Regional water yield calculation method based on InVEST model

Abstract

The invention relates to the technical field of water yield calculation, in particular to a regional water yield calculation method based on an InVEST model, and the method comprises the steps: regularly obtaining a surface remote sensing image of a target region according to a preset sampling period, so as to obtain a surface remote sensing image sequence of the target region, carrying out the preprocessing of the surface remote sensing image sequence of the target region, and obtaining a surface remote sensing image sequence of the target region; obtaining a standard earth surface remote sensing image sequence of the target area; and performing image segmentation on the standard earth surface remote sensing image sequence of the target area to obtain a segmented image sequence set corresponding to the standard earth surface remote sensing image sequence. According to the method, the remote sensing image is preprocessed and segmented, fine land coverage data can be obtained, the data is crucial for accurately evaluating the influence of different land types on the water yield, and the area proportion of the segmented image is input into the trained water yield prediction model, so that the water yield prediction accuracy is improved. And high-precision water yield prediction can be carried out through historical data and a learning result of the model.

Description

Technical Field

[0001] The present invention relates to the technical field of water production calculation, and in particular to a method for calculating regional water production based on an InVEST model. Background Art

[0002] The InVEST model is an open source geographic information system tool used to assess ecosystem services and their value. It simulates different land use and management scenarios, predicts their impact on ecosystem services, and helps evaluate the potential outcomes of policy and planning decisions. In the process of land planning and development, understanding regional water yield helps assess the impact of development activities on water resources and avoid water shortages caused by overdevelopment. Climate change may affect regional precipitation patterns and water resource distribution. By calculating water yield, we can better predict and respond to water resource challenges brought about by climate change.

[0003] Traditional methods often rely on ground measurement data, which is expensive and time-consuming to collect and has difficulty covering wide areas, especially in remote or inaccessible areas. Traditional methods may not fully consider surface changes and dynamic changes in land use, resulting in limited accuracy in water yield estimates. Moreover, due to the cumbersome data collection and processing process, traditional methods are difficult to achieve real-time or dynamic monitoring, which may lead to delayed data updates and failure to reflect environmental changes in a timely manner. Traditional methods also lack adaptability to different terrains, climatic conditions and land use types, making it difficult to adjust and optimize methods under different environments and needs. Summary of the invention

[0004] The purpose of the present invention is to propose a method for calculating regional water production based on the InVEST model in response to the problems existing in the background technology.

[0005] The technical solution of the present invention is a method for calculating regional water production based on the InVEST model, comprising:

[0006] Acquiring a surface remote sensing image of a target area at a preset sampling period to obtain a surface remote sensing image sequence of the target area, and performing a preprocessing operation on the surface remote sensing image sequence of the target area to obtain a standard surface remote sensing image sequence of the target area;

[0007] Performing image segmentation on the standard surface remote sensing image sequence of the target area to obtain a segmented image sequence set corresponding to the standard surface remote sensing image sequence;

[0008] Counting the pixel areas of each segmented image sequence in the segmented image sequence set to obtain a segmented image pixel area sequence set, and calculating the area ratio of the segmented images based on the segmented image pixel area sequence set to obtain a segmented image area ratio sequence set;

[0009] Inputting the segmented image area ratio sequence set into a trained water production prediction model, and outputting a first water production corresponding to the target area through the trained water production prediction model;

[0010] The water production of the target area is calculated by using the InVEST model based on the hydrological and meteorological variables of the target area to obtain a second water production corresponding to the target area, a ground-water material balance model is constructed based on formation parameters, and the water production of the target area is simulated based on the ground-water material balance model to obtain a third water production corresponding to the target area;

[0011] The first water production, the second water production and the third water production corresponding to the target area are weighted on average based on a preset weight coefficient to obtain the water production corresponding to the target area.

[0012] Preferably, the segmented image sequence set includes a grassland remote sensing image sequence, a cultivated land remote sensing image sequence, a forest remote sensing image sequence, a bare land remote sensing image sequence, a town remote sensing image sequence and a water body remote sensing image sequence.

