A method, apparatus, and system for time-varying fusion of multi-source precipitation information considering error variability.

Abstract

This invention discloses a time-varying fusion method, apparatus, and system for multi-source precipitation information considering error variations. The method includes: generating a first satellite-retrieved precipitation product, a second satellite-retrieved precipitation product, and a reanalysis-simulated precipitation product with the same spatiotemporal resolution; determining the time-varying optimal weights of the first satellite-retrieved precipitation product, the second satellite-retrieved precipitation product, and the reanalysis-simulated precipitation product at each time point of each rain gauge, with the objective of minimizing the error between the fused precipitation product and the rain gauge observation data; interpolating the time-varying optimal weights using an inverse distance weighting method to generate time-varying fusion weights; and generating a fused precipitation product based on the first satellite-retrieved precipitation product, the second satellite-retrieved precipitation product, the reanalysis-simulated precipitation product, and the time-varying fusion weights. This invention can effectively fuse multi-source precipitation information in a time-varying manner, providing data support for water resources management and flood disaster monitoring and early warning in data-scarce watersheds.

Description

Technical Field

[0001] This invention belongs to the technical field of multi-source precipitation information fusion, specifically relating to a time-varying fusion method, device, and system for multi-source precipitation information that considers error variations. Background Technology

[0002] High-quality precipitation data is a crucial foundation for global water cycle simulation, regional climate monitoring, watershed flood forecasting and early warning, and water conservancy project planning and design. Traditional precipitation data mainly relies on observations from surface rain gauges. However, rain gauge networks are typically sparsely and unevenly distributed, making it difficult to reliably characterize the spatiotemporal variations of actual precipitation and failing to meet the high-precision hydrological simulation requirements for engineering applications.

[0003] With the rapid development of satellite remote sensing technology and data inversion algorithms, precipitation products based on quantitative estimation using satellite remote sensing have advantages such as wide coverage, strong spatial continuity, and high spatiotemporal resolution. They effectively overcome the spatial limitations of ground-based rain gauges and solve the precipitation observation problem in areas without rain gauges, providing potential precipitation data references for hydrological model simulation and drought / flood monitoring and early warning in data-scarce areas. Meanwhile, based on observational data from ground-based rain gauges, weather radars, UAVs, and satellites, researchers have used numerical models of land surface and climate processes, analysis and prediction systems, and data assimilation techniques to generate spatiotemporally continuous reanalysis products. These products can provide global precipitation data with high spatiotemporal resolution and long time spans.

[0004] However, satellite precipitation products and reanalysis precipitation products are subject to significant uncertainties due to factors such as observation, cloud characteristics, inversion algorithms, assimilation schemes, and model parameterization processes, especially in regions with complex topography and climate. Most scholars have found that multi-source data fusion can effectively improve the reliability of precipitation estimates. However, existing studies generally neglect the true variability of precipitation at small timescales and rarely consider the time-varying characteristics of fusion weights, failing to efficiently integrate the advantages of various precipitation information sources. This hinders watershed water resource planning and management, as well as the joint operation and scheduling of reservoir groups. Summary of the Invention

[0005] To address the aforementioned issues, this invention proposes a time-varying fusion method, apparatus, and system for multi-source precipitation information that considers error variations. This method can effectively fuse multi-source precipitation information over time, providing more accurate and operable data support for water resources management and flood disaster monitoring and early warning in data-scarce watersheds.

[0006] To achieve the above-mentioned technical objectives and effects, the present invention is implemented through the following technical solution:

[0007] In a first aspect, the present invention provides a time-varying fusion method for multi-source precipitation information that considers error variations, comprising:

[0008] Generate a first satellite-retrieved precipitation product, a second satellite-retrieved precipitation product, and a reanalysis-simulated precipitation product with the same spatiotemporal resolution;

[0009] At each time point of each rain gauge station, with the goal of minimizing the error between the fused precipitation product and the rain gauge station observation data, the time-varying optimal weights of the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are determined.

