River-lake water quantity linkage regulation method based on ecological safety

Abstract

The application discloses a river-lake water quantity linkage regulation and control method based on ecological safety, and relates to the technical field of water resources management. The method comprises the following steps: extracting historical multi-source remote sensing image data of a target river-lake basin, constructing a water body extraction model by using a random forest learning algorithm in combination with original spectral bands and various water body indexes, and generating a monthly binary water body graph; screening an influence area of an inflow river on a lake; constructing a nonlinear response model of inflow water quantity-lake area based on a monthly area data sequence; constructing a MIKE hydrodynamic model of the inflow river; setting inflow water quantity of the inflow river to the lake, inputting the MIKE hydrodynamic model, simulating ecological discharge quantity of the inflow river, and constructing an ecological discharge quantity model of the inflow river; regulating ecological discharge water quantity of the inflow river under different ecological targets; formulating a hierarchical regulation scheme of rich, flat and dry year types, and realizing collaborative optimization of ecological benefits and water resource utilization efficiency.

Description

Technical Field

[0001] This invention relates to the field of river-lake ecological water volume regulation, specifically to a method for coordinated regulation of river-lake water volume based on ecological security. Background Technology

[0002] Inland river terminal lakes in arid regions are key components of the "mountain-oasis-desert" system, extremely sensitive to climate change and human activities, and often regarded as "indicators" of watershed ecological security. In recent decades, due to the combined effects of large-scale water and soil development and climate change, many terminal lakes in the arid northwest have continued to shrink or even dry up, leading to shoreline salinization, frequent salt dust storms, and a sharp deterioration in ecological security. While some lakes (such as Lake Taitma) have been restored through ecological water transfer projects, the current scheduling methods have significant technical shortcomings. There is a lack of quantitative response models for lake surface area and inflow volume: existing scheduling relies heavily on experience or simple linear regression, without considering nonlinear characteristics, and cannot accurately answer the question of "how much water is needed for a certain area".

[0003] The existing methods fail to differentiate the spatial contributions of multiple water sources: Lake Taitma is jointly supplied by the Tarim River and the Cherchen River. The two rivers have significantly different entry paths and efficiencies into the lake, but the existing methods do not differentiate regulation based on spatial control zones, resulting in water waste.

[0004] The lack of detailed simulation of water dissipation along the river: Evaporation and infiltration losses account for a large proportion of the water transport process in arid areas. Traditional hydrological models cannot dynamically depict the nonlinear transformation relationship of evaporation, infiltration and lake entry under different discharge scales, which affects scheduling decisions.

[0005] There is a lack of adaptive regulation strategies for different hydrological year types: existing schemes are too simplistic and cannot dynamically adjust targets according to wet, normal, and dry years, making it difficult to balance ecological water demand and efficient water resource utilization. Summary of the Invention

[0006] In view of the above-mentioned shortcomings of the prior art, the present invention provides a river-lake water volume linkage regulation method based on ecological security, so as to achieve precise protection of the ecological security of tailrace lakes.

[0007] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows: A method for coordinated regulation of river and lake water volume based on ecological security is provided, which includes the following steps: S1: Extract historical multi-source remote sensing image data of the target river-lake basin, use the random forest learning algorithm to combine the original spectral bands with multiple water body indices to construct a water body extraction model, extract the monthly lake surface area sequence of the target river-lake basin, and generate a monthly binary water body map; S2: Calculate the historical inundation frequency of each pixel based on the monthly binary water body map of each historical month, and analyze the changing trend of the lake surface water area based on the changing trend of the monthly area data sequence to screen the areas affected by the rivers flowing into the lake. S3: Construct a nonlinear response model of inflow volume-lake surface area based on monthly area data series, and output the total inflow volume required each month according to the target lake surface area required to be maintained for different ecological needs of the lake each month. S4: Construct a MIKE hydrodynamic model of the river flowing into the lake to simulate the evolution of water flow from different cross sections of the target river into the lake to the lake inlet; S5: Set the water volume of the rivers flowing into the lake, input the MIKE hydrodynamic model, simulate the ecological discharge of the rivers flowing into the lake, calculate the actual monthly water volume flowing into the lake during the simulation process, calculate the inflow efficiency of the monthly ecological water volume of the rivers flowing into the lake, and construct the ecological discharge model of the rivers flowing into the lake. S6: Input the total monthly inflow volume required for different ecological objectives into the ecological discharge model, calculate the theoretical ecological discharge volume of the rivers flowing into the lake, and construct a weighted multi-objective function with the theoretical ecological discharge volume as the decision variable based on the influence area of ​​the rivers flowing into the lake, so as to regulate the ecological discharge volume of the rivers flowing into the lake under different ecological objectives.

