear series attenuation model for calculation:

[0068] ;

[0069] in, The first line representing the path of the pollution source towards the lake One grid, is the total number of grids that the path goes through. The physical meaning of this formula is: the remaining pollutants after the reduction of the previous grid, and then the same proportion of reduction of the next grid, and so on, until the end of the path.

[0070] 2. Special path refinement:

[0071] Urban drainage network path correction: For the built-up area of a town, the surface runoff path does not completely follow the natural terrain. The invention combines the pipe network map to correct the path.

[0072] Rain and sewage separation area: The final discharge port of the rainwater pipe network is considered as the end point of a "fast channel". The decay coefficient on its transport path can be assigned a fixed value based on pipe wall adsorption and sedimentation, or still calculated according to the surface runoff but with very low weight.

[0073] Combined rain and sewage area: Focus on identifying combined sewer overflow (CSO). When a storm event occurs that exceeds the capacity of the pipe network, rainwater mixed with high concentration of domestic sewage will directly overflow into the adjacent water body through CSO, almost without any treatment. In this model, this means that the pollution from the service area of CSO, during the overflow period, is considered to bypass all natural and artificial purification facilities completely. Therefore, its corresponding path decay coefficient should be forced to be assigned a value of 0, so as to accurately simulate the explosive pollution of urban point and area into the lake during the rainstorm period.

[0074] Agricultural irrigation and drainage system path correction: Combined with water conservancy facility map, the transport path of agricultural non-point source is refined.

[0075] For farmland with ecological ditches or vegetation buffer zones, the initial section of the water flow path, i.e. the grid where the ditch or buffer zone is located, should be assigned a higher specific value based on the research of the purification efficiency of the facility.

[0076] For areas using drip irrigation, sprinkler irrigation and other efficient water-saving irrigation techniques, the generated surface runoff pollution is very little, and the main path of pollution may be converted to the underground infiltration-soil flow-underground water path. For this, another set of decay model based on soil infiltration and groundwater migration with higher time lag and attenuation effect should be switched to.

[0077] 3. Final lake load integration calculation: Perform grid algebraic operation on the "pollution source generated load spatial distribution map" ( ) obtained in step two and the "path total decay coefficient spatial distribution map" ( ) calculated in this step. The final load contribution of each source grid to the lake is ​​The contribution amount of all source grids is summed up in the whole study area, and the total pollution load of the whole study area at the moment is obtained . .

[0078] ;

[0079] wherein, represents the i-th pollution source grid in the study area, is the total number of grids.

[0080] The application also provides a lake water pollution load analysis and estimation system, which is based on a computer and a GIS platform and is used for automatically executing the foregoing method. The system at least comprises:

[0081] A data preparation and management module is used for receiving, processing and managing various source data required for executing the estimation, performing pretreatment such as format conversion, spatial registration and gridding on the data, and providing standardized data input for subsequent modules.

[0082] A pollution source load generation estimation module is connected with the data preparation and management module, adopts the list method or the coefficient method to calculate and generate a pollution source load spatial distribution layer according to the processed pollution source list data and the related coefficients.

[0083] A space-time dynamic transport and attenuation coefficient modeling module is a core module of the system and is used for constructing a transport and attenuation coefficient model. The module further comprises:

[0084] A spatial basic attenuation coefficient construction submodule is used for calculating the spatial basic attenuation coefficient of each grid unit according to geographic spatial data (LULC, slope, soil, etc.) through a multi-factor weighted model .

[0085] A time dynamic adjustment function construction submodule is used for constructing an adjustment function reflecting seasonal changes according to time series data (rainfall, NDVI, air temperature) .

[0086] A space-time dynamic coefficient synthesis submodule is used for multiplying the spatial basic attenuation coefficient and the adjustment function to generate a series of space-time dynamic transport and attenuation coefficients layers changing dynamically with time .

[0087] A transport path analysis and load integration module is connected with the foregoing modules and is used for completing the final lake load estimation. The module comprises:

[0088] ​​The transport path analysis sub-module: based on DEM, analyzes the pollutant transport path of each area source pollution grid, and can correct the path in combination with engineering facility data.

[0089] The path attenuation accumulation calculation sub-module: along the analyzed transport path, according to the formula , the total path attenuation coefficient of each pollution source is calculated .