[0010] The time-varying optimal weights are interpolated using the inverse distance weighting method to generate time-varying fusion weights, which have the same spatiotemporal resolution as the three precipitation products.

[0011] Based on the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product, as well as the time-varying fusion weights, a fused precipitation product is generated.

[0012] In conjunction with the first aspect, optionally, the method for determining the time-varying optimal weights of the first satellite-retrieved precipitation product, the second satellite-retrieved precipitation product, and the reanalysis-simulated precipitation product includes:

[0013] For each time point of each rain gauge station, three weights w1, w2, and w3, each with a value range of 0 to 1 and a sum of 1, are assigned. At each time point, the precipitation values ​​x1, x2, and x3 from the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are linearly fused into a unique fused precipitation value m. The formula for calculating the fused precipitation value m is as follows: ;

[0014] If the precipitation values ​​of the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are all zero, then w1, w2, and w3 are all taken as 1 / 3;

[0015] If the precipitation values ​​of the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are all lower or higher than the observed precipitation values ​​of the rain gauge station, then the precipitation value of the precipitation product that is closest to the observed precipitation value of the rain gauge station is set to 1, while the precipitation values ​​of the other two precipitation products are set to 0.

[0016] If the observed precipitation value from the rain gauge falls between any two of the precipitation values ​​for all product categories, the weight of the corresponding two precipitation product values ​​is calculated using the following formula; the weight of the remaining precipitation product values ​​is 0:

[0017] ;

[0018] ;

[0019] in, The observed precipitation values ​​from the rain gauge station;

[0020] Based on the three weights at each time point of each rain gauge station, the time-varying optimal weights are generated for the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product.

[0021] In conjunction with the first aspect, optionally, the method for generating the time-varying fusion weights includes:

[0022] At each time point, the Python tools in ArcGIS software were used to programmatically process the time-varying optimal weights of the three precipitation products from each rain gauge station in batches, and interpolated them into grid weights for a specific raster cell size. , , An index representing the weights; during batch processing, the Inverse Distance Weighted Interpolation Toolbox in ArcGIS software is invoked;

[0023] Grid weights are processed using a normalization method. Obtain the time-varying fusion weights of the grid. .

[0024] In conjunction with the first aspect, optionally, the time-varying fusion weights The calculation formula is:

[0025] .

[0026] In conjunction with the first aspect, optionally, the integrated precipitation product can be calculated using the following formula:

[0027]

[0028] Among them, when hour, This indicates the precipitation product retrieved by the first satellite; when hour, This indicates that the second satellite retrieved precipitation products, when hour, This indicates a reanalysis of simulated precipitation products.

[0029] Secondly, the present invention provides a time-varying fusion device for multi-source precipitation information considering error variations, comprising:

[0030] The precipitation product processing module is used to generate a first satellite-retrieved precipitation product, a second satellite-retrieved precipitation product, and a reanalysis simulated precipitation product with the same spatiotemporal resolution;

[0031] The time-varying optimal weight generation module is used to determine the time-varying optimal weights of the first satellite-inverted precipitation product, the second satellite-inverted precipitation product, and the reanalysis simulated precipitation product at each time point of each rain gauge station, with the goal of minimizing the error between the fused precipitation product and the observation data of the rain gauge station.

[0032] The time-varying fusion weight generation module is used to interpolate the time-varying optimal weights using the inverse distance weighting method to generate time-varying fusion weights, which have the same spatiotemporal resolution as the three precipitation products.

[0033] The fused precipitation product generation module is used to generate a fused precipitation product based on the first satellite inverted precipitation product, the second satellite inverted precipitation product, the reanalysis simulated precipitation product, and the time-varying fusion weight.

