A precipitation data downscaling method, device, equipment, medium and program product

Abstract

The present application relates to the technical field of data processing, and discloses a precipitation data downscaling method, device, equipment, medium and program product, the method comprising: constructing a weather scale climate trend surface with a second resolution based on first precipitation observation data, calculating dynamic enhancement factors of multiple grids in a target region based on the weather scale climate trend surface, the spatial resolution of the first precipitation observation data being a first resolution, the observation period being a first period, the spatial resolution of the target climate state data being a second resolution, the statistical period being a second period, the second resolution being greater than the first resolution, and the second period corresponding to the first period; performing nonlinear transformation on the target climate state data based on the dynamic enhancement factors of each grid to obtain a spatial continuous weight of the target region; processing the first precipitation observation data using the spatial continuous weight to obtain precipitation field data; and correcting the precipitation field data to obtain target precipitation data.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, specifically to a method, apparatus, equipment, medium, and program product for downscaling precipitation data. Background Technology

[0002] Precipitation is a crucial flux in the water cycle and essential foundational data for fields such as hydrological simulation, water resource management, drought and flood disaster early warning, and ecological environment assessment. Currently, the main methods for acquiring precipitation data include ground-based rain gauge observations, meteorological radar detection, and satellite remote sensing retrieval. Ground-based observations offer the highest accuracy, but due to limitations in terrain and maintenance costs, station distribution is often sparse and uneven, making it difficult to capture the spatial heterogeneity of precipitation. Meteorological radar coverage is limited by terrain obstruction and beam attenuation. Satellite remote sensing precipitation products provide continuous observations with global coverage, greatly compensating for the shortcomings of ground-based observations. However, limited by sensor detection mechanisms and orbital altitude, mainstream satellite precipitation products typically only provide coarse spatial resolution. This coarse-resolution data often struggles to capture the spatial heterogeneity of precipitation when dealing with small- to medium-scale watershed hydrological simulations or local meteorological disaster analyses, especially in mountainous areas with complex terrain. To address this issue, spatial downscaling techniques have emerged. This technique aims to transform coarse-resolution satellite precipitation data into high-resolution precipitation data by establishing correlations between precipitation and high-resolution environmental factors. In related technologies, a correction strategy based on independent coarse grids is often adopted, which ignores the spatial continuous motion characteristics of atmospheric fluids. This results in a significant numerical step at the boundary of the original grid in the downscaled precipitation field, forming a "mosaic effect" that destroys the data texture. Summary of the Invention

[0003] This invention provides a method, apparatus, device, medium, and program product for downscaling precipitation data, in order to solve the problem in related technologies where downscaling of precipitation data using a correction strategy based on an independent coarse grid results in a significant numerical step at the original grid boundary in the downscaled precipitation data.

[0004] In a first aspect, the present invention provides a method for downscaling precipitation data. The method includes: acquiring first precipitation observation data and target climatological data for a target region, wherein the spatial resolution of the first precipitation observation data is a first resolution, the observation period is a first time period, the spatial resolution of the target climatological data is a second resolution, the statistical period is a second time period, the second resolution is greater than the first resolution, and the second time period corresponds to the first time period; constructing a synoptic-scale climate trend surface with a spatial resolution of the second resolution based on the first precipitation observation data; calculating dynamic enhancement factors for multiple grids within the target region based on the synoptic-scale climate trend surface, the dynamic enhancement factors being used to characterize the magnitude of precipitation intensity; performing a nonlinear transformation on the target climatological data based on the dynamic enhancement factors of each grid to obtain spatially continuous weights for the target region; processing the first precipitation observation data using the spatially continuous weights to obtain precipitation field data; and correcting the precipitation field data to obtain the target precipitation data.

