A method for optimizing land use in multi-outlet watersheds

Abstract

This application provides an optimized regulation method for land use in multi-outlet watersheds. Step 1: Construct a spatial watershed attribute regression model to quantitatively analyze the impact of various land use types on water ecological load within the study area. Step 2: Obtain the corresponding regulation potential based on Step 1; optimize the spatial watershed attribute regression model; use the optimized model to assess the land use regulation potential and ecological restoration regulation potential of the corresponding watershed. Step 3: Set land use regulation scenarios and ecological restoration regulation scenarios; use the optimized spatial watershed attribute regression model to simulate the environmental effects related to these scenarios and formulate optimized restoration measures. This application can simultaneously simulate multiple watershed outlets, comprehensively consider the relationship between water bodies and land areas, quantify the pollution level of pollutants, and intuitively reflect the impact of land use on pollution load.

Description

Technical Field

[0001] This application relates to the field of land use technology, specifically to an optimized regulation method for land use in multi-outflow watersheds. Background Technology

[0002] In the field of land use, non-point source pollution is becoming increasingly prominent, gradually becoming the dominant factor in the deterioration of water quality and eutrophication of surface water bodies such as rivers, lakes and reservoirs. Because the generation and transmission of non-point source pollution are random and intermittent, and the uncertainty is complex, how to carry out comprehensive management of non-point source pollution at the watershed scale has become a key and difficult point in the field of watershed water pollution control.

[0003] Land use change is one of the main factors affecting the aquatic ecosystem of a watershed. It alters processes such as nutrient accumulation, particle deposition, and hydrological conditions in water bodies, thereby impacting the aquatic ecosystem. Its impact is enormous and difficult to control and manage. However, land use is also one of the most controllable factors within a watershed. Territorial spatial planning clearly delineates the control areas and boundaries of urban, agricultural, and ecological spaces in various regions. Scientific and rational land planning can greatly improve and restore the hydrology and habitats of rivers, reduce pollution output, and intercept and transform land pollutants. Therefore, how to optimize and regulate land use in a watershed has become a hot topic.

[0004] In watershed management practices, numerical models are widely used because they can simulate pollutant migration and transformation processes based on pollution formation mechanisms, clarifying the spatiotemporal distribution patterns of pollutant transport. By constructing the relationship between pollution sources and water quality response through modeling, better decision-making support can be provided for refined watershed water environment management. Currently, numerous models have been applied in watershed management with good results. Among them, the spatial watershed attribute regression model is a water quality model that combines mechanistic and statistical models. It uses hybrid statistics and process-based methods to link monitoring data with watershed characteristics and pollutant source information to estimate pollutant migration in the watershed and water bodies, exploring the relationship between human activities, natural processes, and pollutant migration. It has advantages such as low data requirements, transparent structure, and strong universality. This model overcomes the problems of high data requirements, weak interpretability, and poor applicability in constructing complex watershed models. Furthermore, this model can simultaneously simulate multiple watershed outlets, comprehensively consider the relationship between water bodies and land areas in the watershed, analyze the current water environment status of typical functional zones, quantify the degree of pollutant pollution, and reflect the impact of land use on pollution load.

[0005] However, this model cannot assess the regulatory potential of different measures, and therefore cannot formulate a practical and effective spatial regulation scheme for the water environment of the control unit. This application aims to overcome the above-mentioned defects, and after a detailed search, no relevant technical solutions were found.

[0006] Therefore, a new technical solution is needed to solve the above-mentioned technical problems. Summary of the Invention

[0007] This application provides a method for optimizing and regulating land use in multi-outlet watersheds, comprising the following steps:

[0008] Step 1: Obtain basic watershed data for the study area, construct a spatial watershed attribute regression model, and quantitatively analyze the impact of various land use types on water ecological load within the watershed of the study area through the spatial watershed attribute regression model.

[0009] Step Two: Based on the impact of each land use type on water ecological load in Step One, and combined with the rigid boundaries in the territorial spatial planning of the study area, land use schemes and ecological restoration schemes are established. Then, an optimized spatial watershed attribute regression model is obtained by coupling a spatial watershed attribute regression model and an annealing algorithm. This optimized model is used to assess the land use regulation potential and ecological restoration regulation potential of the corresponding watersheds. The rigid boundaries in the territorial spatial planning of the study area are the three zones and three lines of the study area. The three zones are urban space, agricultural space, and ecological space; the three lines are the ecological protection red line, the permanent basic farmland protection red line, and the urban open boundary.

[0010] Step 3: Based on the land use regulation scenario and ecological restoration regulation scenario set in Step 2, use the optimized spatial watershed attribute regression model to simulate the environmental effects related to the land use regulation scenario and ecological restoration regulation scenario, and formulate optimized restoration measures.

[0011] As a preferred embodiment, obtaining the basic watershed data for the study area in step one includes:

[0012] S11: Determine the direction of water flow and the cumulative flow;

[0013] The flow direction of water in each grid is calculated based on the DEM, the number of all incoming water paths upstream of each grid is counted, the cumulative flow of the grid is obtained, and then the threshold of the cumulative flow is determined.

[0014] S12: Refinement and delineation of the river network;

[0015] Based on the cumulative runoff threshold, the river network is extracted from the cumulative runoff data. The nodes of the river network are determined based on the extracted river network. The nodes include the starting point of the river, the confluence point of the tributaries, and the river mouth.