[0034] In conjunction with the second aspect, optionally, the method for determining the time-varying optimal weights of the first satellite-retrieved precipitation product, the second satellite-retrieved precipitation product, and the reanalysis-simulated precipitation product includes:

[0035] For each time point of each rain gauge station, three weights w1, w2, and w3, each with a value range of 0 to 1 and a sum of 1, are assigned. At each time point, the precipitation values ​​x1, x2, and x3 from the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are linearly fused into a unique fused precipitation value m. The formula for calculating the fused precipitation value m is as follows: ;

[0036] If the precipitation values ​​of the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are all zero, then w1, w2, and w3 are all taken as 1 / 3;

[0037] If the precipitation values ​​of the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are all lower or higher than the observed precipitation values ​​of the rain gauge station, then the precipitation value of the precipitation product that is closest to the observed precipitation value of the rain gauge station is set to 1, while the precipitation values ​​of the other two precipitation products are set to 0.

[0038] If the observed precipitation value from the rain gauge falls between any two of the precipitation values ​​for all product categories, the weight of the corresponding two precipitation product values ​​is calculated using the following formula; the weight of the remaining precipitation product values ​​is 0:

[0039] ;

[0040] ;

[0041] in, The observed precipitation values ​​from the rain gauge station;

[0042] Based on the three weights at each time point of each rain gauge station, the time-varying optimal weights are generated for the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product.

[0043] In conjunction with the second aspect, optionally, the method for generating the time-varying fusion weights includes:

[0044] At each time point, the Python tools in ArcGIS software were used to programmatically process the time-varying optimal weights of the three precipitation products from each rain gauge station in batches, and interpolated them into grid weights for a specific raster cell size. , , An index representing the weights; during batch processing, the Inverse Distance Weighted Interpolation Toolbox in ArcGIS software is invoked;

[0045] Grid weights are processed using a normalization method. Obtain the time-varying fusion weights of the grid. .

[0046] In conjunction with the second aspect, optionally, the time-varying fusion weights The calculation formula is:

[0047] ;

[0048] The integrated precipitation product is calculated using the following formula:

[0049]

[0050] Among them, when hour, This indicates the precipitation product retrieved by the first satellite; when hour, This indicates that the second satellite retrieved precipitation products, when hour, This indicates a reanalysis of simulated precipitation products.

[0051] Thirdly, the present invention provides a time-varying fusion system for multi-source precipitation information that considers error variations, including a storage medium and a processor;

[0052] The storage medium is used to store instructions;

[0053] The processor is configured to operate according to the instructions to perform the method according to any one of the first aspects.

[0054] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0055] This invention fully utilizes the advantages of satellite inversion, reanalysis simulation, and ground rain gauge observations to overcome the shortcomings of low performance of satellite inversion precipitation products and reanalysis simulation precipitation products, as well as weak spatial representativeness of ground rain gauges. Through multi-source data time-varying fusion technology, high-quality precipitation products are obtained. These precipitation products provide more accurate and operable data support for water resources management and flood disaster monitoring and early warning in data-free watersheds.

[0056] This invention takes into account the diurnal variability of precipitation and the time-varying characteristics of the fusion weights, improving several performance aspects of precipitation estimation, including correlation, systematic bias, absolute difference, and rainfall detection. This is beneficial for water resources planning and management in watersheds without data and for the joint operation and scheduling of reservoir groups. Attached Figure Description

[0057] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly described below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort, wherein:

[0058] Figure 1 This is a flowchart illustrating the time-varying fusion method for multi-source precipitation information according to an embodiment of the present invention.