[0008] Further, step S1 includes: S11: Extract multi-source remote sensing image data of the target river-lake basin and preprocess it to obtain the surface reflectance dataset of the target river-lake basin; S12: Construct a feature space that integrates multispectral indices and original bands, and input it into a random forest classifier to extract water body pixels; Multispectral indices include the Normalized Difference Water Index (MNDWI), the Automatic Water Extraction Index (AWEIsh), the Multiband Water Index (MBWI), and the Land Water Index (LSWI). The Normalized Differential Water Index (MNDWI), Automatic Water Extraction Index (AWEIsh), Multiband Water Index (MBWI), and Land Water Index (LSWI) are stacked on green, blue, red, shortwave infrared band 1, and shortwave infrared band 2 to obtain a 10-dimensional feature vector. ; ; in, Blue band reflectivity; Obtain the feature vector of each pixel in the remote sensing image. , i For the cell number; Construct a random forest classifier by inputting the feature vector of each pixel into... K In each decision tree, the output is the prediction result for the pixel by each decision tree. 1 indicates that the pixel is a water body, and 0 indicates that the pixel is not a water body; Based on the prediction results of each decision tree for the pixel Calculate the probability that a pixel belongs to a water body. ; ; in, k Number the decision tree; S13: Based on the area of ​​each cell A i Calculate the monthly surface water area of ​​the target river-lake basin. The monthly area data sequence of the lake surface water body is obtained, and a monthly binary water body map is generated. In the monthly binary water body map, each pixel is marked as 1 or 0 to indicate whether it belongs to the water body. ; in, This represents the probability of belonging to a body of water. pixels, N The probability that it belongs to a body of water. The number of pixels, t For months.

[0009] Furthermore, the multispectral indices include the Normalized Difference Water Index (MNDWI), the Automatic Water Extraction Index (AWEIsh), the Multiband Water Index (MBWI), and the Land Water Index (LSWI). ; in, For green band reflectivity, The reflectivity is 1 for shortwave infrared. ; in, Near-infrared reflectance, For shortwave infrared reflectivity; ; in, Reflectivity in the red band; .

[0010] Further, step S2 includes: S21: Based on whether the pixels in the monthly binary water body maps of each historical month are binary values ​​of water bodies, determine their binarized values. Calculate the flooding frequency for each pixel's history. ; ; in, T Total number of months in history The coordinates of the pixel; like Then the corresponding pixel will be divided into a permanent water body area, if If the corresponding pixel is in the seasonal water body area, then the corresponding pixel will be classified as a non-submerged area; otherwise, the corresponding pixel will be classified as a non-submerged area. S22: Calculate the trend of lake surface water area change for any two months based on the monthly data series of lake surface water area. S ; ; in, Any two months The surface area of ​​the lake, It is a symbolic function; S23: Utilizing the trend of lake surface water area changes S Calculate the variance of the trend of lake surface water area change in monthly area data series. ; ; in, p This refers to the classification of months with the same lake surface area in the monthly area data series. The number of months in the categorized month group. g The number of categories for each month group; S24: Variance based on the trend of lake surface water area change Calculate the standardized test statistic Z ; ; S25: If If the monthly area data series shows a significant upward or downward trend in lake surface area, then the lake surface area shows a significant upward or downward trend; otherwise, there is no significant upward or downward trend. S26: Based on the inundation frequency of pixels within permanent and seasonal water bodies. Screening the pixel inundation frequency of rivers flowing into the lake during independent replenishment periods. Screening the areas of influence of rivers flowing into the lake on the lake .

[0011] Further, step S3 includes: S31: Calculate the lake surface area for each month in the monthly area data series. Total inflow of water into the lake by each month Nonlinear regression was performed to establish a nonlinear response model of inflow volume versus lake surface area. ; in, These are the nonlinear fitting coefficients; S32: Input the target lake surface area required to be maintained for different ecological needs each month into the nonlinear response model of lake inflow-lake surface area, and calculate the total inflow required for each month under different ecological objectives.

[0012] Further, step S5 includes: S51: Based on the annual runoff characteristics and ecological regulation needs of the rivers flowing into the lake, the water discharge volume of the rivers into the lake is set, and different monthly discharge volumes are set as discharge step sizes. The ecological discharge volume of the rivers flowing into the lake is simulated, and the monthly discharge volume is converted into a daily flow process and input into the MIKE hydrodynamic model to calculate the daily inflow volume of the rivers flowing into the lake. The daily inflow volume is then integrated to obtain the actual monthly inflow volume. ; S52: Calculate the total monthly evaporation of rivers flowing into the lake. V evap and total monthly infiltration V inf Calculate the theoretical ecological discharge volume of the rivers flowing into the lake. V release ; ; in, This represents the monthly change in the water storage of rivers flowing into the lake. S53: Based on theoretical ecological discharge volume V release and the actual monthly inflow of water into the lake Calculate the inflow efficiency of monthly ecological water volume of rivers flowing into the lake ; S54: Obtain the actual monthly inflow to the lake under each simulated discharge step condition. Theoretical ecological discharge volume V release A piecewise function was used to fit the relationship between the actual monthly inflow into the lake and the theoretical ecological outflow, thus obtaining the ecological outflow model of the rivers flowing into the lake. ; in, The slope The intercept is... This represents the critical threshold for segmentation.