[0013] Preferably, preprocessing the surface remote sensing image sequence of the target area to obtain a standard surface remote sensing image sequence of the target area includes:

[0014] Binarizing the surface remote sensing image sequence to obtain a surface remote sensing binary image sequence;

[0015] Traversing each pixel point in the surface remote sensing binary image, and when traversing to a current pixel point, judging whether the current pixel point is an isolated point by using a preset isolated point judgment rule;

[0016] If the current pixel is an isolated point, the current pixel is deleted; otherwise, the next pixel of the current pixel is traversed until each pixel in the surface remote sensing binary image is traversed to obtain a standard surface remote sensing image.

[0017] Preferably, the isolated point judgment rule includes a first isolated point judgment formula, a second isolated point judgment formula, a third isolated point judgment formula and a fourth isolated point judgment formula, and the first isolated point judgment formula is as follows:

[0018] B(P0)=1;

[0019] Wherein, P0 represents the current pixel point, and B(P0) represents the pixel value of the current pixel point;

[0020] The second isolated point judgment formula is as follows:

[0021]

[0022] Where n represents the serial number of the pixel adjacent to the current pixel, N4 represents the first digital sequence, and N4 = {1, 2, 3, 4}, a 2n-1 Indicates whether the pixel value of the adjacent pixel numbered 2n-1 is 0, and

[0023] The third isolated point judgment formula is as follows:

[0024]

[0025] Wherein, N8 represents a second digital sequence, and N8={1,2,...,8};

[0026] The fourth isolated point determination formula is as follows:

[0027]

[0028] Among them, C n Indicates whether the pixel value of the adjacent pixel with sequence number n is 1, and

[0029] Preferably, performing image segmentation on the standard surface remote sensing image sequence of the target area to obtain a set of segmented image sequences corresponding to the standard surface remote sensing image sequence includes:

[0030] A plurality of initial segmentation thresholds are preset, and grayscale mapping is performed on the standard surface remote sensing image based on the plurality of initial segmentation thresholds, and the grayscale mapping formula is as follows:

[0031]

[0032] Among them, f(x,y) represents the grayscale value of the standard surface remote sensing image, l0,l1,...,l n-1 ,l n Represents the grayscale of the standard surface remote sensing image, t1, t2, ..., t n-1 ,t n represents multiple preset initial segmentation thresholds, and L represents the total number of gray levels of the standard surface remote sensing image;

[0033] An objective function is constructed based on the multiple initial segmentation thresholds and the grayscale mapping, and the objective function is as follows:

[0034]

[0035] Among them, H(t1,t2,...,t n-1 ,tn ) represents the objective function of the standard surface remote sensing image, p j Represents the probability that the gray value of the standard surface remote sensing image is j.

[0036] Preferably, performing image segmentation on the standard surface remote sensing image sequence of the target area to obtain a set of segmented image sequences corresponding to the standard surface remote sensing image sequence also includes:

[0037] Solving the objective function based on an optimization algorithm to obtain an optimal segmentation threshold combination to obtain multiple optimal segmentation thresholds;

[0038] The standard surface remote sensing image sequence of the target area is segmented based on the multiple optimal segmentation thresholds to obtain a segmented image sequence set corresponding to the standard surface remote sensing image sequence.

[0039] Preferably, solving the objective function based on an optimization algorithm to obtain an optimal segmentation threshold combination includes:

[0040] Set the initialization parameters of the optimization algorithm to initialize the population position. The population position initialization formula is as follows:

[0041]

[0042] Among them, x t and x t+1 represents the position of the individual at the t-th iteration and the t+1-th iteration, and p represents the preset parameter for controlling the nonlinear mapping;

[0043] Calculate the fitness value of the individual and calculate the elite individual position. The elite individual position calculation formula is as follows:

[0044]

[0045] in, represents the reverse solution of the j-th dimension individual at the t-th iteration, x i,j (t) represents the elite solution of the j-th dimension individual at the t-th iteration, k represents a random number in the interval [0,1], u j (t) represents the maximum position of the individual at the tth iteration, l j (t) represents the minimum position of the individual at the tth iteration;

[0046] Marking a preset number of the individuals with the best fitness values ​​as leader individuals, and updating the positions of the remaining individuals in the population;

[0047] Calculate and sort the updated individual fitness values, obtain elite individuals, and update the elite individual positions;

[0048] Repeat the above steps until the optimal solution is found or the maximum number of iterations of the optimization algorithm is reached to obtain the optimal segmentation threshold combination.