[0005] The precipitation data downscaling method provided by this invention combines first precipitation observation data with lower spatial resolution with target climatological data with higher resolution to construct a synoptic-scale climate trend surface corresponding to a second resolution. Based on the synoptic-scale climate trend surface, a dynamic enhancement factor is calculated, and a spatially continuous weight is obtained through nonlinear transformation. This weight is not a discrete correction factor based on independent coarse grid division as in related technologies, but rather a continuously varying parameter generated point-by-point for each fine grid. This accurately matches the continuous motion characteristics of atmospheric fluids, achieving a nonlinear response between precipitation intensity and local climate background. This allows the enhancement or weakening effect of the precipitation field to exhibit a natural and gradual change, avoiding the blocky distribution defects caused by independent coarse grid correction in related technologies. Subsequently, the precipitation field is obtained and corrected by processing the first precipitation observation data with spatially continuous weights, solving the "mosaic effect" problem of downscaling data in related technologies. This ensures that the final output target precipitation data possesses both the reliability of macroscopic precipitation trends and the accuracy of microscopic spatial distribution.

[0006] In one optional implementation, the step of constructing a synoptic-scale climate trend surface with a second spatial resolution based on the first precipitation observation data includes: mapping the first precipitation observation data to a grid at the target scale using a preset interpolation algorithm to obtain second precipitation observation data, wherein the spatial resolution of the second precipitation observation data is the second resolution; and filtering the second precipitation observation data to obtain the synoptic-scale climate trend surface.

[0007] The method provided in this optional implementation first maps low-resolution first precipitation observation data to a second high-resolution grid using a preset high-precision interpolation algorithm, replacing the traditional simple interpolation method that is prone to discrepancies, and avoiding numerical steps at the grid boundary. Then, filtering is used to further smooth out the small fluctuations left by the interpolation, so that the constructed synoptic-scale climate trend surface is more in line with the continuous motion characteristics of atmospheric fluids. At the same time, the high-resolution grid shape also lays the foundation for accurately characterizing the heterogeneity of small and medium-scale precipitation in combination with the local climate background, thus optimizing the spatial continuity and physical rationality of downscaling data from the source.

[0008] In one alternative implementation, the target climatological data is determined by the following steps: obtaining initial climatological data for the target region; filling invalid value regions in the initial climatological data using the nearest neighbor difference method to obtain the target climatological data.

[0009] The method provided in this optional implementation fills in the invalid value region of the initial climatological data using the nearest neighbor interpolation method. This can quickly fill in the missing parts of the data, ensuring the spatial integrity and continuity of the target climatological data. At the same time, nearest neighbor interpolation can preserve the original local climate characteristics of the initial climatological data to the greatest extent, avoiding the introduction of additional biases due to complex interpolation algorithms. This provides reliable climatological background support for subsequent calculation of spatially continuous weights and construction of accurate high-resolution precipitation fields, ensuring that the downscaling process proceeds steadily based on missing baseline data.

[0010] In one optional implementation, the precipitation field data is corrected to obtain target precipitation data, including: determining at least one target grid among the multiple grids based on the precipitation values ​​of multiple grids in the synoptic-scale climate trend surface and a preset trace precipitation threshold, wherein the target grid is the grid whose precipitation value is less than the trace precipitation threshold; correcting the climatological precipitation values ​​of each target grid in the precipitation field data to obtain corrected precipitation field data; and correcting the corrected precipitation field data using a preset mass balance constraint algorithm to obtain the target precipitation data.

[0011] The method provided in this optional implementation first identifies the target grid based on the synoptic-scale climate trend surface and the trace precipitation threshold, achieving precise targeting of the correction operation, avoiding indiscriminate intervention in areas of heavy precipitation, and reducing unnecessary calculation errors. Then, it corrects the climatological precipitation values ​​of the target grid, which can accurately suppress false trace precipitation in arid areas, eliminate the jagged effect of dry-wet boundaries, optimize the spatial continuous distribution characteristics of the precipitation field, and solve the problem that traditional correction methods are prone to destroying data texture. Finally, through a preset mass balance constraint algorithm, it can avoid the deviation of macroscopic precipitation total caused by local correction, ensuring that the target precipitation data not only conforms to the local microclimate background characteristics, but also is consistent with the macroscopic water volume characteristics of the original observation data, providing high-precision and physically reasonable basic data for applications such as small- and medium-scale watershed hydrological simulation and drought and flood disaster early warning.