[0016] S13: Extract sub-basin boundaries;

[0017] At each node of the river network, a water catchment outlet is set up. Based on the water flow direction data, all grids upstream of each water catchment outlet that flow towards that outlet are traced in reverse. The set of these grids is the sub-basin corresponding to that water catchment outlet.

[0018] S14: Obtain continuous monitoring hydrological and water quality data, meteorological data, soil condition data, land use data, and statistical data of various pollution sources within each sub-basin.

[0019] As a preferred embodiment, in step S12, the river network is extracted from the confluence accumulation data and then calibrated with existing river network data.

[0020] As a preferred embodiment, the calculation formula for the spatial watershed attribute regression model in step one is as follows:

[0021] in, This represents the pollution flux flowing through sub-basin i, expressed in kg / year. This indicates the upstream river segment adjacent to river segment in, where river segment in is one of the river segments divided from sub-basin i; Indicates the index of the upstream river section, This indicates the pollution flux in the upstream section of the river; This indicates the proportion of upstream flux transported to the river segment in. This represents the attenuation function of pollutant transport along the river channel. Different equations exist for rivers and lakes; if the water body is a river, then the corresponding equation is... and If the water body is a reservoir or lake, then the corresponding and ,in, The parameter matrix representing the attenuation of river transport. This represents a parameter matrix related to reservoir / lake transport attenuation; This represents the coefficients to be fitted for river transport attenuation. NS represents the coefficients to be fitted for the transport attenuation of a reservoir or lake; NS represents the total number of pollution sources in sub-basin i, where n is the index of the pollution source. These represent the amount produced by each pollution source. This represents the emission coefficient of the pollution source. Indicates the soil-water transport factor. Represents the parameter matrix, This represents the coefficient corresponding to parameter D. This represents the attenuation function of locally generated pollutants in this section of the river. It has different equation forms for rivers and lakes. If the water body is a river, then it corresponds to... and If the water body is a reservoir or lake, then the corresponding and , where θ represents the coefficient to be fitted.

[0022] As a preferred embodiment, the optimized spatial watershed attribute regression model in step two is obtained by setting constraints on the model based on the initially established spatial watershed attribute regression model as the objective function. The constraints are as follows:

[0023] ; Where N represents the number of sub-basins, and i is the index of the sub-basin. This represents the amount of pollution generated by each pollution source. Sopt represents the target value for the pollution source. Si and Sj are the upper and lower boundaries of the pollution source, respectively.

[0024] As a preferred embodiment, step three includes:

[0025] S31: The existing land type holdings and existing land use within the existing rigid boundaries shall be used as the control boundaries for each sub-basin, and the area for setting control measures shall reach the short-term or long-term goals.

[0026] S32: Using the established spatial watershed attribute regression model as the objective function, with the goal of minimizing the inflow load of each sub-watershed into the lake, the spatial priority restoration area for restoration measures is calculated using the optimized spatial watershed attribute regression model.

[0027] S33: Based on the spatial priority repair area, derive the spatial optimization repair scheme for each repair measure.

[0028] The present invention has the following beneficial effects:

[0029] I. Constructing a spatially based watershed attribute regression model requires relatively little data, has strong interpretability, and is suitable for use in most regions.

[0030] Second, the spatial watershed attribute regression model can simultaneously simulate multiple watershed outlets, comprehensively consider the relationship between water bodies and land areas in the watershed, quantify the degree of pollution of pollutants, and intuitively reflect the impact of land use on pollution load.

[0031] Third, based on the spatial watershed attribute regression model, this application uses the mature annealing algorithm to optimize the spatial watershed attribute regression model, which can evaluate different remediation measures and formulate practical and effective spatial remediation schemes for the water environment of control units, thereby guiding local areas to carry out targeted measures and countermeasures. Attached Figure Description

[0032] Figure 1 This is the logic block diagram of this application. Detailed Implementation

[0033] The following combination Figure 1 The specific embodiments of the present invention will be described in detail below. It should be noted that the specific embodiments described herein are for illustrative and explanatory purposes only and are not intended to limit the present invention. Example 1

[0034] This embodiment provides an optimized regulation method for land use in multi-outlet watersheds, comprising the following steps:

[0035] Step 1: Obtain basic watershed data for the study area and construct a spatial watershed attribute regression model, also known as the SPARROW model; use the spatial watershed attribute regression model to quantitatively analyze the impact of various land use types on water ecological load within the watershed of the study area.

[0036] The acquisition of basic watershed data for the study area specifically includes:

[0037] S11: Determine the direction of water flow and the cumulative flow;

[0038] The flow direction of each raster is calculated based on Digital Elevation Model (DEM) data downloaded from the Chinese Academy of Sciences scientific database. Based on the flow direction data, the number of all upstream flow paths entering each raster is counted to obtain the cumulative flow for that raster, thus determining the cumulative flow threshold. The calculation method for the cumulative flow of a raster can be similar to that in ArcGIS (Arc Geographic Information System), using the flow tool in hydrological analysis. Inputting the flow direction data and running the algorithm generates the cumulative flow raster data. A higher number of upstream flow paths for each raster indicates a potential river channel, i.e., a river network. The prototype is as follows: Preferably, before determining the water flow direction of each grid, the DEM data is first processed by filling depressions, such as using the depression filling tool in GIS software to process the DEM data; in ArcGIS, the depression filling tool in the hydrological analysis toolkit can be used to remove depressions in the DEM data and generate a depression-free DEM to ensure the accuracy of the water flow direction. After processing the DEM data, the water flow direction of each grid is calculated based on the depression-free DEM data using algorithms such as the maximum slope method. In ArcGIS, this can be achieved through the flow direction tool in the hydrological analysis module to obtain the water flow direction grid data.