[0090] The lake inflow load integration calculation sub-module: according to the formula , the final total lake inflow pollution load is integrated and calculated, and the result visualization mapping, report output and space-time analysis are supported.

[0091] The present application also provides a lake area water pollution load analysis and estimation device, which can be a general or special computing device (such as a server, workstation) containing a central processing unit (CPU), a graphics processing unit (GPU), a memory (RAM, hard disk) and an input / output interface. The memory stores computer executable program instructions, which can completely implement all steps of the aforementioned pollution load analysis and estimation method when the processor executes these instructions. The functional units of the device can be logically divided into: data acquisition unit, initial load calculation unit, attenuation coefficient calculation unit and load integration calculation unit, which correspond to the main calculation links in the method.

[0092] The beneficial technical effects of the present application are as follows: by constructing a space-time dynamic transport and attenuation coefficient model, the attenuation of pollutants in the transport process is divided into a spatial basic attenuation coefficient determined by geographical elements and a time dynamic adjustment function driven by seasonal environmental factors, so that the spatial heterogeneity and time dynamics of the pollutant attenuation ability can be quantitatively characterized. Compared with the traditional method of using fixed attenuation coefficient, the consistency of the lake pollution load estimation result and the physical process is improved; at the same time, the present application can generate spatial distribution maps of pollution sources, attenuation coefficients and lake inflow load contributions. These spatialized results can identify the key paths of pollution transport and the areas that play a major purification role, providing quantitative spatial decision basis for the site selection and layout of pollution prevention measures; in addition, the basic data acquisition approach relied on by this method is relatively extensive, and under the premise of maintaining mechanism, compared with the complex mechanism model with strict data requirements, it has better applicability. DETAILED DESCRIPTION

[0093] The technical solutions of the present application will be described comprehensively and in detail below in combination with a specific embodiment. It should be noted that the present embodiment is only one of the many specific embodiments of the present application, and is intended to fully demonstrate the present application by taking the lakeside area of Poyang Lake as an example. Those skilled in the art can make adaptive modifications or combinations to the model parameters, factor selection, weight method, etc. under the guidance of the core idea of the present application, and these changes should not deviate from the scope of the present application.

[0094] The water environment quality of Poyang Lake is crucial to the ecological safety of the middle and lower reaches of the Yangtze River. In recent years, the problem of eutrophication has become prominent, and agricultural non-point source pollution is considered to be the main contributor. The present embodiment aims to apply the method of the present application to finely estimate the rural agricultural non-point source TN load into Poyang Lake in the lakeside area of Poyang Lake, and to further analyze the spatio-temporal dynamic characteristics.

[0095] Step one: definition of the study area and preparation of basic data:

[0096] 1. Study area: the study area of the present embodiment is defined as the lakeside area of Poyang Lake, which is the direct catchment area of pollutants into the main lake body of Poyang Lake. This area mainly considers all or part of the areas of 23 counties (cities, districts) involved in the “five rivers and seven estuaries” (the lower plain area of the five major rivers of Xiuhe, Ganjiang, Fuhe, Xinjiang and Raohe, and other direct estuaries into the lake), with a total area of about 23,400 square kilometers. The area is relatively flat, with dense river network, and land use is mainly agriculture and water area, which is a hot area of non-point source pollution.

[0097] 2. Data preparation and processing:

[0098] DEM: 30-meter spatial resolution DEM data released by SRTM is used. In the GIS software, the data is preprocessed by splicing, cutting, filling, etc.

[0099] LULC: Based on Landsat TM / ETM+ remote sensing images, a land use map of the study area is made by combining supervised classification and visual interpretation. The main types include: paddy field, dry land, forest land, grassland, urban construction land, rural residential land, water area, beach / tidal flat, bare land, etc.

[0100] Soil map: the main soil types (such as red soil, paddy soil, chao soil, etc.) and their physical and chemical properties (such as texture, organic matter content) in the study area are obtained.

[0101] Water system map: the water bodies of all levels are extracted from the LULC map, and the river network generated by hydrological analysis of DEM is combined to construct a complete water system network vector data.

[0102] Pollution source list: the TN and TP production of rural domestic sewage, domestic waste, farmland runoff, livestock and poultry breeding, and aquaculture is subdivided.