[0059] Figure 2 This refers to the feasible regions of weights w1 and w2 in the fusion method of this embodiment of the invention, as well as the possible optimal values ​​of fused precipitation. Detailed Implementation

[0060] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0061] Furthermore, if the embodiments of this invention involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0062] Example 1

[0063] This invention provides a time-varying fusion method for multi-source precipitation information that considers error variations, comprising the following steps:

[0064] (1) The original first satellite inverted precipitation products, the original second satellite inverted precipitation products and the original reanalysis simulated precipitation products are interpolated to generate the first satellite inverted precipitation products, the second satellite inverted precipitation products and the reanalysis simulated precipitation products with the same spatiotemporal resolution;

[0065] (2) At each time point of each rain gauge station, with the goal of minimizing the error between the fused precipitation product and the rain gauge station observation data, determine the time-varying optimal weights of the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product;

[0066] (3) The time-varying optimal weights are interpolated using the inverse distance weighting method to generate time-varying fusion weights, which have the same spatiotemporal resolution as the three precipitation products;

[0067] (4) Based on the first satellite inverted precipitation product, the second satellite inverted precipitation product and the reanalysis simulated precipitation product, and the time-varying fusion weight, generate a fused precipitation product.

[0068] In one specific embodiment of the present invention, the method for determining the time-varying optimal weights of the first satellite-retrieved precipitation product, the second satellite-retrieved precipitation product, and the reanalysis simulated precipitation product includes:

[0069] For each time point of each rain gauge station, three weights w1, w2, and w3, each with a value range of 0 to 1 and a sum of 1, are assigned. At each time point, the precipitation values ​​x1, x2, and x3 from the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are linearly fused into a unique fused precipitation value m. The formula for calculating the fused precipitation value m is as follows: ;

[0070] If the precipitation values ​​of the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are all zero, then w1, w2, and w3 are all taken as 1 / 3;

[0071] If the precipitation values ​​of the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are all lower or higher than the observed precipitation values ​​of the rain gauge station, then the precipitation value of the product closest to the observed precipitation value of the rain gauge station will be weighted at 1, while the precipitation values ​​of the other two products will be weighted at 0.

[0072] If the observed precipitation value from the rain gauge falls between any two of the precipitation values ​​for all products, the weight of the corresponding two product precipitation values ​​is calculated using the following formula; the weight of the remaining product precipitation values ​​is 0:

[0073] ;

[0074] ;

[0075] in, The observed precipitation values ​​from the rain gauge station;

[0076] Based on the three weights at each time point of each rain gauge station, the time-varying optimal weights are generated for the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product.

[0077] In one specific embodiment of the present invention, the method for generating the second time-varying optimal weight matrix includes:

[0078] At each time point, the Python tools in ArcGIS software were used to programmatically process the time-varying optimal weights of the three precipitation products from each rain gauge station in batches, and interpolated them into grid weights for a specific raster cell size. , , An index representing the weights; during batch processing, the Inverse Distance Weighted Interpolation Toolbox in ArcGIS software is invoked;

[0079] Grid weights are processed using a normalization method. Obtain the time-varying fusion weights of the grid. .

[0080] In one specific embodiment of the present invention, the time-varying fusion weight The calculation formula is:

[0081] .

[0082] In one specific embodiment of the present invention, the fused precipitation product is calculated using the following formula:

[0083]

[0084] Among them, when hour, This indicates the precipitation product retrieved by the first satellite; when hour, This indicates that the second satellite retrieved precipitation products, when hour, This indicates a reanalysis of simulated precipitation products.

[0085] The following detailed description of the time-varying fusion method for multi-source precipitation information considering error variations in this invention is based on a specific implementation method.

[0086] refer to Figure 1 This invention provides a time-varying fusion method for multi-source precipitation information considering error variations. The method first collects limited rain gauge observation data, satellite-retrieved precipitation products, and reanalysis-simulated precipitation products from data-scarce areas, and interpolates these three data points to achieve the same suitable spatiotemporal resolution. Second, at each time point of the rain gauge, a linear combination method is applied to optimize the weights of all time points to obtain the optimal time-varying weights, aiming to minimize the error between the fused precipitation value and the station observation data. Subsequently, an inverse distance weighting method is used to interpolate the optimal time-varying weights into grid cells of appropriate size, and normalization is performed to obtain the time-varying fusion weights of the grid. Finally, at the same spatiotemporal resolution, the time-varying fusion weights are linearly weighted with the three precipitation products to generate a higher-precision fused precipitation product.