[0013] Further, step S6 includes: S61: Input the total monthly inflow volume required for different ecological objectives into the ecological discharge model of the inflowing rivers to calculate the theoretical ecological discharge volume of the inflowing rivers. ; S62: Based on the impact area of ​​rivers flowing into the lake on the lake Calculate the area of ​​impact of rivers flowing into the lake on the lake under different ecological objectives. ,W For the affected area The number of internal pixels; S63: Construct a weighted multi-objective function with the theoretical ecological discharge volume as the decision variable. J With weighted multi-objective function J With the goal of minimizing the value, the theoretical ecological discharge volume of rivers flowing into the lake under different ecological objectives is optimized, and the ecological discharge volume of rivers flowing into the lake under different ecological objectives is regulated. ; in, The ideal impact area of ​​rivers flowing into the lake under different ecological objectives. This represents the baseline natural flow of rivers flowing into the lake for the corresponding months under different ecological objectives. The number of months maintained for different ecological objectives. These are the weighting coefficients for the area affected by rivers flowing into the lake and the flow smoothing coefficients, respectively.

[0014] The beneficial effects of this invention are as follows: 1. Establish a nonlinear frequency response relationship between lake surface area and inflow volume to achieve quantitative decision-making based on water availability; 2. Clarify the contribution of multi-source rivers to different spatial areas of lakes and construct a differentiated joint regulation model; 3. Accurately simulate the process of water evaporation, infiltration, and transformation into water entering the lake along the route, and identify the water conveyance efficiency threshold; 4. Formulate graded regulation schemes for wet, normal, and dry years to achieve optimal synergy between ecological benefits and water resource utilization efficiency. Attached Figure Description

[0015] Figure 1 This is a flowchart of a river-lake water quantity linkage regulation method based on ecological security.

[0016] Figure 2 This is a spatial remote sensing image of Lake Taitma in 2023.

[0017] Figure 3 This is a spatial remote sensing image of Lake Taitma in 2024.

[0018] Figure 4 This is a bar chart showing the area of ​​Lake Taitma in 2024.

[0019] Figure 5 This is a map showing the area of ​​Lake Taitma in 2024.

[0020] Figure 6 This is a map showing the frequency of flooding and the corresponding distribution of different water bodies for different periods from 2005 to 2025. Detailed Implementation

[0021] The specific embodiments of the present invention are described below to enable those skilled in the art to understand the present invention. However, it should be understood that the present invention is not limited to the scope of the specific embodiments. For those skilled in the art, various changes are obvious as long as they are within the spirit and scope of the present invention as defined and determined by the appended claims. All inventions utilizing the concept of the present invention are protected.

[0022] Lake Taitma is a typical inland terminal lake jointly fed by the Tarim River and the Cherchen River. Its water volume changes directly affect the stability of the ecological corridor in the lower reaches of the Tarim River and the ecological security of the desert region. In recent years, influenced by climate change and human activities, the inflow process has changed significantly, and the lake area exhibits strong interannual and seasonal fluctuations, necessitating research on the lake's water volume variation patterns and ecological regulation mechanisms. This embodiment takes Lake Taitma as the research object, comprehensively analyzing the relationship between lake surface change characteristics and inflow response through remote sensing data, hydrological statistical analysis, and hydrodynamic numerical simulation. Based on this analysis, a joint regulation scheme between the two rivers under multi-objective protection is proposed.

[0023] like Figure 1 As shown, a river-lake water quantity linkage regulation method based on ecological security includes the following steps: S1: Extract historical multi-source remote sensing image data of the target river-lake basin, and use the random forest learning algorithm to combine the original spectral bands with various water body indices to construct a water body extraction model, extract the monthly lake surface area sequence of the target river-lake basin, and generate a monthly binary water body map.

[0024] Step S1 specifically includes: S11: Extract multi-source remote sensing image data of the target river-lake basin, such as Figure 2 and Figure 3 As shown, the surface reflectance dataset of the target river-lake basin is obtained by preprocessing. In this embodiment, Landsat TM / ETM+ / OLI (30m resolution, 16-day revisit) and Sentinel-2 MSI (10m resolution, 5-day revisit) satellite images of the target river-lake basin from 2005 to 2025 were acquired, along with MODIS daily reflectance products (250m resolution) to fill in the gaps. Radiometric calibration, atmospheric correction (using the 6S model), cloud masking (using the Fmask algorithm), and geometric registration were performed on the images to obtain the surface reflectance dataset of the target river-lake basin.