[0103] Time series data: Monthly rainfall and monthly average temperature data were collected from 15 meteorological stations in and around the study area; MODIS NDVI products (MOD13Q1, 250-meter resolution, 16-day composite) for the same period were downloaded from NASA’s website, and monthly NDVI raster data were generated using the maximum value composite method.

[0104] Engineering facility data: To conduct in-depth analysis, data on drainage systems in major towns within the study area and irrigation facilities information in some modern agricultural demonstration zones were collected.

[0105] Step Two: Pollution Source Load Generation ( Spatialization estimation:

[0106] According to the report data, the total TN load from rural and agricultural non-point source pollution in the Poyang Lake area in a certain year was 40,433.54 tons. To spatialize this total data into each 30-meter grid, the following strategy was adopted:

[0107] 1. Source Separation: The report indicates that in the composition of TN (Total Nitrogen) load, livestock and poultry farming accounts for 46.2%, farmland runoff accounts for 23.4%, domestic waste accounts for 14.8%, aquaculture accounts for 8.1%, and domestic sewage accounts for 7.5%. The total load is first decomposed into the load of each source according to this proportion.

[0108] 2. Space allocation:

[0109] Farmland runoff load: The total farmland runoff load is weighted and allocated to each farmland grid based on the area of ​​the paddy field and dry land grids in the LULC map.

[0110] Livestock and poultry farming load: Based on the statistics of farming scale in each county and the distribution map of rural settlements, the pollution load of farming is mainly allocated to rural settlements and surrounding farmland areas.

[0111] Domestic waste / garbage load: Based on the area of ​​rural settlement patches, the rural domestic pollution load is allocated to the corresponding grid.

[0112] Aquaculture load: Based on the area of ​​aquaculture water bodies such as ponds and reservoirs in the LULC map, the load is allocated.

[0113] 3. Load Integration: In GIS, the load raster layers of the above-mentioned sub-sources are overlaid to generate the final "TN-generated load spatial distribution map" ( ).

[0114] Step 3: Construct a spatiotemporal dynamic transport attenuation coefficient model for the Poyang Lake lakeside area ( ):

[0115] 3.1 Constructing the spatial basis attenuation coefficient (

[0116] 1. Factor selection and standardization: LULC, slope, soil type, distance to river were selected as the core factors.

[0117] LULC standardization: According to the TN purification capacity of different land use types, an evaluation value between 0 and 1 was given. Through expert consultation and literature reference, the values were as follows: tidal flat / wetland (0.95), forest land (0.80), grassland (0.65), paddy field (0.40, considering denitrification), dry land (0.25), water area (0.10), rural residential land (0.05), urban construction land (0.02), bare land (0.0).

[0118] Slope standardization: The slope map was calculated from DEM. A negative linear function was used for standardization: The larger the slope, the lower the evaluation value.

[0119] Soil type standardization: According to soil texture, the TN adsorption and denitrification potential were valued: paddy soil (0.8), red soil (silty) (0.6), aquic soil (sandy) (0.3).

[0120] Distance to river standardization: The Euclidean distance (Distance) of each grid to the nearest water system was calculated. The farther the distance, the longer the purification path of pollutants. An attenuation exponential function was used for standardization: . The coefficient 0.005 is an empirical value commonly used in similar studies of plain river network areas in the field, which can also be optimized by fitting the measured water quality data of different distance samples in the study area.

[0121] 2. Weight determination (AHP method): Five experts in hydrology, soil, and ecology who have been engaged in the research of Poyang Lake Basin for a long time were invited to compare and score the relative importance of the above four factors, and to construct the judgment matrix.

[0122] ;

[0123] ;

[0124] Calculation process:

[0125] 1. Normalize the C matrix by column.

[0126] 2. Sum the normalized matrix by row.

[0127] 3. Normalize the sum vector to get the weight vector W.​

[0128] ;

[0129] Consistency check: largest eigenvalue , consistency index , consistency ratio , pass the test.

[0130] The maximum weight is: , , , .

[0131] 3. Calculate : In the grid calculator of GIS, apply the weighted sum formula:

[0132] .

[0133] 3.2 Deep analysis of seasonal dynamic estimation:

[0134] To fully demonstrate the dynamic analysis capability of the invention, we select two typical months, summer (July) and winter (January), to calculate their dynamic adjustment functions . Assume that the weight of each seasonal driving factor is: .