[0087] The following examples, in conjunction with the appendix, illustrate the concepts. Figure 1 The technical solution of the present invention will be further described in detail below:

[0088] Step 1: Collect limited rain gauge observation data, satellite-retrieved precipitation products, and reanalysis-simulated precipitation products from areas lacking data;

[0089] This embodiment uses two satellite precipitation products, GPM IMERG Final and PERSIANN-CDR, as well as the ERA5 reanalysis precipitation product. Due to the spatiotemporal scale mismatch among these three precipitation products, bilinear interpolation is used to ensure they have the same suitable spatiotemporal resolution for easier calculations. Specifically, this embodiment uses a spatial resolution of 0.25° and a daily timescale.

[0090] Step two involves applying a linear combination method at each time point of the rain gauge to optimize the weights for all time points, aiming to minimize the error between the fused precipitation product and the rain gauge observation data, thus obtaining the time-varying optimal weights. This step is detailed below:

[0091] Without any assumptions, each rain gauge is assigned three weights (w1, w2, w3) with values ​​ranging from 0 to 1 and summing to 1. At each time step, the precipitation values ​​(x1, x2, x3) of the three precipitation products are linearly merged into a unique merged precipitation value (m):

[0092] (1)

[0093] The absolute error (d) between the fused precipitation value and the precipitation value (r) from the rain gauge is calculated using a function. This indicates that the goal of the function is to minimize the absolute error by optimizing the weights, thus obtaining the optimal fused precipitation value. There are typically three scenarios when estimating reference precipitation data for multi-source precipitation products at each time step: all product precipitation values ​​underestimate the reference precipitation value, meaning x1, x2, and x3 are all less than r; all product precipitation values ​​overestimate the reference precipitation value, meaning any one of x1, x2, and x3 is greater than r; and there is both underestimation and overestimation of the reference precipitation value by the three product precipitation values, meaning the value of r lies between the minimum and maximum values ​​of x1, x2, and x3.

[0094] To find the optimal weights in these three cases, it is necessary to determine the numerical range of the merged precipitation values. Based on the constraint of the sum of weights, any weight can be represented by the other two weights, such as... Then equation (1) can be updated to:

[0095] (2)

[0096] By combining the graphical method of linear programming, the feasible region of weights w1 and w2 can be deduced, such as... Figure 2 As shown in the figure, the feasible region of these two weights is a triangular convex set, meaning that the three vertices of the triangle are the optimal solutions to equation (2). Regardless of the direction and magnitude of the linear slope in equation (2), the possible optimal values ​​for the merged precipitation are x1 (w1 is 1 and w2 is 0), x2 (w1 is 0 and w2 is 1), and x3 (w1 and w2 are both 0). In short, the range of the merged precipitation values ​​is from the minimum to the maximum of x1, x2, and x3. Therefore, when all product precipitation values ​​are lower or higher than the reference precipitation value, the optimal weight for the product precipitation value closest to the reference precipitation value is 1, while the optimal weight for other product precipitation values ​​is 0.

[0097] When the reference precipitation value falls within the range of the merged precipitation value, the absolute error value can reach the ideal value, i.e., d = 0. However, equation (2) has infinite solutions for weights in this case. The merged precipitation value can be calculated using two product precipitation values ​​that are closer to the reference precipitation value, but the reference precipitation value must be between these two product precipitation values ​​(e.g., x1 and x2). The weights of the remaining product precipitation values ​​(e.g., x3) are set to 0 to remove redundant precipitation information. Therefore, when the merged precipitation value equals the reference precipitation value, the corresponding absolute error value is 0, and equation (2) can be simplified to:

[0098] (3)

[0099] Combining the constraints of equation (3) The expression for the optimal weight can be written as:

[0100] , (4)

[0101] In summary, the simplified process of obtaining the optimal weight at each time point is as follows: if the precipitation values ​​of all original precipitation products are zero, then their weights are all 1 / 3; if the precipitation values ​​of all products are lower or higher than the reference precipitation value, then the precipitation value of the product closest to the reference precipitation value has a weight of 1, while the weights of other product precipitation values ​​are 0; if the reference precipitation value is between any two of the product precipitation values, then the weights of the corresponding two product precipitation values ​​are calculated using equation (4), and the weights of the remaining product precipitation values ​​are 0. Thus, the time-varying optimal weights of the rain gauge station can be obtained.