[0025] S12: Construct a feature space that integrates multispectral indices and original bands, and input it into a random forest classifier to extract water body pixels; In this embodiment, the multispectral indices include the Normalized Difference Water Index (MNDWI), the Automatic Water Extraction Index (AWEIsh), the Multiband Water Index (MBWI), and the Land Water Index (LSWI). ; in, The reflectivity is for the green band (Landsat band 3, center wavelength 0.56µm). Reflectance is measured in shortwave infrared 1 (Landsat band 6, center wavelength 1.65µm). The Normalized Difference Water Index (MNDWI) is used to enhance water body signals and suppress buildings and vegetation.

[0026] ; in, The reflectance is for near-infrared (Landsat band 5, center wavelength 0.86µm). The reflectance is measured in shortwave infrared 2 (Landsat band 7, center wavelength 2.2µm). The Automatic Water Extraction Index (AWEIsh) can effectively distinguish water bodies from shadows and dark surfaces.

[0027] ; in, Reflectivity in the red band (Landsat band 4, center wavelength 0.66µm); .

[0028] The Normalized Differential Water Index (MNDWI), Automatic Water Extraction Index (AWEIsh), Multiband Water Index (MBWI), and Land Water Index (LSWI) are stacked on green, blue, red, shortwave infrared band 1, and shortwave infrared band 2 to obtain a 10-dimensional feature vector. ; ; in, This refers to the blue band reflectivity.

[0029] Obtain the feature vector of each pixel in the remote sensing image. , i This is the cell number.

[0030] Construct a random forest classifier by inputting the feature vector of each pixel into... K In each decision tree, the output is the prediction result for the pixel by each decision tree. 1 indicates that the pixel is a water body, and 0 indicates that the pixel is not a water body; Based on the prediction results of each decision tree for the pixel Calculate the probability that a pixel belongs to a water body. ; ; in, k Number the decision tree.

[0031] The random forest classifier in this embodiment is composed of... K It consists of several decision trees, each of which draws samples from the training set based on a bootstrap sampling method, and randomly selects at each node. m try The features were used for splitting. Training samples were generated through high-resolution image visual interpretation and field surveys (validated by drones in 2024), collecting a total of 4000 samples (2000 for water bodies and 2000 for non-water bodies), which were divided into training and validation sets in a 7:3 ratio. Parameters were optimized using grid search: the number of decision trees. K =150, maximum depth 20, number of node splitting features m try =3; After classification, a 3×3 median filter is used to eliminate isolated noise.

[0032] S13: Based on the area of ​​each cell A i (Landsat is 900 m², Sentinel is 100 m²), calculate the monthly lake surface area of ​​the target river-lake basin. The monthly area data sequence of the lake surface water body is obtained, and a monthly binary water body map is generated. In the monthly binary water body map, each pixel is marked as 1 or 0 to indicate whether it belongs to the water body. ; in, This represents the probability of belonging to a body of water. pixels, N The probability that it belongs to a body of water. The number of pixels, t For months.

[0033] This invention employs a multi-index fusion random forest algorithm, which fully utilizes spectral difference features to overcome the shortcomings of the single threshold method in misjudging complex surfaces, thus providing a highly reliable data foundation for subsequent analysis.

[0034] like Figure 4 and Figure 5As shown, considering both the area process and spatial pattern, the surface changes of Lake Taitma in 2024 exhibited a typical annual rhythm of "spring expansion—summer contraction—rapid recovery / expansion in autumn." In terms of area, the lake surface was generally at a high level from January to April and gradually expanded (due to the combined effects of spring replenishment and low evaporation). From May to early September, the lake surface contracted significantly and reached its lowest value of the year at the end of summer, reflecting the lake's sensitive response to changes in water volume when high temperatures and strong evaporation are combined with weak inflow. Starting in late September, the lake surface rapidly recovered, and significant expansion occurred in October and November (especially in the latter part of the month), reaching its peak for the year. This indicates that the phased replenishment / inflow process in autumn played a decisive role in the lake's recovery. Spatially, at the beginning of the year, the lake surface consisted mainly of a relatively stable central area plus scattered shallow patches on the periphery. Spring expansion was characterized by an increase in shallow waters and peripheral wetland patches, along with enhanced connectivity. During the summer shrinkage phase, the water body became significantly fragmented, with a large amount of shallow water receding from the periphery, leaving only the central lake and some deep / depression areas. In the autumn recovery phase, the water body first rapidly spread along the inflow channels and the northeastern recharge area, then expanded into the low-lying areas of the lake basin and connected with the central lake area, forming a larger area of ​​continuous water surface and wetland patches. Overall, the lake surface changes in 2024 reflected both the strong constraints of climate evaporation on the tailrace lake and the dominant role of the autumn recharge process in the rapid expansion of the lake surface and the improvement of spatial connectivity.