[0135] Summer (July) scenario:

[0136] Climate and vegetation characteristics: typical wet season, concentrated rainfall and often heavy rain (monthly rainfall 300 mm), high temperature (monthly mean temperature 30℃), and vegetation in the most lush growth period (NDVI value up to 0.8).

[0137] Subfunction calculation:

[0138] : Heavy rain leads to intensified erosion, and the attenuation ability is severely weakened, with a standardized evaluation value of 0.2.

[0139] : The vegetation is lush and has strong absorption capacity, with a standardized evaluation value of 0.9.

[0140] : High temperature is conducive to microbial denitrification, with a standardized evaluation value of 0.8.

[0141] Integrated time adjustment function value: .

[0142] Temporal and spatial dynamic attenuation coefficient: Results show that in summer, although the positive effects of vegetation and temperature are strong, the overwhelming negative effect of heavy rainfall leads to a nearly 50% decrease in the overall attenuation capacity compared to the annual average (1 as the baseline).

[0143] Winter (January) scenario:

[0144] Climate and vegetation characteristics: Typical dry season, with little rainfall (50 mm of monthly average rainfall) and low temperatures (5℃ of monthly average temperature), most of the vegetation is in dormancy or decay (NDVI value as low as 0.2).

[0145] Sub-function calculation:

[0146] : Little rainfall, weak surface runoff, and longer residence time for pollutants, with a standardized evaluation value of 0.9.

[0147] : Vegetation decay, weak absorption capacity, with a standardized evaluation value of 0.1.

[0148] : Low temperature inhibits microbial activity, with a standardized evaluation value of 0.2.

[0149] Integrated time adjustment function value: .

[0150] Temporal and spatial dynamic attenuation coefficient: Results show that in winter, although the runoff conditions are conducive to purification, the limiting effects of vegetation and temperature are more prominent, leading to a low level of overall attenuation capacity.

[0151] Through this dynamic analysis, we can reveal the different factors that control the attenuation process in different seasons and quantify the seasonal risk of pollution entering the lake.

[0152] Step four: transport path analysis and integration of lake load calculation:

[0153] 1. Annual load calculation: We first perform annual-scale calculations. Here we set an annual average adjustment coefficient as the baseline, directly using for calculation.

[0154] Path attenuation calculation: For all TN load-producing non-point source grids, use the GIS hydrological analysis module (based on flow direction grids) to track their water flow paths to the Poyang Lake body. For each path, apply the series attenuation model to calculate the total path attenuation coefficient: This calculation is automatically completed in the GIS background for all source grids, resulting in an "annual path total attenuation coefficient map" covering the entire lake area.

[0155] Final load calculation:

[0156] Grid operation between "TN production load spatial distribution map" and "annual path total attenuation coefficient map". Through area-weighted average of the value of the entire region , the TN comprehensive path attenuation coefficient of the region is about 0.47. This value is a scientific result calculated by the model based on spatial heterogeneity information such as geography, soil, vegetation, etc., rather than a simple experience.

[0157] Accordingly, the final annual load into the lake is estimated:

[0158] ;

[0159] 2. Special path processing deep analysis:

[0160] Assume that the old urban area of a county (such as Duchang County) in the study area is a combined rainwater and sewage system, and during a heavy rain event in summer, multiple CSO points in the county overflow.

[0161] Scenario setting: On a certain heavy rain day in July, the daily rainfall reaches 150 mm.

[0162] Model processing:

[0163] 1. Identify all pollution source grids within the combined system range of the old urban area of the county.

[0164] 2. During the time step of this heavy rain day, for these grids, the path attenuation coefficient is forced to be assigned a value of 0.

[0165] 3. For other areas, use to perform normal path attenuation calculation.