[0039] Methods for calculating the confluence accumulation threshold include, but are not limited to: 1. Empirical judgment method: The threshold can be determined based on the topographic features of the study area and relevant research or industry experience in similar areas. For example, in mountainous areas with steep terrain and rapid runoff convergence, a small confluence accumulation may form a water flow path, so the threshold can be set to a smaller value, such as 500-2000 grid cells; in plains with flat terrain, a larger confluence accumulation is required to form a significant water flow, so the threshold can be set to a larger value, such as 5000-10000 grid cells or even larger; 2. Iterative trial method: By setting a series of different confluence accumulation thresholds, the river network is extracted, and then the extracted results are compared with the actual river network obtained through field surveys, high-resolution remote sensing image interpretation, etc., and the threshold that best matches the actual river network is selected; 3. Statistical analysis method: Using the statistical functions of GIS software, the relationship between confluence accumulation and parameters such as river length and number of tributaries is analyzed, scatter plots are drawn, and function fitting is performed. 1. Identify abrupt changes in the curve and use the corresponding cumulative flow value as a threshold. For example, as the threshold increases, the river length and number of tributaries initially increase slowly, then increase rapidly after reaching a certain value. This value can be used as a preliminary threshold reference, and then adjusted according to the actual situation. 2. Based on the river network density method: Determine the target river network density according to the research needs, that is, the river length per unit area. First, set different thresholds to extract the river network and calculate its river network density. Then, select the threshold whose river network density is closest to the target value as the final threshold. For example, if the research requires a medium-density river network, the threshold whose river network density meets the requirements can be found through experiments.

[0040] S12: Refinement and delineation of the river network;

[0041] Based on the confluence accumulation threshold, the river network is extracted from the confluence accumulation data. The nodes of the river network are determined based on the extracted river network, including the river's starting point, tributary confluence points, and river mouths. Furthermore, if existing river network data exists, it can be calibrated with the existing river network data.

[0042] Based on the topographic features and actual needs of the study area, a suitable threshold for runoff accumulation is set. Grids with runoff accumulation greater than or equal to this threshold are considered potential water flow paths. The network formed by these grids is the raster river network. In ArcGIS, map algebra tools can be used to extract grids that meet the threshold conditions through the "greater than or equal to" operation, generating raster river network data. The extracted raster river network is then converted into vector format for subsequent editing, analysis, and mapping. In ArcGIS, the raster to vector tool in ArcToolbox can be used to convert the raster river network data into vector river network data, ultimately obtaining a river network distribution map.

[0043] S13: Extract sub-basin boundaries;

[0044] A catchment outlet is set at each node of the river network, such as at the node where a tributary flows into the main stream. The runoff of each sub-basin eventually flows into the river network from this point. Based on the water flow direction data, all the grids upstream of each catchment outlet that flow towards that outlet are traced in reverse. The set of these grids is the sub-basin corresponding to that catchment outlet. The rasterized sub-basin boundary is converted into a vector polygon and the boundary is corrected in combination with the actual terrain, such as mountains and artificial embankments, to ensure that it is consistent with the watershed. The above S11-S13 can be implemented using ArcGIS, specifically by using the Hydrology Tools toolkit to sequentially execute "fill depressions, water flow direction, runoff accumulation, river network extraction, confirmation of river network nodes, and sub-basin division". Alternatively, it can be implemented using QGIS (Quantum Geographic Information System), specifically by using the GRASS plugin (Geographic Resource Analysis Support System plugin) or Whitebox Tools in the Processing toolbox to achieve similar terrain analysis and sub-basin extraction.

[0045] S14: Based on the above S11-S13, obtain the continuous monitoring hydrological data and water quality data, meteorological data, soil condition data, land use type data, and various pollution source data in each sub-basin.

[0046] The aforementioned continuously monitored hydrological and water quality data were obtained from local environmental protection departments or websites within the study sub-basin area; meteorological data were preferably daily meteorological data, obtained through the China Meteorological Science Data Sharing Service Network; soil condition data could be obtained through the World Harmonious Soil Database, the Soil Branch Center of the National Earth System Science Data Center, websites of the Ministry of Natural Resources, and websites of the Ministry of Ecology and Environment; land use type data were mainly obtained from remote sensing interpretation or environmental protection departments; and data on various pollution sources were mainly obtained from statistical yearbooks or local environmental protection departments.