[0102] Step 3: Use the inverse distance weighting method to interpolate the time-varying optimal weights into grid cells of appropriate size (turning point data into grid data), and at the same time, obtain the time-varying fusion weights of the grid through normalization processing;

[0103] At each time point, the Python tools in ArcGIS software are used to programmatically process the time-varying optimal weights of the stations in batches, interpolating them into grid weights for a specific raster cell size. During batch processing, the inverse distance weighted interpolation toolbox in ArcGIS software is invoked. Then, a normalization method is used to process the grid weights to obtain the time-varying fused weights of the grids. :

[0104] (5)

[0105] Step 4: Under the same spatiotemporal resolution, the time-varying fusion weights are linearly weighted with the original precipitation product to generate a higher-precision fused precipitation product.

[0106] Based on three precipitation products The time-varying fusion weights of the grid are used to generate a fused precipitation product with spatiotemporal resolution through weighted averaging. (0.25° / 1 day):

[0107] (6)

[0108] Among them, when hour, This indicates the precipitation product retrieved by the first satellite; when hour, This indicates that the second satellite retrieved precipitation products, when hour, This indicates a reanalysis of simulated precipitation products.

[0109] Example 2

[0110] Based on the same inventive concept as in Embodiment 1, this embodiment of the invention provides a time-varying fusion device for multi-source precipitation information that considers error variations, comprising:

[0111] The precipitation product processing module is used to generate a first satellite-retrieved precipitation product, a second satellite-retrieved precipitation product, and a reanalysis simulated precipitation product with the same spatiotemporal resolution;

[0112] The time-varying optimal weight generation module is used to determine the time-varying optimal weights of the first satellite-inverted precipitation product, the second satellite-inverted precipitation product, and the reanalysis simulated precipitation product at each time point of each rain gauge station, with the goal of minimizing the error between the fused precipitation product and the observation data of the rain gauge station.

[0113] The time-varying fusion weight generation module is used to interpolate the time-varying optimal weights using the inverse distance weighting method to generate time-varying fusion weights, which have the same spatiotemporal resolution as the three precipitation products.

[0114] The fused precipitation product generation module is used to generate a fused precipitation product based on the first satellite inverted precipitation product, the second satellite inverted precipitation product, the reanalysis simulated precipitation product, and the time-varying fusion weight.

[0115] In one specific embodiment of the present invention, the method for determining the time-varying optimal weights of the first satellite-retrieved precipitation product, the second satellite-retrieved precipitation product, and the reanalysis simulated precipitation product includes:

[0116] For each time point of each rain gauge station, three weights w1, w2, and w3, each with a value range of 0 to 1 and a sum of 1, are assigned. At each time point, the precipitation values ​​x1, x2, and x3 from the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are linearly fused into a unique fused precipitation value m. The formula for calculating the fused precipitation value m is as follows: ;

[0117] If the precipitation values ​​of the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are all zero, then w1, w2, and w3 are all taken as 1 / 3;

[0118] If the precipitation values ​​of the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are all lower or higher than the observed precipitation values ​​of the rain gauge station, then the precipitation value of the precipitation product that is closest to the observed precipitation value of the rain gauge station is set to 1, while the precipitation values ​​of the other two precipitation products are set to 0.

[0119] If the observed precipitation value from the rain gauge falls between any two of the precipitation values ​​for all product categories, the weight of the corresponding two precipitation product values ​​is calculated using the following formula; the weight of the remaining precipitation product values ​​is 0:

[0120] ;

[0121] ;

[0122] in, The observed precipitation values ​​from the rain gauge station;

[0123] Based on the three weights at each time point of each rain gauge station, the time-varying optimal weights are generated for the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product.