[0049] Preferably, the water yield prediction model adopts a BP neural network, and the water yield prediction model includes an input layer, a hidden layer and an output layer, wherein the input layer is used to input an area ratio sequence of grassland remote sensing images, cultivated land remote sensing images, forest land remote sensing images, bare land remote sensing images, urban remote sensing images and water body remote sensing images,

[0050] The hidden layer is used to learn the nonlinear relationship between the area proportion sequence of grassland remote sensing images, cultivated land remote sensing images, forest land remote sensing images, bare land remote sensing images, urban remote sensing images and water body remote sensing images and the water yield, and the output layer is used to quantify and output the learned nonlinear relationship between the area proportion sequence of grassland remote sensing images, cultivated land remote sensing images, forest land remote sensing images, bare land remote sensing images, urban remote sensing images and water body remote sensing images and the water yield, so as to obtain the first water yield corresponding to the target area.

[0051] Preferably, the water yield of the target area is calculated by using the InVEST model based on the hydrological and meteorological variables of the target area to obtain a second water yield corresponding to the target area, including:

[0052] A water production model is constructed based on the InVEST model. The water production model is as follows:

[0053]

[0054] Among them, Y(x) represents water yield, Kc(x) represents vegetation evapotranspiration coefficient, ETO(x) represents reference evapotranspiration, P(x) represents precipitation, and W represents non-physical parameters;

[0055] The water production of the target area is calculated through a water production model based on the hydrological and meteorological variables of the target area to obtain a second water production corresponding to the target area.

[0056] Preferably, the groundwater material balance model is as follows:

[0057]

[0058] Among them, p R represents the gas layer pressure, Z represents the gas deviation coefficient under the gas layer pressure, p i represents the original formation pressure, Z i represents the gas deviation coefficient under the original formation pressure, G represents the total cumulative gas production and storage in the target area, and G p Indicates the accumulated gas volume, T sc Indicates the temperature under standard conditions, psc Indicates the pressure under standard conditions, W p Indicates the third water production, B w Represents the volume coefficient of formation water.

[0059] Compared with the prior art, the above technical solution of the present invention has the following beneficial technical effects:

[0060] 1. The present invention can obtain detailed land cover data by preprocessing and segmenting remote sensing images. These data are crucial for accurately evaluating the impact of different land types on water production. By inputting the area ratio of the segmented image into the trained water production prediction model, it is helpful to make high-precision water production predictions through historical data and model learning results. It combines remote sensing image data, InVEST model and ground-water material balance model, and comprehensively considers the impact of multiple factors on water production. This comprehensive method can overcome the limitations of a single method and improve the comprehensiveness and accuracy of water production estimation. In addition, by regularly acquiring and processing surface remote sensing images, it can monitor changes in the target area in real time and adjust the water production prediction in time, which can cope with the impact of natural environmental changes on water resources and improve the timeliness and reliability of the prediction.

[0061] 2. The present invention can provide more detailed and accurate information for water resource management through multi-faceted calculation and analysis of water production, helping decision makers make scientific and reasonable water resource management decisions. By relying on remote sensing images and model predictions, it can reduce dependence on traditional ground data collection and reduce data acquisition costs. At the same time, it can effectively evaluate water production in areas that are difficult to reach or where data collection is difficult. In addition, the remote sensing image acquisition cycle, model parameters and weight coefficients can be adjusted according to the characteristics and needs of different regions, so that the method can adapt to changes in different environments and needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0062] Figure 1 The figure is a flow chart of an overall method in one embodiment of the present invention. DETAILED DESCRIPTION

[0063] Embodiment 1, as Figure 1 As shown, the present invention proposes a method for calculating regional water production based on the InVEST model, including:

[0064] S1, acquiring the surface remote sensing image of the target area at a preset sampling period to obtain a surface remote sensing image sequence of the target area, and performing a preprocessing operation on the surface remote sensing image sequence of the target area to obtain a standard surface remote sensing image sequence of the target area;

[0065] S2. performing image segmentation on the standard surface remote sensing image sequence of the target area to obtain a segmented image sequence set corresponding to the standard surface remote sensing image sequence;

[0066] S3, counting the pixel areas of each segmented image sequence in the segmented image sequence set to obtain a segmented image pixel area sequence set, and calculating the area ratio of the segmented images based on the segmented image pixel area sequence set to obtain a segmented image area ratio sequence set;

[0067] S4, inputting the segmented image area ratio sequence set into the trained water production prediction model, and outputting the first water production corresponding to the target area through the trained water production prediction model;

[0068] S5. Based on the hydrological and meteorological variables of the target area, the water production of the target area is calculated by using the InVEST model to obtain a second water production corresponding to the target area, a ground-water material balance model is constructed based on the formation parameters, and the water production of the target area is simulated based on the ground-water material balance model to obtain a third water production corresponding to the target area;

[0069] S6. Performing average weighting on the first water production, the second water production, and the third water production corresponding to the target area based on a preset weight coefficient to obtain the water production corresponding to the target area.