[0047] The calculation formula for the spatial-based watershed attribute regression model is as follows:

[0048] in, This represents the pollution flux flowing through sub-basin i, expressed in kg / year. This indicates the upstream river segment adjacent to river segment in, where river segment in is one of the river segments divided from sub-basin i; Represents the index of the upstream river segment, traversing The upstream section of the river in the collection, This indicates the pollution flux in the upstream section of the river; This indicates the proportion of upstream flux transported to the river segment in. This represents the attenuation function of pollutant transport along the river channel. Different equations exist for rivers and lakes; if the water body is a river, then the corresponding equation is... and If the water body is a reservoir or lake, then the corresponding and ,in, The parameter matrix representing the attenuation of river transport. This represents a parameter matrix related to reservoir / lake transport attenuation; This represents the coefficients to be fitted for river transport attenuation. NS represents the coefficients to be fitted for the transport attenuation of a reservoir or lake; NS represents the total number of pollution sources in sub-basin i, and n is the index of the pollution source. These represent the amount generated by each pollution source, which can represent both the actual amount of pollution emitted by the source and the area of ​​different types of land. This represents the emission coefficient of the pollution source. Indicates the soil-water transport factor. Represents the parameter matrix, This represents the coefficient corresponding to parameter D (the transmission coefficient for pollution sources that directly discharge into the river, such as point sources, is 1). This represents the attenuation function of locally generated pollutants in this section of the river. It has different equation forms for rivers and lakes. If the water body is a river, then it corresponds to... and If the water body is a reservoir or lake, then the corresponding and , where θ represents the coefficient to be fitted; although functions A and A' are similar in form, they consider different river retention times. The former assumes that pollutants are attenuated by the entire river segment, while the latter assumes that pollutants enter the river segment i halfway and are only attenuated by halfway.

[0049] Using continuous monitoring of hydrological and water quality data, meteorological data, soil conditions, land use type data, and various pollution source data within each sub-basin, a system of simultaneous equations was established. The nonlinear least squares method was applied to fit the parameters, and a spatial watershed attribute regression model was established. The relevant model parameters were calibrated, and the impact of different pollution sources and land use types on water ecological load within the watershed was quantitatively analyzed through the established spatial watershed attribute regression model.

[0050] The impact of different pollution sources and land use types on water ecological load varies significantly. The following analysis is conducted from two dimensions: pollution source type and land use type. Water ecological load includes, but is not limited to, the pollution pressure and structural damage load borne by the water ecosystem.

[0051] The impact of different pollution sources on the water ecological load: Pollution sources directly increase the load on the water ecosystem by inputting pollutants into water bodies or changing hydrological processes. The main impacts are as follows: Pollutants enter water bodies through dispersed pathways such as surface runoff, farmland drainage, and atmospheric deposition. Pollutants mainly include agricultural fertilizers and pesticides, urban surface runoff, and rural domestic sewage. Their treatment is difficult and requires comprehensive measures such as land use adjustment and the construction of vegetation buffer zones, rather than being solved by a single project.

[0052] The impact of different land use types on water ecological load: Land use types indirectly affect water ecological load by altering land cover, hydrological processes, and the distribution of pollution sources, as detailed below:

[0053] Natural ecological land (forest land, grassland, wetlands, etc.) can reduce soil erosion through vegetation interception and root stabilization, and reduce non-point source pollution load by absorbing pollutants (such as nitrogen and phosphorus); wetlands can purify water bodies through sedimentation and microbial decomposition, buffer flood peaks, and maintain the diversity of aquatic ecosystems. However, the over-development of natural ecological land, such as deforestation for cultivation, will result in the loss of ecological functions and exacerbate the pollution load.

[0054] Agricultural land (cultivated land, orchards, etc.) is a major source of non-point source pollution due to the excessive use of chemical fertilizers and pesticides, which leads to nitrogen and phosphorus loss. Exposed cultivated land is prone to soil erosion, carrying sediment and pollutants into water bodies, exacerbating river siltation and water turbidity. Measures such as returning farmland to forest and constructing ecological ditches can reduce its pollution load.

[0055] Urban construction land (cities, towns, industrial land, etc.) has a high proportion of impermeable surfaces and a large surface runoff coefficient, making it easy to carry urban waste, oil pollution, heavy metals, etc. into water bodies, causing initial rainwater pollution; it also encroaches on natural water bodies, such as filling in lakes to create land, which will damage the integrity of aquatic ecosystems and reduce the self-purification capacity of water bodies.

[0056] Unused land (bare land, wasteland, etc.) may have different characteristics depending on whether it is native bare land, such as desert or Gobi, which are ecologically fragile but contribute little to pollution; or degraded land, such as mining wasteland, which may release heavy metals or saline substances due to soil erosion, increasing the pollution load on water bodies.

[0057] Step Two: Based on the impact of each land use type on water ecological load in Step One, and combined with the rigid boundaries in the territorial spatial planning of the study area, land use schemes and ecological restoration schemes are established. Then, an optimized spatial watershed attribute regression model is obtained by coupling a spatial watershed attribute regression model and an annealing algorithm. This optimized model is used to assess the land use regulation potential and ecological restoration regulation potential of the corresponding watersheds. The rigid boundaries in the territorial spatial planning of the study area are the three zones and three lines of the study area. The three zones are urban space, agricultural space, and ecological space; the three lines are the ecological protection red line, the permanent basic farmland protection red line, and the urban open boundary.

[0058] In this step, the land use regulation potential and ecological restoration regulation potential of the corresponding sub-basins are assessed through an optimized spatial watershed attribute regression model. This ensures that limited funds can be allocated more accurately in the future, prioritizing the treatment of areas with "low investment and quick results" to achieve efficient improvement of water quality at the watershed outlet section. Because different land use patterns, such as farmland, towns, and forests, have significantly different effects on pollution reduction, for example, farmland with reduced fertilizer use can reduce nitrogen and phosphorus pollution more quickly than scattered villages, thus directly affecting the water quality at the watershed outlet section. Moreover, the treatment funds are limited during the treatment process, making it impossible to apply equal effort to all areas. This embodiment uses an optimized spatial watershed attribute regression model to assess the two types of potential in each area within the sub-basin, prioritizing the investment in areas with high regulation potential to achieve "the best water quality improvement effect with the least amount of money".