[0035] S2: Calculate the historical inundation frequency of each pixel based on the monthly binary water body map of each historical month, and analyze the changing trend of the lake surface water area based on the changing trend of the monthly area data sequence to screen the areas affected by the rivers flowing into the lake.

[0036] Step S2 specifically includes: S21: Based on whether the pixels in the monthly binary water body maps of each historical month are binary values ​​of water bodies, determine their binarized values. Calculate the flooding frequency for each pixel's history (between 2005 and 2025). ; ; in, T Total number of months in history The coordinates of the pixel; like Then the corresponding pixel will be divided into a permanent water body area, if If the condition is met, the corresponding pixel will be classified as a seasonal water body area; otherwise, the corresponding pixel will be classified as a non-submerged area.

[0037] S22: Calculate the trend of lake surface water area change for any two months based on the monthly data series of lake surface water area. S ; ; in, Any two months The surface area of ​​the lake, It is a symbolic function; The sign function transforms the monthly area difference into a ternary determination of "+1, 0, -1", eliminating dimensional differences.

[0038] S23: Utilizing the trend of lake surface water area changes S Calculate the variance of the trend of lake surface water area change in monthly area data series. ; ; in, p This refers to the classification of months with the same lake surface area in the monthly area data series. The number of months in the categorized month group. g The number of categories for each month group; Categorical month groups group equal-area values ​​in a monthly area data series together. For example, if multiple months in a monthly area data series have an area of ​​"0.35 km²" (multiple data points with the same value), then these data points constitute a categorical month group. The existence of categorical month groups reduces variance and requires correction.

[0039] S24: Variance based on the trend of lake surface water area change Calculate the standardized test statistic Z ; ; S25: If If the monthly area data series shows a significant upward or downward trend in lake surface area, then the lake surface area shows a significant upward or downward trend; otherwise, there is no significant upward or downward trend.

[0040] In this embodiment, the standardized test statistics for the period 2005-2025 are... Z =4.32>1.96, confirming a highly significant upward trend in the surface area of ​​Lake Taitma, thus demonstrating the positive effect of long-term ecological water transfer.

[0041] S26: Based on the inundation frequency of pixels within permanent and seasonal water bodies. Screening the pixel inundation frequency of rivers flowing into the lake during independent replenishment periods. Screening the areas of influence of rivers flowing into the lake on the lake .

[0042] like Figure 6As shown, this embodiment uses flood inundation frequency (WIF) to analyze the changes in lake surface area in different regions of Lake Taitma from 2005 to 2025. It also divides the lake into ten-year periods to obtain the flood inundation frequency and spatial distribution of different water body types for the preceding and following periods. Flood inundation frequency indicates the spatial distribution of the percentage of flood inundations in different periods; darker blue indicates a lower flood inundation frequency, and darker red indicates a higher flood inundation frequency. Considering the extremely significant intra-annual variation of Lake Taitma, a WIF greater than 50% is defined as a high-frequency permanent water body (hereinafter referred to as permanent water body), and a WIF between 5% and 50% is defined as a low-frequency seasonal water body (hereinafter referred to as seasonal water body).

[0043] S3: Construct a nonlinear response model of inflow volume-lake surface area based on monthly area data series, and output the total inflow volume required each month according to the target lake surface area required to be maintained for different ecological needs of the lake each month.

[0044] Step S3 specifically includes: S31: Calculate the lake surface area for each month in the monthly area data series. Total inflow of water into the lake by each month Nonlinear regression was performed to establish a nonlinear response model of inflow volume versus lake surface area. ; in, These are the nonlinear fitting coefficients; S32: Input the target lake surface area required to be maintained for different ecological needs each month into the nonlinear response model of lake inflow-lake surface area, and calculate the total inflow required for each month under different ecological objectives.

[0045] In this embodiment, the data table of different target lake surface areas and required inflow volume for Lake Taitma is shown in Table 1 below: Table 1. Data on the required inflow volume of different target lake areas in Lake Taitma

[0046] S4: Construct a MIKE hydrodynamic model of the river flowing into the lake to simulate the evolution of the water flow at different cross-sections (Yottatjang) of the target river flowing into the lake (Chelsen River) to the lake inlet (Konlak cross-section); The governing equations of the MIKE hydrodynamic model: Continuity equation: ; Momentum equation: ; in, Q The cross-sectional flow rate of the river flowing into the lake. X The horizontal coordinates are along the direction of the river flowing into the lake. The width of the water surface. h For the cross-sectional water depth, Lateral inflow along the river channel (including sources and sinks such as rainfall, evaporation, and infiltration). A The cross-sectional area of ​​the river flowing into the lake. g It is the acceleration due to gravity. C For the Xie Cai coefficient, R The radius is the hydraulic radius.