[0124] In one specific embodiment of the present invention, the method for generating the time-varying fusion weights includes:

[0125] At each time point, the Python tools in ArcGIS software were used to programmatically process the time-varying optimal weights of the three precipitation products from each rain gauge station in batches, and interpolated them into grid weights for a specific raster cell size. , , An index representing the weights; during batch processing, the Inverse Distance Weighted Interpolation Toolbox in ArcGIS software is invoked;

[0126] Grid weights are processed using a normalization method. Obtain the time-varying fusion weights of the grid. .

[0127] In one specific embodiment of the present invention, the time-varying fusion weight The calculation formula is:

[0128] ;

[0129] The integrated precipitation product is calculated using the following formula:

[0130]

[0131] Among them, when hour, This indicates the precipitation product retrieved by the first satellite; when hour, This indicates that the second satellite retrieved precipitation products, when hour, This indicates a reanalysis of simulated precipitation products.

[0132] Example 3

[0133] This invention provides a time-varying fusion system for multi-source precipitation information that considers error variations, including a storage medium and a processor;

[0134] The storage medium is used to store instructions;

[0135] The processor is configured to operate according to the instructions to execute the method according to any one of Embodiment 1.

[0136] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0137] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0138] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0139] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0140] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

[0141] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A time-varying fusion method for multi-source precipitation information considering error variations, characterized in that, include: Generate a first satellite-retrieved precipitation product, a second satellite-retrieved precipitation product, and a reanalysis-simulated precipitation product with the same spatiotemporal resolution; At each time point of each rain gauge station, with the goal of minimizing the error between the fused precipitation product and the rain gauge station observation data, the time-varying optimal weights of the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are determined. The time-varying optimal weights are interpolated using the inverse distance weighting method to generate time-varying fusion weights, which have the same spatiotemporal resolution as the three precipitation products. Based on the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product, as well as the time-varying fusion weight, a fused precipitation product is generated. The method for determining the time-varying optimal weights of the first satellite-retrieved precipitation product, the second satellite-retrieved precipitation product, and the reanalysis simulated precipitation product includes: For each time point of each rain gauge station, three weights w1, w2, and w3, each with a value range of 0 to 1 and a sum of 1, are assigned. At each time point, the precipitation values ​​x1, x2, and x3 from the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are linearly fused into a unique fused precipitation value m. The formula for calculating the fused precipitation value m is as follows: ; If the precipitation values ​​of the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are all zero, then w1, w2, and w3 are all taken as 1 / 3; If the precipitation values ​​of the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are all lower or higher than the observed precipitation values ​​of the rain gauge station, then the precipitation value of the precipitation product that is closest to the observed precipitation value of the rain gauge station is set to 1, while the precipitation values ​​of the other two precipitation products are set to 0. If the observed precipitation value from the rain gauge falls between any two of the precipitation values ​​for all product categories, the weight of the corresponding two precipitation product values ​​is calculated using the following formula; the weight of the remaining precipitation product values ​​is 0: ； ； in, The observed precipitation values ​​from the rain gauge station; Based on the three weights at each time point of each rain gauge station, the time-varying optimal weights are generated for the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product.

2. The time-varying fusion method for multi-source precipitation information considering error variations according to claim 1, characterized in that: The method for generating the time-varying fusion weights includes: At each time point, the Python tools in ArcGIS software were used to programmatically process the time-varying optimal weights of the three precipitation products from each rain gauge station in batches, and interpolated them into grid weights for a specific raster cell size. , , An index representing the weights; during batch processing, the Inverse Distance Weighted Interpolation Toolbox in ArcGIS software is invoked; Grid weights are processed using a normalization method. Obtain the time-varying fusion weights of the grid. .