[0059] Land use regulation potential refers to the maximum possible reduction in pollution or improvement in water quality that can be achieved by adjusting land use patterns (such as changing land use type, optimizing layout, etc.). For example, if a region is a high-density farmland with heavy use of chemical fertilizers and pesticides and serious non-point source pollution, converting part of it into ecological buffer zones, such as wetlands and grasslands, can significantly intercept pollutants. The amount of emission reduction brought about by this conversion from farmland to buffer zones is the land use regulation potential of that region. The key is whether there is room for optimization of land use patterns in a region, and how much pollutants can be reduced after the adjustment.

[0060] Ecological restoration and regulation potential refers to the maximum possible reduction in pollution or improvement in water quality that can be achieved through ecological restoration measures (such as restoring damaged ecosystems and enhancing natural purification capacity). For example, if a region is a degraded riverbank with sparse vegetation that cannot intercept pollutants from surface runoff, planting aquatic plants and restoring riparian forests can enhance the adsorption and degradation capacity of pollutants. The emission reduction brought about by this enhancement of ecosystem function is the ecological restoration and regulation potential of that region. Similarly, if a region is a dried-up wetland that has lost its original purification function, replenishing water and restoring the wetland hydrology can allow it to once again function as a natural wastewater treatment plant. This purification space restored by the ecosystem also falls into this category of potential. The key is the extent of damage to the ecosystem of a specific region, such as wetlands, woodlands, or rivers, and how much purification capacity can be enhanced after restoration, such as how much pollutants can be intercepted and how much water quality indicators can be improved.

[0061] By using an optimized spatial watershed attribute regression model, we can assess the land use regulation potential and ecological restoration regulation potential of the corresponding sub-watersheds, identify the areas with the "highest cost-effectiveness" for governance, and concentrate limited funds on places where the water quality at the outlet section can be improved the fastest and to the greatest extent, thus avoiding waste of resources. The essence of these two regulation potentials is to assess how much water quality improvement space can be brought about by human intervention in nature and land, which is a key basis for scientific decision-making.

[0062] More specifically: The assessment of land use regulation potential is mainly based on the planning of the three zones and three lines in the national land planning, to formulate corresponding land use regulation potential, which is then input into a spatial watershed attribute regression model constructed based on the current land use status, and compared with the simulation results of the model calculated using the original land use type (utilization rate), thereby identifying controllable space and obtaining the land use regulation potential; The assessment of ecological restoration regulation potential mainly uses an optimized spatial watershed attribute regression model, and the regulation threshold of the relevant regulation potential is substituted into the optimized spatial watershed attribute regression model to find the relevant ecological restoration regulation potential.

[0063] Among them, the annealing algorithm is a stochastic optimization algorithm based on the Monte Carlo iterative solution strategy. Its starting point is based on the similarity between the annealing process of solid materials in physics and general combinatorial optimization problems. The simulated annealing algorithm starts from a relatively high initial temperature. As the temperature parameter decreases, it combines the probabilistic jump characteristic to randomly search for the global optimal solution of the objective function in the solution space. That is, the local optimal solution can be jumped out of the local optimal solution and eventually tend to the global optimal solution.

[0064] The above optimized spatial watershed attribute regression model is obtained by setting the constraints of the initially established spatial watershed attribute regression model as the objective function. The constraints are as follows:

[0065] ; ;

[0066] Where N represents the number of sub-basins, i is the index of the sub-basin, and i is used to traverse the sub-basins from the 1st sub-basin to the Nth sub-basin, representing the summation and constraint of the pollution source generation in each sub-basin; This represents the amount of pollution generated by each pollution source (such as the area of ​​cultivated land or the load generated by cultivated land), Sopt represents the control target value of the pollution source, and Si and Sj are the upper and lower boundaries of the pollution source, respectively.

[0067] Step 3: Based on the land use control scenario and ecological restoration control scenario set in Step 2, use the optimized spatial watershed attribute regression model to simulate and evaluate the relevant environmental effects, and formulate optimized restoration measures.

[0068] Step two has already assessed the land use regulation potential and ecological restoration regulation potential of different areas within the sub-basin, i.e., it is known which areas are more effective in adjusting land use and which areas are more effective in restoring the ecology. Based on the land use regulation potential and ecological restoration regulation potential, corresponding land use regulation scenarios and ecological restoration regulation scenarios are set. The optimized spatial watershed attribute regression model is used to simulate the environmental effects that the above regulation scenarios can bring, such as how much pollutants are reduced and how much water quality is improved. Finally, based on the simulation results, the "best effect and lowest cost" scheme is selected to form an optimized restoration measure scheme.

[0069] Setting up land use regulation scenarios and ecological restoration regulation scenarios specifically involves designing specific and operable governance plans based on the regulation potential identified in step two. More specifically:

[0070] Land use regulation scenario: Specific plans designed to adjust land use patterns, such as converting 10% of the highly polluted farmland (with high fertilizer usage) in the watershed into ecological wetlands; returning farmland to forest can reduce fertilizer usage by reducing the area of ​​cultivated land, thereby having a positive impact on water quality.