[0047] S5: Set the water volume of the rivers flowing into the lake, input the MIKE hydrodynamic model, simulate the ecological discharge of the rivers flowing into the lake, calculate the actual monthly water volume flowing into the lake during the simulation process, calculate the inflow efficiency of the monthly ecological water volume of the rivers flowing into the lake, and construct the ecological discharge model of the rivers flowing into the lake.

[0048] Step S5 specifically includes: S51: Based on the annual runoff characteristics and ecological regulation needs of the rivers flowing into the lake, the water discharge volume of the rivers into the lake is set, and different monthly discharge volumes are set as discharge step sizes. The ecological discharge volume of the rivers flowing into the lake is simulated, and the monthly discharge volume is converted into a daily flow process and input into the MIKE hydrodynamic model to calculate the daily inflow volume of the rivers flowing into the lake. The daily inflow volume is then integrated to obtain the actual monthly inflow volume. ; In this embodiment, based on the multi-year runoff characteristics of the Cherchen River and the needs of ecological regulation, a total of 10 discharge steps are set, ranging from 100 million m³ to 550 million m³, with a step size of 50 million m³.

[0049] S52: Calculate the total monthly evaporation of rivers flowing into the lake. V evap and total monthly infiltration V inf Calculate the theoretical ecological discharge volume of the rivers flowing into the lake. V release ; ; in, The monthly change in the water storage of the rivers flowing into the lake (0 is taken in long-term simulation). S53: Based on theoretical ecological discharge volume V release and the actual monthly inflow of water into the lake Calculate the inflow efficiency of monthly ecological water volume of rivers flowing into the lake ; S54: Obtain the actual monthly inflow to the lake under each simulated discharge step condition. Theoretical ecological discharge volume V release A piecewise function was used to fit the relationship between the actual monthly inflow into the lake and the theoretical ecological outflow, thus obtaining the ecological outflow model of the rivers flowing into the lake. ; in, The slope (reflects the marginal efficiency of lake inflow (dimensionless)). The intercept is (in billions of m³). The critical threshold for segmentation (in billions of m³) is determined by minimizing the sum of squared residuals from the segmented regression.

[0050] In this embodiment, a simulation data table of the conversion between theoretical ecological discharge volume and actual monthly inflow volume for different discharge stages (different discharge scenarios) is obtained, as shown in Table 2 below: Table 2. Data on ecological outflow and actual monthly inflow into the lake.

[0051] The parameter fitting results of the ecological discharge model of rivers flowing into the lake are as follows: ; when At that time, the marginal inflow efficiency was only 0.28, indicating that more than 70% of the water volume was dissipated through infiltration and evaporation during transport; when Subsequently, the marginal efficiency jumped to 0.52, indicating a significant improvement in the efficiency of water inflow into the lake. Therefore, it is necessary to determine the critical threshold for optimal economic and ecological outflow. =300 million m³, meaning that ecological scheduling should prioritize ensuring that the discharge volume is not lower than this threshold in order to avoid the ineffective dissipation of water resources during transportation along the route.

[0052] By setting different discharge volumes to simulate different discharge scenarios, the study revealed the "threshold effect" of water conveyance in wide and shallow rivers in arid regions. It provides management decision-makers with a clear quantitative red line: releasing water below the critical threshold will result in huge waste of water resources (lake inflow efficiency less than 30%), while releasing water above the critical threshold significantly improves the lake replenishment efficiency (over 40%), providing a physical basis for determining economically reasonable ecological discharge volumes.

[0053] S6: Input the total monthly inflow volume required for different ecological objectives into the ecological discharge model of the rivers flowing into the lake, calculate the theoretical ecological discharge volume of the rivers flowing into the lake, and construct a weighted multi-objective function with the theoretical ecological discharge volume as the decision variable based on the influence area of ​​the rivers flowing into the lake, so as to regulate the ecological discharge volume of the rivers flowing into the lake under different ecological objectives.

[0054] Step S6 specifically includes: S61: Input the total monthly inflow volume required for different ecological objectives into the ecological discharge model of the inflowing rivers to calculate the theoretical ecological discharge volume of the inflowing rivers. ; S62: Based on the impact area of ​​rivers flowing into the lake on the lake Calculate the area of ​​impact of rivers flowing into the lake on the lake under different ecological objectives. , W For the affected area The number of internal pixels; S63: Construct a weighted multi-objective function with the theoretical ecological discharge volume as the decision variable. J With weighted multi-objective function J With the goal of minimizing the value, the theoretical ecological discharge volume of rivers flowing into the lake under different ecological objectives is optimized, and the ecological discharge volume of rivers flowing into the lake under different ecological objectives is regulated. ; in, The ideal impact area of ​​rivers flowing into the lake under different ecological objectives. The baseline natural flow of the rivers flowing into the lake for the corresponding months under different ecological objectives. We can take the average natural inflow of historical rivers flowing into the lake under different ecological objectives. The number of months maintained for different ecological objectives. The weighting coefficients for the area affected by rivers flowing into the lake and the flow smoothing coefficients are respectively taken as 0.1~0.3 and 0.9~0.7.