[0012] In one optional implementation, the climatological precipitation values ​​of each target grid in the precipitation field data are corrected to obtain corrected precipitation field data. This includes: extracting multiple non-zero values ​​from the target climatological data; sorting the multiple non-zero values ​​to obtain a sorting result; determining a precipitation critical value based on the sorting result and a preset ratio; inputting the precipitation critical value, the climatological precipitation values ​​of each target grid in the precipitation field data, and a preset transition width coefficient into a pre-constructed smoothing correction function to solve for the correction coefficient of the corresponding target grid; and using the correction coefficient of each target grid to correct the climatological precipitation values ​​of the corresponding target grid in the precipitation field data to obtain corrected precipitation field data.

[0013] The method provided in this optional implementation determines the precipitation threshold based on the non-zero value sorting and preset ratio of the target climatological data. This accurately adapts to the regional climatological background characteristics, effectively distinguishes between truly arid areas and areas with normal trace precipitation, and avoids false precipitation misjudgments. Furthermore, by combining the precipitation threshold, local climatological values, and transition width coefficient, a smoothing correction function is used to solve for the continuously gradually changing correction coefficient, eliminating the sawtooth effect of the dry-wet boundary and making the precipitation intensity transition more consistent with the actual atmospheric motion patterns. At the same time, the correction is only performed on the target grid, avoiding indiscriminate intervention in areas of heavy precipitation and reducing additional calculation errors.

[0014] In one optional implementation, the modified precipitation field data is corrected using a preset mass balance constraint algorithm to obtain target precipitation data. This includes: aggregating the modified precipitation field data to obtain a simulated precipitation field, where the spatial resolution of the simulated precipitation field is a first resolution; determining the first correction coefficients of each grid in the target area at the first resolution based on the simulated precipitation field and the first precipitation observation data; mapping the first correction coefficients of multiple grids to grids at the target scale using a preset interpolation algorithm to obtain the second correction coefficients of each grid in the target area at a second resolution; and correcting the modified precipitation field data using the second correction coefficients of each grid in the target area at the second resolution to obtain the target precipitation data.

[0015] Secondly, the present invention provides a precipitation data downscaling device, comprising: an acquisition module for acquiring first precipitation observation data and target climatological data of a target area, wherein the spatial resolution of the first precipitation observation data is a first resolution and the observation period is a first time period, the spatial resolution of the target climatological data is a second resolution and the statistical period is a second time period, the second resolution is greater than the first resolution, and the second time period corresponds to the first time period; a construction module for constructing a synoptic-scale climate trend surface with a spatial resolution of the second resolution based on the first precipitation observation data; a calculation module for calculating dynamic enhancement factors of multiple grids within the target area based on the synoptic-scale climate trend surface, wherein the dynamic enhancement factors are used to characterize the magnitude of precipitation intensity; a transformation module for performing a nonlinear transformation on the target climatological data based on the dynamic enhancement factors of each grid to obtain spatially continuous weights for the target area; a processing module for processing the first precipitation observation data using the spatially continuous weights to obtain precipitation field data; and a correction module for correcting the precipitation field data to obtain target precipitation data.

[0016] Thirdly, the present invention provides an electronic device, comprising: a memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the precipitation data downscaling method of the first aspect or any corresponding embodiment described above.

[0017] Fourthly, the present invention provides a computer-readable storage medium storing computer instructions for causing a computer to execute the precipitation data downscaling method of the first aspect or any corresponding embodiment described above.