[0071] Ecological restoration and regulation scenarios: Specific plans designed for the restoration of ecosystems, such as: restoring riparian vegetation (planting reeds, willows, etc.) within 100 meters of both banks of the river, covering 50% of the river section; replenishing water and rebuilding vegetation in three degraded wetlands within the basin to restore their purification functions.

[0072] The optimized spatial watershed attribute regression model is used to simulate relevant environmental effects. This involves quantifying the environmental impact under different scenarios using the model, determining which scenario is more effective, and combining the optimized spatial watershed attribute regression model with geospatial characteristics (such as soil, topography, and pollution sources in different regions) to calculate, after each scenario is implemented: how much pollutant emissions can be reduced (such as specific indicators like nitrogen, phosphorus, and COD); how much the water quality at the watershed outlet section can be improved (such as from Class IV to Class III); the difference in contribution between different regions (such as how much higher the emission reduction in region A is compared to region B under the same scenario); and using data to illustrate which scenario has a higher cost-effectiveness, such as scenario A investing 1 million to reduce nitrogen emissions by 20 tons, while scenario B investing 1.2 million to reduce nitrogen emissions by 25 tons, requiring comparison to select the appropriate solution.

[0073] Develop optimized remediation measures and, based on the model simulation results, select the scenario combination with the best environmental effect, lowest cost, and highest feasibility to form the final remediation plan. For example, if the simulation finds that the combination of converting 20% ​​of highly polluted farmland into wetlands (land use scenario) and restoring 30% of the riparian zone (ecological restoration scenario) can improve the water quality at the outlet section by one level at the lowest cost, then this combination will be identified as the optimized remediation measures to ensure that the limited investment of funds can maximize the improvement of water quality.

[0074] Specifically:

[0075] S31: The existing land type reserves and existing land use within the existing rigid boundaries shall be used as the control boundaries for each sub-basin, and the area for which restoration measures are set shall reach the short-term or long-term goals; specifically, the existing rigid boundaries are the rigid boundaries specified in the existing three zones and three lines.

[0076] When formulating watershed pollution control or ecological restoration plans, the land type restrictions defined by the three zones and three lines and the current land use status must be used as hard boundaries. At the same time, different phased implementation scenarios should be designed according to the restoration goals and actual feasibility. Restoration measures cannot be accomplished overnight and need to achieve short-term and long-term goals in stages. Therefore, specific plans for different stages should be designed.

[0077] Land type and quantity: Based on the three zones and three lines, the minimum / maximum area that a certain type of land must maintain is stipulated. For example, the permanent basic farmland in a certain watershed shall not be less than 500,000 mu, and the proportion of ecological land shall not be less than 30%.

[0078] Current land use status: refers to the actual distribution of various types of land within the current watershed, such as 600,000 mu of existing cultivated land, 200,000 mu of forest land, and 100,000 mu of urban construction land, etc.

[0079] Regulation boundaries: Restoration measures cannot exceed these boundaries, which are limited by the amount of land types and existing land use. For example, if the permanent basic farmland control line requires no less than 500,000 mu of cultivated land and there are 600,000 mu of existing cultivated land, then the maximum area of ​​returning farmland to forest cannot exceed 100,000 mu, which is the regulation boundary. If a certain area belongs to the ecological protection red line, then urban construction cannot be carried out in the area, and only ecological restoration can be carried out, such as planting trees and replenishing water.

[0080] Setting restoration measures to achieve near-term or long-term goals: This refers to dividing the restoration task into phases over time, specifying the area to be completed in each phase. For example, the long-term goal for a certain watershed is a total of 100,000 mu of land to be converted from farmland to forest. However, considering factors such as farmers' livelihood transition and the phased allocation of funds, the near-term goal is to complete 30,000 mu within 3 years, the medium-term goal is to complete another 50,000 mu within 5 years, and the long-term goal is to complete the remaining 20,000 mu within 10 years. The goals for each phase must meet the regulatory boundaries. For example, after converting 30,000 mu of farmland in the near term, the remaining farmland still needs to meet the requirements of permanent basic farmland, which is no less than 500,000 mu.

[0081] The conversion of farmland to forest needs to be carried out gradually, so different scenarios are set up here, as shown in the following examples:

[0082] Scenario for near-term goals: From 2025 to 2028, 30,000 mu of easily eroded farmland with a slope of more than 25 degrees will be converted back to forest and planted with trees, with supporting small-scale water conservancy facilities.

[0083] The medium-term target scenario: From 2029 to 2033, another 50,000 mu of farmland with a slope of 15 to 25 degrees will be converted from farmland to mixed forests of trees and shrubs, and understory economy will be developed, such as planting medicinal herbs, to ensure farmers' income.

[0084] Long-term goal scenario: By 2034-2040, complete the conversion of the remaining 20,000 mu of farmland back to forest, while optimizing the vegetation in the areas converted from farmland in the first two phases to improve the stability of the ecosystem.

[0085] This step ensures that the solution is legal and compliant, and that the remediation process is smooth and effective, avoiding problems caused by a one-size-fits-all approach or rash governance.

[0086] S32: Using the established spatial watershed attribute regression model as the objective function, with the goal of minimizing the inflow load of each sub-watershed into the lake, the spatial priority restoration area for restoration measures is calculated using the optimized spatial watershed attribute regression model.