[0055] The ecological objectives of this embodiment include low water threshold, water balance, and high water management; based on a weighted multi-objective function. J The ecological discharge control strategies for the Cherchen River and the Tarim River are shown in Table 3 below: Table 3. Ecological Discharge Regulation Strategies for the Cherchen River and Tarim River

[0056] This invention incorporates the hydrological differences and spatial control zoning of the two rivers into a single optimization framework. By setting differentiated weights (emphasizing minimum water levels in dry years and controlling water levels in wet years) and flow smoothing constraints, it generates a scheduling scheme that meets both ecological needs and is engineering-feasible. Compared to traditional "one-size-fits-all" scheduling, the joint regulation method of this invention saves 18% to 25% of the total outflow and improves the lake inflow efficiency by 12 to 18 percentage points.

Claims

1. A method for coordinated regulation of river and lake water volume based on ecological security, characterized in that, Includes the following steps: S1: Extract historical multi-source remote sensing image data of the target river-lake basin, use the random forest learning algorithm to combine the original spectral bands with multiple water body indices to construct a water body extraction model, extract the monthly lake surface area sequence of the target river-lake basin, and generate a monthly binary water body map; S2: Calculate the historical inundation frequency of each pixel based on the monthly binary water body map of each historical month, and analyze the changing trend of the lake surface water area based on the changing trend of the monthly area data sequence to screen the areas affected by the rivers flowing into the lake. S3: Construct a nonlinear response model of inflow volume-lake surface area based on monthly area data series, and output the total inflow volume required each month according to the target lake surface area required to be maintained for different ecological needs of the lake each month. S4: Construct a MIKE hydrodynamic model of the river flowing into the lake to simulate the evolution of water flow from different cross sections of the target river into the lake to the lake inlet; S5: Set the water volume of the rivers flowing into the lake, input the MIKE hydrodynamic model, simulate the ecological discharge of the rivers flowing into the lake, calculate the actual monthly water volume flowing into the lake during the simulation process, calculate the inflow efficiency of the monthly ecological water volume of the rivers flowing into the lake, and construct the ecological discharge model of the rivers flowing into the lake. S6: Input the total monthly inflow volume required for different ecological objectives into the ecological discharge model, calculate the theoretical ecological discharge volume of the rivers flowing into the lake, and construct a weighted multi-objective function with the theoretical ecological discharge volume as the decision variable based on the influence area of ​​the rivers flowing into the lake, so as to regulate the ecological discharge volume of the rivers flowing into the lake under different ecological objectives.

2. The river-lake water quantity linkage regulation method based on ecological security according to claim 1, characterized in that, Step S1 includes: S11: Extract multi-source remote sensing image data of the target river-lake basin and preprocess it to obtain the surface reflectance dataset of the target river-lake basin; S12: Construct a feature space that integrates multispectral indices and original bands, and input it into a random forest classifier to extract water body pixels; Multispectral indices include the Normalized Difference Water Index (MNDWI), the Automatic Water Extraction Index (AWEIsh), the Multiband Water Index (MBWI), and the Land Water Index (LSWI). The Normalized Differential Water Index (MNDWI), Automatic Water Extraction Index (AWEIsh), Multiband Water Index (MBWI), and Land Water Index (LSWI) are stacked on green, blue, red, shortwave infrared band 1, and shortwave infrared band 2 to obtain a 10-dimensional feature vector. ; ； in, Blue band reflectivity; Obtain the feature vector of each pixel in the remote sensing image. , i For the cell number; Construct a random forest classifier by inputting the feature vector of each pixel into... K In each decision tree, the output is the prediction result for the pixel by each decision tree. 1 indicates that the pixel is a water body, and 0 indicates that the pixel is not a water body; Based on the prediction results of each decision tree for the pixel Calculate the probability that a pixel belongs to a water body. ; ； in, k Number the decision tree; S13: Based on the area of ​​each cell A i Calculate the monthly surface water area of ​​the target river-lake basin. The monthly area data sequence of the lake surface water body is obtained, and a monthly binary water body map is generated. In the monthly binary water body map, each pixel is marked as 1 or 0 to indicate whether it belongs to the water body. ； in, This represents the probability of belonging to a body of water. pixels, N The probability that it belongs to a body of water. The number of pixels, t For months.