[0018] Fifthly, the present invention provides a computer program product, including computer instructions for causing a computer to execute the precipitation data downscaling method of the first aspect or any corresponding embodiment described above. Attached Figure Description

[0019] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of an application scenario according to an embodiment of the present invention;

[0021] Figure 2 This is a schematic diagram of the first process of a precipitation data downscaling method according to an embodiment of the present invention;

[0022] Figure 3 This is a schematic diagram of the second process of the precipitation data downscaling method according to an embodiment of the present invention;

[0023] Figure 4 This is a schematic diagram of the third process of the precipitation data downscaling method according to an embodiment of the present invention;

[0024] Figure 5 This is a flowchart illustrating a specific example of the precipitation data downscaling method in the embodiments of this application;

[0025] Figure 6 This is a schematic diagram of coarse-scale daily precipitation observation data;

[0026] Figure 7 This is a schematic diagram of high-resolution precipitation data;

[0027] Figure 8 This is a structural block diagram of a precipitation data downscaling device according to an embodiment of the present invention;

[0028] Figure 9 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, 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 some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0031] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0032] As an optional application scenario of this invention, the specific application environment architecture or specific hardware architecture on which the precipitation data downscaling method depends is described here. For example... Figure 1 As shown, the architecture system may include at least one terminal device and at least one server. Figure 1 The system is illustrated in the example, which includes a computer 101, a mobile terminal 102, and a server 103, and the terminal devices such as the computer 101 and the mobile terminal 102 are connected to the server 103 through a network 110.

[0033] Specifically, the terminal device can be a smartphone, tablet, laptop, PDA, desktop computer, game console, smart TV, smart wearable device, in-vehicle terminal, VR (Virtual Reality) device, AR (Augmented Reality) device, etc. Server 103 can be a standalone physical server, a server cluster, a distributed system, or a cloud server providing cloud services. Network 110 can be a wired or wireless network, examples of which include, but are not limited to, the Internet, corporate intranet, local area network, wide area network, mobile communication network, and combinations thereof.

[0034] Currently, the main methods for acquiring precipitation data include ground-based rain gauge observations, meteorological radar detection, and satellite remote sensing inversion. However, limited by sensor detection mechanisms and orbital altitude, mainstream satellite precipitation products typically only provide coarse spatial resolution. This coarse-resolution data often struggles to capture the spatial heterogeneity of precipitation when dealing with small- to medium-scale watershed hydrological simulations or local meteorological disaster analyses, especially in mountainous areas with complex terrain. To address this issue, spatial downscaling techniques have emerged. This technique aims to transform coarse-resolution satellite precipitation data into high-resolution precipitation data by establishing a correlation between precipitation and high-resolution environmental factors. Many related techniques employ correction strategies based on independent coarse grids, neglecting the continuous spatial motion characteristics of atmospheric fluids. This results in a significant numerical step at the original grid boundaries in the downscaled precipitation field, creating a "mosaic effect" that disrupts the data texture.

[0035] In view of this, this application provides a precipitation data downscaling method that can be applied to a server to achieve downscaling of coarse-scale precipitation data. The method provided in this application combines first precipitation observation data with lower spatial resolution with target climatological data with higher resolution to construct a synoptic-scale climate trend surface corresponding to a second resolution. Based on the synoptic-scale climate trend surface, a dynamic enhancement factor is calculated, and spatially continuous weights are obtained through nonlinear transformation. These weights are not discrete correction factors based on independent coarse grid divisions as in related technologies, but rather continuously varying parameters generated point-by-point for each fine grid. This accurately matches the continuous motion characteristics of atmospheric fluids, achieving a nonlinear response between precipitation intensity and local climate background. This allows the enhancement or weakening effect of the precipitation field to exhibit natural and gradual changes, avoiding the blocky distribution defects caused by independent coarse grid correction in related technologies. Subsequently, the precipitation field is obtained and corrected by processing the first precipitation observation data with spatially continuous weights, solving the "mosaic effect" problem of downscaling data in related technologies. This ensures that the final output target precipitation data possesses both the reliability of macroscopic precipitation trends and the accuracy of microscopic spatial distribution.