[0070] In the present invention, area proportion refers to the calculation of the relative area of ​​each segmented area in the image, usually expressed as a percentage; the InVEST model is a tool for evaluating ecosystem services (such as water output); and average weighted means combining the water volume prediction values ​​from different sources into a comprehensive prediction value according to the set weights.

[0071] Embodiment 2: A method for calculating regional water yield based on the InVEST model proposed in the present invention. Compared with Embodiment 1, this embodiment also includes: a segmented image sequence set including a grassland remote sensing image sequence, a cultivated land remote sensing image sequence, a forest remote sensing image sequence, a bare land remote sensing image sequence, a town remote sensing image and a water body remote sensing image sequence.

[0072] In an optional embodiment, a preprocessing operation is performed on a surface remote sensing image sequence of a target area to obtain a standard surface remote sensing image sequence of the target area, including:

[0073] A1. Binarizing the surface remote sensing image sequence to obtain a surface remote sensing binary image sequence;

[0074] A2, traversing each pixel point in the surface remote sensing binary image, when traversing to the current pixel point, judging whether the current pixel point is an isolated point according to the preset isolated point judgment rule;

[0075] A3. If the current pixel is an isolated point, the current pixel is deleted. Otherwise, the next pixel of the current pixel is traversed until each pixel in the surface remote sensing binary image is traversed to obtain a standard surface remote sensing image.

[0076] It should be noted that a surface remote sensing image sequence refers to a series of images acquired from remote sensing equipment (such as satellites or drones) that cover the same geographic area but are taken at different time points. Such sequences are used to monitor surface changes, analyze dynamic processes, etc. Binarization is an operation in image processing that converts each pixel value in an image into one of two possible values ​​(usually 0 and 1, representing black and white). This process distinguishes the target and background in the image and simplifies subsequent processing. Common methods include threshold processing, in which points in the image with pixel values ​​above a certain threshold are set to 1 (or white), and points below the threshold are set to 0 (or black); isolated points refer to pixels that exist alone in the image without neighboring points in a binary image. They may be caused by noise or discontinuous target areas. Isolated points are usually considered unimportant or unreliable. In many applications, these isolated points need to be deleted or removed through post-processing steps to improve the accuracy of image analysis.

[0077] In an optional embodiment, the isolated point judgment rule includes a first isolated point judgment formula, a second isolated point judgment formula, a third isolated point judgment formula and a fourth isolated point judgment formula, and the first isolated point judgment formula is as follows:

[0078] B(P0)=1;

[0079] Wherein, P0 represents the current pixel point, and B(P0) represents the pixel value of the current pixel point;

[0080] The second isolated point judgment formula is as follows:

[0081]

[0082] Where n represents the serial number of the pixel adjacent to the current pixel, N4 represents the first digital sequence, and N4 = {1, 2, 3, 4}, a 2n-1 Indicates whether the pixel value of the adjacent pixel numbered 2n-1 is 0, and

[0083] The third isolated point judgment formula is as follows:

[0084]

[0085] Wherein, N8 represents a second digital sequence, and N8={1,2,...,8};

[0086] The fourth isolated point judgment formula is as follows:

[0087]

[0088] Among them, C n Indicates whether the pixel value of the adjacent pixel with sequence number n is 1, and

[0089] In an optional embodiment, image segmentation is performed on a standard surface remote sensing image sequence of a target area to obtain a segmented image sequence set corresponding to the standard surface remote sensing image sequence, including:

[0090] B1. Pre-set multiple initial segmentation thresholds, and perform grayscale mapping on the standard surface remote sensing image based on the multiple initial segmentation thresholds. The grayscale mapping formula is as follows:

[0091]

[0092] Among them, f(x,y) represents the grayscale value of the standard surface remote sensing image, l0,l1,...,l n-1 ,l n Represents the grayscale of the standard surface remote sensing image, t1, t2, ..., t n-1 ,t n represents multiple preset initial segmentation thresholds, and L represents the total number of gray levels of the standard surface remote sensing image;

[0093] B2. Construct an objective function based on multiple initial segmentation thresholds and grayscale mapping. The objective function is as follows:

[0094]

[0095]

[0096] Among them, H(t1,t2,...,t n-1 ,t n ) represents the objective function of the standard surface remote sensing image, p j Represents the probability that the gray value of the standard surface remote sensing image is j.