[0166] Result impact: Through this processing, the model can accurately simulate that on this heavy rain day, the mixed sewage (domestic sewage + surface runoff) produced by the old urban area of the county contributes 100% of the TN load of Poyang Lake, thereby capturing the instantaneous and severe impact of this extreme event on the total load into the lake. This has direct quantitative support for assessing the urgency and environmental benefits of urban drainage system reconstruction.​

Claims

1. A method for analyzing and estimating pollution load of a lake area water area, characterized by, The method comprises the following steps: Step 1: Research area definition and basic data preparation; Step two: pollution source load estimation, calculate the total amount of pollutants generated by each pollution source in the study area And form the pollution source load spatial distribution map; Step three: constructing a time-space dynamic transport attenuation coefficient model to calculate a time-space dynamic transport attenuation coefficient representing the proportion of pollutants reduced by the environment during transport wherein, , is a grid cell coordinate, is time, is a spatial base attenuation coefficient, is a time dynamic adjustment function, and the spatial base attenuation coefficient is constructed by a multi-factor weighted model, and the calculation formula is ; wherein, is a base attenuation coefficient value of a grid cell, is the number of key geographical factors affecting attenuation, is the weight of the th influencing factor, is a function for standardizing the value of the th factor, is the specific value of the th factor on the grid cell; the key geographical factors at least include one or more of land use / cover type, terrain factor, soil property, and hydrological connectivity; the time dynamic adjustment function is constructed by weighted summation of multiple adjustment sub-functions driven by seasonal changes in environmental factors, and the calculation formula is ; wherein, is an adjustment sub-function driven by rainfall, is an adjustment sub-function driven by normalized vegetation index, is an adjustment sub-function driven by temperature; are the weights of the respective seasonal driving factors. Step four: transport path analysis and integrated calculation of lake load, tracking the transport path of each pollution source to the lake, and calculating the total path attenuation coefficient along the path Finally, according to the formula The total lake pollution load The calculation method of the total path attenuation coefficient is to use a series attenuation model, and the calculation formula is ; wherein, represents the th grid on the path of the pollution source flowing to the lake, is the total number of grids passed by the path.

2. The method of claim 1, wherein, The weight The determination method includes one of analytic hierarchy process (AHP), entropy weight method or principal component analysis method, or a combination of these methods.

3. The method of claim 1, wherein, The step four further comprises a special path refinement step; when the transport path involves a municipal sewer network, the path is corrected in combination with the sewer network map; for a combined rainwater and sewage area, the overflow outlet of the combined system is identified, and in the event of a rainwater overflow event, the path attenuation coefficient corresponding to the pollution from the service range of the overflow outlet is reduced Forced assignment to 0.

4. A system for analyzing and estimating pollution load of a lake area water area, characterized by, The system is used for executing the method as claimed in any one of claims 1 to 3, and the system comprises: a data preparation and management module for receiving, processing and managing multi-source data required for performing estimation; a pollution source generation load estimation module for calculating and generating a pollution source generation load spatial distribution layer; a time-space dynamic transport and decay coefficient modeling module for constructing a transport and decay coefficient model and generating a time-space dynamic transport and decay coefficient layer; a transport path analysis and load integration module for analyzing a transport path, cumulatively calculating a path decay coefficient, and finally integrating and calculating a total lake pollution load.

5. The system of claim 4, wherein, The time-space dynamic transport and decay coefficient modeling module comprises: The spatial base attenuation coefficient construction submodule is configured to calculate the spatial base attenuation coefficient of each grid unit through a multi-factor weighting model according to the geographic spatial data ; The time dynamic adjustment function construction submodule is configured to construct an adjustment function reflecting seasonal changes according to time series data ; a spatiotemporal dynamic coefficient synthesis sub-module, configured to synthesize with each other, to generate a series of spatiotemporal dynamic transport attenuation coefficients layers.

6. The system of claim 5, wherein, The transport path analysis and load integration module comprises: a transport path analysis submodule for analyzing a pollution transport path of each area source pollution grid based on DEM, and being capable of correcting the path in combination with engineering facility data; A path attenuation accumulation calculation sub-module is configured to calculate, along the analyzed conveying path, a total path attenuation coefficient according to The layer adopts a series attenuation model to calculate a total path attenuation coefficient ; a lake load integration and calculation submodule for integrating a generation load and a path decay coefficient to calculate a total lake pollution load.

7. A device for analyzing and estimating pollution load in lake areas, characterized in that, The device is a computing device comprising a processor and a memory, and the memory stores computer program instructions, and when the processor executes the computer program instructions, the lake water pollution load analysis and estimation method as claimed in any one of claims 1 to 3 is realized.

Images