3. The time-varying fusion method for multi-source precipitation information considering error variations according to claim 2, characterized in that: The time-varying fusion weight The calculation formula is: 。 4. The time-varying fusion method for multi-source precipitation information considering error variations according to claim 3, characterized in that: The integrated precipitation product is calculated using the following formula: Among them, when hour, This indicates the precipitation product retrieved by the first satellite; when hour, This indicates that the second satellite retrieved precipitation products, when hour, This indicates a reanalysis of simulated precipitation products.

5. A time-varying fusion device for multi-source precipitation information considering error variations, characterized in that, include: The precipitation product processing module is used to generate a first satellite-retrieved precipitation product, a second satellite-retrieved precipitation product, and a reanalysis simulated precipitation product with the same spatiotemporal resolution; The time-varying optimal weight generation module is used to determine the time-varying optimal weights of the first satellite-inverted precipitation product, the second satellite-inverted precipitation product, and the reanalysis simulated precipitation product at each time point of each rain gauge station, with the goal of minimizing the error between the fused precipitation product and the observation data of the rain gauge station. The time-varying fusion weight generation module is used to interpolate the time-varying optimal weights using the inverse distance weighting method to generate time-varying fusion weights, which have the same spatiotemporal resolution as the three precipitation products. The fused precipitation product generation module is used to generate a fused precipitation product based on the first satellite inverted precipitation product, the second satellite inverted precipitation product, the reanalysis simulated precipitation product, and the time-varying fusion weight; The method for determining the time-varying optimal weights of the first satellite-retrieved precipitation product, the second satellite-retrieved precipitation product, and the reanalysis simulated precipitation product includes: For each time point of each rain gauge station, three weights w1, w2, and w3, each with a value range of 0 to 1 and a sum of 1, are assigned. At each time point, the precipitation values ​​x1, x2, and x3 from the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are linearly fused into a unique fused precipitation value m. The formula for calculating the fused precipitation value m is as follows: ; If the precipitation values ​​of the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are all zero, then w1, w2, and w3 are all taken as 1 / 3; If the precipitation values ​​of the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product are all lower or higher than the observed precipitation values ​​of the rain gauge station, then the precipitation value of the precipitation product that is closest to the observed precipitation value of the rain gauge station is set to 1, while the precipitation values ​​of the other two precipitation products are set to 0. If the observed precipitation value from the rain gauge falls between any two of the precipitation values ​​for all product categories, the weight of the corresponding two precipitation product values ​​is calculated using the following formula; the weight of the remaining precipitation product values ​​is 0: ； ； in, The observed precipitation values ​​from the rain gauge station; Based on the three weights at each time point of each rain gauge station, the time-varying optimal weights are generated for the first satellite inverted precipitation product, the second satellite inverted precipitation product, and the reanalysis simulated precipitation product.

6. The time-varying fusion device for multi-source precipitation information considering error variations according to claim 5, characterized in that, The method for generating the time-varying fusion weights includes: At each time point, the Python tools in ArcGIS software were used to programmatically process the time-varying optimal weights of the three precipitation products from each rain gauge station in batches, and interpolated them into grid weights for a specific raster cell size. , , An index representing the weights; during batch processing, the Inverse Distance Weighted Interpolation Toolbox in ArcGIS software is invoked; Grid weights are processed using a normalization method. Obtain the time-varying fusion weights of the grid. .

7. The time-varying fusion device for multi-source precipitation information considering error variations according to claim 6, characterized in that, The time-varying fusion weight The calculation formula is: ； The integrated precipitation product is calculated using the following formula: Among them, when hour, This indicates the precipitation product retrieved by the first satellite; when hour, This indicates that the second satellite retrieved precipitation products, when hour, This indicates a reanalysis of simulated precipitation products.

8. A time-varying fusion system for multi-source precipitation information considering error variations, characterized in that, Including storage media and processor; The storage medium is used to store instructions; The processor is configured to operate according to the instructions to perform the method according to any one of claims 1-4.

Images