[0097] In an optional embodiment, performing image segmentation on a standard surface remote sensing image sequence of a target area to obtain a segmented image sequence set corresponding to the standard surface remote sensing image sequence further includes:

[0098] B3, solving the objective function based on the optimization algorithm to obtain the optimal segmentation threshold combination, so as to obtain multiple optimal segmentation thresholds;

[0099] B4. Performing image segmentation on the standard surface remote sensing image sequence of the target area based on multiple optimal segmentation thresholds to obtain a set of segmented image sequences corresponding to the standard surface remote sensing image sequence.

[0100] In an optional embodiment, solving the objective function based on an optimization algorithm to obtain an optimal segmentation threshold combination includes:

[0101] C1. Set the initialization parameters of the optimization algorithm to initialize the population position. The population position initialization formula is as follows:

[0102]

[0103] Among them, x t and x t+1 represents the position of the individual at the t-th iteration and the t+1-th iteration, and p represents the preset parameter for controlling the nonlinear mapping;

[0104] C2. Calculate the fitness value of the individual and the position of the elite individual. The calculation formula for the position of the elite individual is as follows:

[0105]

[0106] in, represents the reverse solution of the j-th dimension individual at the t-th iteration, x i,j (t) represents the elite solution of the j-th dimension individual at the t-th iteration, k represents a random number in the interval [0,1], u j (t) represents the maximum position of the individual at the tth iteration, l j (t) represents the minimum position of the individual at the tth iteration;

[0107] C3, mark the preset number of individuals with the best fitness values ​​as leader individuals, and update the positions of the remaining individuals in the population;

[0108] C4, calculate and sort the updated individual fitness values, obtain elite individuals, and update the elite individual positions;

[0109] C5. Repeat the above steps until the optimal solution is found or the maximum number of iterations of the optimization algorithm is reached to obtain the optimal segmentation threshold combination.

[0110] In an optional embodiment, the water yield prediction model adopts a BP neural network, and the water yield prediction model includes an input layer, a hidden layer and an output layer. The input layer is used to input the area proportion sequence of grassland remote sensing images, cultivated land remote sensing images, forest remote sensing images, bare land remote sensing images, urban remote sensing images and water body remote sensing images. The hidden layer is used to learn the nonlinear relationship between the area proportion sequence of grassland remote sensing images, cultivated land remote sensing images, forest remote sensing images, bare land remote sensing images, urban remote sensing images and water body remote sensing images and the water yield. The output layer is used to quantify and output the learned nonlinear relationship between the area proportion sequence of grassland remote sensing images, cultivated land remote sensing images, forest remote sensing images, bare land remote sensing images, urban remote sensing images and water body remote sensing images and the water yield, so as to obtain the first water yield corresponding to the target area.

[0111] It should be noted that during the training process of the water production prediction model, the weights between each layer are adjusted through the back propagation algorithm so that the prediction results gradually approach the actual water production, thereby improving the prediction accuracy of the model.

[0112] In an optional embodiment, the water yield of the target area is calculated by the InVEST model based on the hydrological and meteorological variables of the target area to obtain a second water yield corresponding to the target area, including:

[0113] D1. Construct a water production model based on the InVEST model. The water production model is as follows:

[0114]

[0115] Among them, Y(x) represents water yield, Kc(x) represents vegetation evapotranspiration coefficient, ETO(x) represents reference evapotranspiration, P(x) represents precipitation, and W represents non-physical parameters;

[0116] D2. Based on the hydrological and meteorological variables of the target area, the water production of the target area is calculated through the water production model to obtain a second water production corresponding to the target area.