3. The river-lake water quantity linkage regulation method based on ecological security according to claim 2, characterized in that, The multispectral indices include the Normalized Difference Water Index (MNDWI), the Automatic Water Extraction Index (AWEIsh), the Multiband Water Index (MBWI), and the Land Water Index (LSWI). ； in, For green band reflectivity, The reflectivity is 1 for shortwave infrared. ； in, Near-infrared reflectance, For shortwave infrared reflectivity; ； in, Reflectivity in the red band; 。 4. The river-lake water quantity linkage regulation method based on ecological security according to claim 2, characterized in that, Step S2 includes: S21: Based on whether the pixels in the monthly binary water body maps of each historical month are binary values ​​of water bodies, determine their binarized values. Calculate the flooding frequency for each pixel's history. ; ； in, T Total number of months in history The coordinates of the pixel; like Then the corresponding pixel will be divided into a permanent water body area, if If the corresponding pixel is in the seasonal water body area, then the corresponding pixel will be classified as a non-submerged area; otherwise, the corresponding pixel will be classified as a non-submerged area. S22: Calculate the trend of lake surface water area change for any two months based on the monthly data series of lake surface water area. S ; ； in, Any two months The surface area of ​​the lake, It is a symbolic function; S23: Utilizing the trend of lake surface water area changes S Calculate the variance of the trend of lake surface water area change in monthly area data series. ; ； in, p This refers to the classification of months with the same lake surface area in the monthly area data series. The number of months in the categorized month group. g The number of categories for each month group; S24: Variance based on the trend of lake surface water area change Calculate the standardized test statistic Z ; ； S25: If If the monthly area data series shows a significant upward or downward trend in lake surface area, then the lake surface area shows a significant upward or downward trend; otherwise, there is no significant upward or downward trend. S26: Based on the inundation frequency of pixels within permanent and seasonal water bodies. Screening the pixel inundation frequency of rivers flowing into the lake during independent replenishment periods. Screening the areas of influence of rivers flowing into the lake on the lake .

5. The river-lake water quantity linkage regulation method based on ecological security according to claim 4, characterized in that, Step S3 includes: S31: Calculate the lake surface area for each month in the monthly area data series. Total inflow of water into the lake by each month Nonlinear regression was performed to establish a nonlinear response model of inflow volume versus lake surface area. ； in, These are the nonlinear fitting coefficients; S32: Input the target lake surface area required to be maintained for different ecological needs each month into the nonlinear response model of lake inflow-lake surface area, and calculate the total inflow required for each month under different ecological objectives.

6. The river-lake water quantity linkage regulation method based on ecological security according to claim 5, characterized in that, Step S5 includes: S51: Based on the annual runoff characteristics and ecological regulation needs of the rivers flowing into the lake, the water discharge volume of the rivers into the lake is set, and different monthly discharge volumes are set as discharge step sizes. The ecological discharge volume of the rivers flowing into the lake is simulated, and the monthly discharge volume is converted into a daily flow process and input into the MIKE hydrodynamic model to calculate the daily inflow volume of the rivers flowing into the lake. The daily inflow volume is then integrated to obtain the actual monthly inflow volume. ; S52: Calculate the total monthly evaporation of rivers flowing into the lake. V evap and total monthly infiltration V inf Calculate the theoretical ecological discharge volume of the rivers flowing into the lake. V release ; ； in, This represents the monthly change in the water storage of rivers flowing into the lake. S53: Based on theoretical ecological discharge volume V release and the actual monthly inflow of water into the lake Calculate the inflow efficiency of monthly ecological water volume of rivers flowing into the lake ; S54: Obtain the actual monthly inflow to the lake under each simulated discharge step condition. Theoretical ecological discharge volume V release A piecewise function was used to fit the relationship between the actual monthly inflow into the lake and the theoretical ecological outflow, thus obtaining the ecological outflow model of the rivers flowing into the lake. ； in, The slope The intercept is... This represents the critical threshold for segmentation.

7. The river-lake water quantity linkage regulation method based on ecological security according to claim 6, characterized in that, Step S6 includes: S61: Input the total monthly inflow volume required for different ecological objectives into the ecological discharge model of the inflowing rivers to calculate the theoretical ecological discharge volume of the inflowing rivers. ; S62: Based on the impact area of ​​rivers flowing into the lake on the lake Calculate the area of ​​impact of rivers flowing into the lake on the lake under different ecological objectives. , W For the affected area The number of internal pixels; S63: Construct a weighted multi-objective function with the theoretical ecological discharge volume as the decision variable. J With weighted multi-objective function J With the goal of minimizing the value, the theoretical ecological discharge volume of rivers flowing into the lake under different ecological objectives is optimized, and the ecological discharge volume of rivers flowing into the lake under different ecological objectives is regulated. ； in, The ideal impact area of ​​rivers flowing into the lake under different ecological objectives. This represents the baseline natural flow of rivers flowing into the lake for the corresponding months under different ecological objectives. The number of months maintained for different ecological objectives. These are the weighting coefficients for the area affected by rivers flowing into the lake and the flow smoothing coefficients, respectively.

Images