[0036] According to an embodiment of the present invention, a method for downscaling precipitation data is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0037] This embodiment provides a precipitation data downscaling method, which can be used in the aforementioned server. Figure 2 This is a flowchart of a precipitation data downscaling method according to an embodiment of the present invention, as shown below. Figure 2 As shown, the process includes the following steps:

[0038] Step S201: Obtain the first precipitation observation data and the target climatological data for the target area. The spatial resolution of the first precipitation observation data is the first resolution, and the observation period is the first time period. The spatial resolution of the target climatological data is the second resolution, and the statistical period is the second time period. The second resolution is greater than the first resolution, and the second time period corresponds to the first time period.

[0039] For example, the first precipitation observation data refers to the actual precipitation observation data with low spatial resolution within the target area. It is the basic data source for downscaling and is usually a precipitation product retrieved from satellite remote sensing or aggregated data from low-density ground stations. Its core function is to provide the macroscopic precipitation situation for the target period. The first resolution refers to the spatial resolution of the first precipitation observation data, which characterizes the fineness of the data grid. The value is usually large (e.g., 10km), meaning that the geographical side of a single data grid is long and the spatial accuracy is low. The first period refers to the specific observation time range of the first precipitation observation data. It is a weather-scale period with a clear time node, which can be a single day, a single month, or the duration of a certain precipitation process. In this embodiment, the first period can be a single day. The target climatological data refers to the long-term climate average data with high spatial resolution within the target area. It is the core data that characterizes the local climate background features and is usually obtained by averaging observation data from the same period over many years. In this embodiment, the target climatological data can be the climatological data of a certain month. The second time period refers to the statistical time range of the target climatological data. It is a "climatological scale period" used to calculate the climatological average and must correspond to the time dimension of the first time period. For example, if the first time period is "June 10, 2026 (single day)", then the second time period must be "June". That is, when it is necessary to downscale the precipitation observation data of a certain day, it is necessary to simultaneously obtain the climatological data of the corresponding month as the target climatological data to ensure that the target climatological data can match the local climate characteristics of the specific observation period.

[0040] Step S202: Construct a synoptic-scale climate trend surface with a spatial resolution of second resolution based on the first precipitation observation data.

[0041] For example, low-resolution first precipitation observation data is transformed into a high-resolution, spatially continuous macroscopic precipitation trend framework, i.e., a synoptic-scale climate trend surface. This application does not limit the specific construction process of the synoptic-scale climate trend surface, as long as it is reasonable.

[0042] Step S203: Calculate the dynamic enhancement factor of multiple grids within the target area based on the synoptic-scale climate trend surface. The dynamic enhancement factor is used to characterize the magnitude of precipitation intensity.

[0043] For example, in the embodiments of this application, the dynamic enhancement factor can be calculated by the following formula:

[0044]

[0045] in, 1 represents the enhancement factor of the i-th grid in the synoptic-scale climate trend surface; 1.0 represents the baseline value, ensuring that when the synoptic-scale precipitation is 0, there is no additional enhancement effect, and the precipitation field directly uses the baseline value of the trend surface or climate state. The sensitivity coefficient is a calibrable empirical parameter used to control the intensity of the enhancement effect; it can be set to 0.1. Let represent the precipitation value of the weather-scale climate trend surface on the i-th grid. This formula makes the precipitation value suitable for heavy rainfall scenarios... The value increases significantly, thus simulating the nonlinear sensitivity of severe convective weather to topographic forced uplift.

[0046] Step S204: Based on the dynamic enhancement factors of each grid, perform a nonlinear transformation on the target climatological data to obtain the spatial continuous weights of the target region.