[0117] In an optional embodiment, the groundwater material balance model is as follows:

[0118]

[0119] Among them, p R represents the gas layer pressure, Z represents the gas deviation coefficient under the gas layer pressure, p i represents the original formation pressure, Z i represents the gas deviation coefficient under the original formation pressure, G represents the total cumulative gas production and storage in the target area, and G p Indicates the accumulated gas volume, T sc Indicates the temperature under standard conditions, p scIndicates the pressure under standard conditions, W p Indicates the third water production, B w Represents the volume coefficient of formation water.

[0120] The embodiments of the present invention are described in detail above with reference to the accompanying drawings, but the present invention is not limited thereto, and various changes can be made within the knowledge scope of technicians in the relevant technical field without departing from the purpose of the present invention.

Claims

1. A method for calculating regional water production based on the InVEST model, characterized in that: include: Acquiring a surface remote sensing image of a target area at a preset sampling period to obtain a surface remote sensing image sequence of the target area, and performing a preprocessing operation on the surface remote sensing image sequence of the target area to obtain a standard surface remote sensing image sequence of the target area; Performing image segmentation on the standard surface remote sensing image sequence of the target area to obtain a segmented image sequence set corresponding to the standard surface remote sensing image sequence; Counting the pixel areas of each segmented image sequence in the segmented image sequence set to obtain a segmented image pixel area sequence set, and calculating the area ratio of the segmented images based on the segmented image pixel area sequence set to obtain a segmented image area ratio sequence set; Inputting the segmented image area ratio sequence set into a trained water production prediction model, and outputting a first water production corresponding to the target area through the trained water production prediction model; The water production of the target area is calculated by using the InVEST model based on the hydrological and meteorological variables of the target area to obtain a second water production corresponding to the target area, a ground-water material balance model is constructed based on formation parameters, and the water production of the target area is simulated based on the ground-water material balance model to obtain a third water production corresponding to the target area; The first water production, the second water production and the third water production corresponding to the target area are weighted on average based on a preset weight coefficient to obtain the water production corresponding to the target area.

2. The method for calculating regional water production based on the InVEST model according to claim 1, characterized in that: The segmented image sequence set includes a grassland remote sensing image sequence, a cultivated land remote sensing image sequence, a forest remote sensing image sequence, a bare land remote sensing image sequence, a town remote sensing image sequence and a water body remote sensing image sequence.

3. The method for calculating regional water production based on the InVEST model according to claim 2, characterized in that: The surface remote sensing image sequence of the target area is preprocessed to obtain a standard surface remote sensing image sequence of the target area, including: Binarizing the surface remote sensing image sequence to obtain a surface remote sensing binary image sequence; Traversing each pixel point in the surface remote sensing binary image, and when traversing to a current pixel point, judging whether the current pixel point is an isolated point by using a preset isolated point judgment rule; If the current pixel is an isolated point, the current pixel is deleted; otherwise, the next pixel of the current pixel is traversed until each pixel in the surface remote sensing binary image is traversed to obtain a standard surface remote sensing image.

4. The method for calculating regional water production based on the InVEST model according to claim 3 is characterized in that: The isolated point judgment rule includes a first isolated point judgment formula, a second isolated point judgment formula, a third isolated point judgment formula and a fourth isolated point judgment formula, and the first isolated point judgment formula is as follows: B(P0)=1; Wherein, P0 represents the current pixel point, and B(P0) represents the pixel value of the current pixel point; The second isolated point judgment formula is as follows: Where n represents the serial number of the pixel adjacent to the current pixel, N4 represents the first digital sequence, and N4 = {1, 2, 3, 4}, a 2n-1 Indicates whether the pixel value of the adjacent pixel numbered 2n-1 is 0, and The third isolated point judgment formula is as follows: Wherein, N8 represents a second digital sequence, and N8={1,2,...,8}; The fourth isolated point determination formula is as follows: Among them, C n Indicates whether the pixel value of the adjacent pixel with sequence number n is 1, and 5. The method for calculating regional water production based on the InVEST model according to claim 4, characterized in that: Performing image segmentation on the standard surface remote sensing image sequence of the target area to obtain a segmented image sequence set corresponding to the standard surface remote sensing image sequence, including: A plurality of initial segmentation thresholds are preset, and grayscale mapping is performed on the standard surface remote sensing image based on the plurality of initial segmentation thresholds, and the grayscale mapping formula is as follows: Among them, f(x,y) represents the gray value of the standard surface remote sensing image, l0,l1,...,l n-1 ,l n Represents the grayscale of the standard surface remote sensing image, t1, t2, ..., t n-1 ,t n represents multiple preset initial segmentation thresholds, and L represents the total number of gray levels of the standard surface remote sensing image; An objective function is constructed based on the multiple initial segmentation thresholds and the grayscale mapping, and the objective function is as follows: Among them, H(t1,t2,...,t n-1 ,t n ) represents the objective function of the standard surface remote sensing image, p j Represents the probability that the gray value of the standard surface remote sensing image is j.