[0087] S33: Based on the spatial priority repair area, derive the spatial optimization repair scheme for each repair measure. Example 2

[0088] This embodiment provides a specific application scenario, using a lake basin in the Yangtze River basin as a case study:

[0089] Step 1: Based on the basic watershed data within the study area, construct a spatial watershed attribute regression model. In this embodiment, download a 30m resolution DEM of the watershed area of ​​a certain lake with five rivers and seven outlets from the Chinese Academy of Sciences scientific database. Based on the DEM data, generate the river network and sub-watershed partitions within the study area. The data within the sub-watersheds are obtained through the following methods: meteorological data obtained from the China Meteorological Science Data Sharing Service Network; continuous monitoring hydrological data, water quality data, land use type data, and various pollution source data obtained from the environmental protection department of the province corresponding to the lake; and soil condition data obtained from the website of the ecological and environmental protection department of the province corresponding to the lake.

[0090] Using the data unified to the sub-basin, a system of simultaneous equations is established, and the nonlinear least squares method is applied for parameter fitting. Taking into account the relationship between the water body and the land area of ​​the basin, a spatial-based watershed attribute regression model is constructed.

[0091] Based on water quality monitoring data, land use data, and pollution source statistics for total phosphorus in the Wuhe Qikou watershed in 2017, relevant model parameters were calibrated; among them, the determination coefficient R of the established spatial watershed attribute regression model was... 2 =0.8639, Nash efficiency coefficient is 0.8558, root mean square error is 0.22, indicating that the spatial watershed attribute regression model has good simulation effect.

[0092] After completing the modeling work of the spatial watershed attribute regression model, the impact of different pollution sources and land use types on water ecological load within the watershed was quantitatively analyzed using the spatial watershed attribute regression model. The results showed that in 2017, the pollution source with the highest proportion of total phosphorus load entering the river in a certain lake watershed was cultivated land, accounting for 25.59%; followed by urban domestic sources, accounting for 23.21%, which were the main sources of total phosphorus pollution in the lake watershed in 2017 and should become the main direction of pollution source control.

[0093] Step 2: Based on the rigid boundaries of the three zones and three lines in the territorial spatial planning of the study area, establish the regulation potential (land use scheme and ecological restoration scheme), and then use the spatial watershed attribute regression model and annealing algorithm to optimize the spatial watershed attribute regression model. Use the optimized spatial watershed attribute regression model to evaluate the land use regulation potential and ecological restoration regulation potential of the corresponding watershed.

[0094] According to the "Three Zones and Three Lines" plan in the territorial spatial planning of a certain province, the area of ​​the urban development boundary is 5,101 square kilometers. In the future, urban pollution will be further aggravated. Therefore, it is difficult to optimize and regulate urban land use. This embodiment mainly focuses on assessing the potential for land use regulation of arable land and the potential for ecological restoration regulation of arable land.

[0095] The plan stipulates that the area of ​​cultivated land within the agricultural space should not be less than 29,273 square kilometers. Based on the proportion of the basin with five rivers and seven outlets in a certain lake to the whole basin, and combined with the county and district ledgers in the "Three Zones and Three Lines" plan, relevant conversion calculations are performed to obtain the basic farmland area of ​​each sub-basin. The above data is then used to simulate the basin attribute regression model based on space to obtain the reduction in the load transmitted to the downstream outlet due to cultivated land regulation, which is the land use regulation potential.

[0096] The difference between the area for land use regulation and the minimum area for arable land preservation is added to the ecological space area, which is the maximum area of ​​ecological space. The minimum area for arable land preservation and the maximum area of ​​ecological space are then incorporated into the established regulation potential assessment model. Based on the calibration of relevant parameters, the reduction in total phosphorus pollution after returning farmland to forest is simulated, which serves as the regulation potential of returning farmland to forest, i.e., the ecological restoration regulation potential assessment.

[0097] Step 3: Based on the land use control scenario and ecological restoration control scenario set in Step 2, use the optimized spatial watershed attribute regression model to simulate and evaluate the relevant environmental effects, and formulate optimized restoration measures.

[0098] Land use optimization and regulation are mainly based on the aforementioned spatial watershed attribute regression model, combined with linear programming methods for evaluation. First, the conversion of farmland to forest uses the amount of cultivated land and the existing cultivated land status in the "three zones and three lines" as the control boundary for each sub-watershed, setting the area for conversion to forest to achieve the short-term goal (or long-term goal). Then, the established spatial watershed attribute regression model is used as the objective function, with the goal of minimizing the inflow load into the lake in each sub-watershed. The optimized spatial watershed attribute regression model is used to calculate the spatial priority control area for conversion to forest, thus deriving the spatial optimization control scheme for conversion to forest.

[0099] In summary, due to the adoption of the above technical solution, this application has the following advantages:

[0100] I. Constructing a spatially based watershed attribute regression model requires relatively little data, has strong interpretability, and is suitable for use in most regions.

[0101] Second, the spatial watershed attribute regression model can simultaneously simulate multiple watershed outlets, comprehensively consider the relationship between water bodies and land areas in the watershed, quantify the degree of pollution of pollutants, and intuitively reflect the impact of land use on pollution load.