6. The method for calculating regional water production based on the InVEST model according to claim 5, characterized in that: Performing image segmentation on the standard surface remote sensing image sequence of the target area to obtain a segmented image sequence set corresponding to the standard surface remote sensing image sequence, further comprising: Solving the objective function based on an optimization algorithm to obtain an optimal segmentation threshold combination to obtain multiple optimal segmentation thresholds; The standard surface remote sensing image sequence of the target area is segmented based on the multiple optimal segmentation thresholds to obtain a segmented image sequence set corresponding to the standard surface remote sensing image sequence.

7. The method for calculating regional water production based on the InVEST model according to claim 6, characterized in that: Solving the objective function based on an optimization algorithm to obtain an optimal segmentation threshold combination includes: Set the initialization parameters of the optimization algorithm to initialize the population position. The population position initialization formula is as follows: Among them, x t and x t+1 represents the position of the individual at the t-th iteration and the t+1-th iteration, and p represents the preset parameter for controlling the nonlinear mapping; Calculate the fitness value of the individual and calculate the elite individual position. The elite individual position calculation formula is as follows: in, represents the reverse solution of the j-th dimension individual at the t-th iteration, x i,j (t) represents the elite solution of the j-th dimension individual at the t-th iteration, k represents a random number in the interval [0,1], u j (t) represents the maximum position of the individual at the tth iteration, l j (t) represents the minimum position of the individual at the tth iteration; Marking a preset number of the individuals with the best fitness values ​​as leader individuals, and updating the positions of the remaining individuals in the population; Calculate and sort the updated individual fitness values, obtain elite individuals, and update the elite individual positions; Repeat the above steps until the optimal solution is found or the maximum number of iterations of the optimization algorithm is reached to obtain the optimal segmentation threshold combination.

8. The method for calculating regional water production based on the InVEST model according to claim 1, characterized in that: The water yield prediction model adopts a BP neural network, and the water yield prediction model includes an input layer, a hidden layer and an output layer. The input layer is used to input an area ratio sequence of grassland remote sensing images, cultivated land remote sensing images, forest remote sensing images, bare land remote sensing images, urban remote sensing images and water body remote sensing images; the hidden layer is used to learn the nonlinear relationship between the area ratio sequence of grassland remote sensing images, cultivated land remote sensing images, forest remote sensing images, bare land remote sensing images, urban remote sensing images and water body remote sensing images and the water yield; the output layer is used to quantify and output the learned nonlinear relationship between the area ratio sequence of grassland remote sensing images, cultivated land remote sensing images, forest remote sensing images, bare land remote sensing images, urban remote sensing images and water body remote sensing images and the water yield, so as to obtain a first water yield corresponding to the target area.

9. The method for calculating regional water production based on the InVEST model according to claim 1, characterized in that: The water yield of the target area is calculated by using the InVEST model based on the hydrological and meteorological variables of the target area to obtain a second water yield corresponding to the target area, including: A water production model is constructed based on the InVEST model. The water production model is as follows: Among them, Y(x) represents water yield, Kc(x) represents vegetation evapotranspiration coefficient, ETO(x) represents reference evapotranspiration, P(x) represents precipitation, and W represents non-physical parameters; The water production of the target area is calculated through a water production model based on the hydrological and meteorological variables of the target area to obtain a second water production corresponding to the target area.

10. The method for calculating regional water production based on the InVEST model according to claim 1, characterized in that: The groundwater material balance model is as follows: Among them, p R represents the gas layer pressure, Z represents the gas deviation coefficient under the gas layer pressure, p i represents the original formation pressure, Z i represents the gas deviation coefficient under the original formation pressure, G represents the total cumulative gas production and storage in the target area, and G p Indicates the accumulated gas volume, T sc Indicates the temperature under standard conditions, p sc Indicates the pressure under standard conditions, W p Indicates the third water production, B w Represents the volume coefficient of formation water.

Images