[0102] Third, based on the spatial watershed attribute regression model, this application uses the mature annealing algorithm to optimize the spatial watershed attribute regression model, which can evaluate different remediation measures and formulate practical and effective spatial remediation schemes for the water environment of control units, thereby guiding local areas to carry out targeted measures and countermeasures.

[0103] The preferred embodiments of this application have been described in detail above with reference to the accompanying drawings. However, this application is not limited to the specific details of the above embodiments. Within the scope of the technical concept of this application, various simple modifications can be made to the technical solution of this application, and these simple modifications all fall within the protection scope of this application.

[0104] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable way without contradiction. In order to avoid unnecessary repetition, the various possible combinations in this application will not be described separately.

[0105] Furthermore, various different implementations of this application can be combined in any way, as long as they do not violate the spirit of this application, and such combinations should also be regarded as the content disclosed in this application.

Claims

1. A method for optimizing and regulating land use in a multi-outflow watershed, characterized in that, Includes the following steps: Step 1: Obtain basic watershed data for the study area, construct a spatial watershed attribute regression model, and quantitatively analyze the impact of various land use types on water ecological load within the watershed of the study area through the spatial watershed attribute regression model. Step 2: Based on the impact of each land use type on water ecological load in Step 1, establish land use plans and ecological restoration plans; An optimized spatial watershed attribute regression model is obtained by coupling a spatial watershed attribute regression model with an annealing algorithm. This optimized model is then used to assess the land use regulation potential and ecological restoration regulation potential of the corresponding watershed. The optimized model is derived by setting constraints based on the initially established spatial watershed attribute regression model as the objective function. The constraints are as follows: ； ； Where N represents the number of sub-basins, and i is the index of the sub-basin. This represents the amount of pollution generated by each pollution source, Sopt represents the target value for the pollution source, and Si and Sj represent the upper and lower boundaries of the pollution source, respectively. Step 3: Based on the land use regulation scenario and ecological restoration regulation scenario set in Step 2, use the optimized spatial watershed attribute regression model to simulate the environmental effects related to the land use regulation scenario and ecological restoration regulation scenario, and formulate optimized restoration measures.

2. The method for optimizing and regulating land use in a multi-outlet watershed according to claim 1, characterized in that, The step one, which involves obtaining basic watershed data for the study area, includes: S11: Determine the direction of water flow and the cumulative flow; The flow direction of water in each grid is calculated based on the DEM, the number of all incoming water paths upstream of each grid is counted, the cumulative flow of the grid is obtained, and then the threshold of the cumulative flow is determined. S12: Refinement and delineation of the river network; Based on the cumulative runoff threshold, the river network is extracted from the cumulative runoff data. The nodes of the river network are determined based on the extracted river network. The nodes include the starting point of the river, the confluence point of the tributaries, and the river mouth. S13: Extract sub-basin boundaries; At each node of the river network, a water catchment outlet is set up. Based on the water flow direction data, all grids upstream of each water catchment outlet that flow towards that outlet are traced in reverse. The set of these grids is the sub-basin corresponding to that water catchment outlet. S14: Obtain continuous monitoring hydrological and water quality data, meteorological data, soil condition data, land use data, and statistical data of various pollution sources within each sub-basin.

3. The method for optimizing and regulating land use in a multi-outlet watershed according to claim 2, characterized in that, In step S12, the river network is extracted from the confluence accumulation data and then calibrated with the existing river network data.

4. The method for optimizing and regulating land use in a multi-outlet watershed according to claim 1, characterized in that, The calculation formula for the spatial watershed attribute regression model in step one is as follows: ;in, This represents the pollution flux flowing through sub-basin i, expressed in kg / year. This indicates the upstream river segment adjacent to river segment in, where river segment in is one of the river segments divided from sub-basin i; Indicates the index of the upstream river section, This indicates the pollution flux in the upstream section of the river; This indicates the proportion of upstream flux transported to the river segment in. This represents the attenuation function of pollutant transport along the river channel. Different equations exist for rivers and lakes; if the water body is a river, then the corresponding equation is... and If the water body is a reservoir or lake, then the corresponding and ,in, The parameter matrix representing the attenuation of river transport. This represents a parameter matrix related to transport attenuation in a reservoir or lake. This represents the coefficients to be fitted for river transport attenuation. NS represents the coefficients to be fitted for the transport attenuation of a reservoir or lake; NS represents the total number of pollution sources in sub-basin i, and n is the index of the pollution source. These represent the amount produced by each pollution source. This represents the emission coefficient of the pollution source. Indicates the soil-water transport factor. Represents the parameter matrix, This represents the coefficient corresponding to parameter D. This represents the attenuation function of locally generated pollutants in this section of the river. It has different equation forms for rivers and lakes. If the water body is a river, then it corresponds to... and If the water body is a reservoir or lake, then the corresponding and .

5. The method for optimizing and regulating land use in a multi-outlet watershed according to claim 1, characterized in that, Step three includes: S31: The existing land types and land use status are used as the control boundaries for each sub-basin, and the area for control measures is set to achieve the short-term or long-term goals. S32: Using the established spatial watershed attribute regression model as the objective function, with the goal of minimizing the inflow load of each sub-watershed into the lake, the spatial priority restoration area for restoration measures is calculated using the optimized spatial watershed attribute regression model. S33: Based on the spatial priority repair area, derive the spatial optimization repair scheme for each repair measure.

Images