[0047] For example, a nonlinear weighted transformation is performed on the high-resolution target climatological data using the dynamic enhancement factor of each grid, generating a spatially continuously varying weight field (spatial continuous weight). This allows the weights to dynamically adjust with the macro-precipitation intensity (amplifying the impact of local climatology during heavy precipitation and maintaining basic characteristics during light precipitation) while preserving the spatial continuity of the local climatological background, avoiding the grid boundary step caused by traditional discrete weights. In this embodiment, the target climatological data can be nonlinearly transformed using the following formula:

[0048]

[0049] in, This represents the weight of the i-th fine mesh, where a fine mesh refers to the mesh at the second resolution; This represents the climatological data for the i-th fine grid. This indicates a background-scale Gaussian filtering operation; The smoothing coefficient is used for filtering fine-scale climatological data. The background scaling is performed by exponentiation, yielding the average response against a coarse-scale background, which is then used to normalize the weights. It can be set to 5.0; This indicates that for all fine meshes After performing background scale Gaussian filtering, the first The filtering result corresponding to each fine grid represents the average climatological response of the coarse-scale background to which the grid is located. This represents a regularization term (a small constant to prevent division by zero), avoiding calculation errors when the denominator is 0 and ensuring numerical stability.

[0050] Step S205: Process the first precipitation observation data using spatial continuous weights to obtain precipitation field data.

[0051] For example, in this embodiment of the application, the spatially continuous weights are fused with the first precipitation observation data grid by grid, and the nonlinear enhancement effect reflecting the current synoptic-scale precipitation intensity is combined with the long-term precipitation background characteristics of the local microclimate to generate a high-resolution preliminary precipitation field that has both macro-precipitation trends and micro-spatial heterogeneity. Moreover, because the weights change continuously and gradually, the precipitation field also maintains spatial continuity.

[0052] Step S206: Correct the precipitation field data to obtain the target precipitation data.

[0053] For example, in this embodiment of the application, precipitation field data is corrected according to a preset method to obtain target precipitation data that meets the requirements. This embodiment of the application does not limit the specific content of the preset method, as long as it is reasonable.

[0054] The precipitation data downscaling method provided in this embodiment combines first precipitation observation data with lower spatial resolution with target climatological data with higher resolution to construct a synoptic-scale climate trend surface corresponding to a second resolution. Based on the synoptic-scale climate trend surface, a dynamic enhancement factor is calculated, and a spatially continuous weight is obtained through nonlinear transformation. This weight is not a discrete correction factor based on independent coarse grid division as in related technologies, but rather a continuously varying parameter generated point-by-point for each fine grid. This accurately matches the continuous motion characteristics of atmospheric fluids, achieving a nonlinear response between precipitation intensity and local climate background. This allows the enhancement or weakening effect of the precipitation field to exhibit a natural and gradual change, avoiding the blocky distribution defects caused by independent coarse grid correction in related technologies. Subsequently, the precipitation field is obtained and corrected by processing the first precipitation observation data with spatially continuous weights, solving the "mosaic effect" problem of downscaling data in related technologies. This ensures that the final output target precipitation data possesses both the reliability of macroscopic precipitation trends and the accuracy of microscopic spatial distribution.

[0055] This embodiment provides a precipitation data downscaling method, which can be used in the aforementioned server. Figure 3 This is a flowchart of a precipitation data downscaling method according to an embodiment of the present invention, as shown below. Figure 3 As shown, the process includes the following steps:

[0056] Step S301: Acquire the first precipitation observation data and the target climatological data for the target area. The spatial resolution of the first precipitation observation data is a first resolution, and the observation period is a first time period. The spatial resolution of the target climatological data is a second resolution, and the statistical period is a second time period. The second resolution is greater than the first resolution, and the second time period corresponds to the first time period. For details, please refer to [link to details]. Figure 2 Step S201 of the illustrated embodiment will not be described again here.

[0057] In some alternative implementations, the target climatological data are determined through the following steps:

[0058] Step a1: Obtain initial climatological data for the target area.

[0059] For example, in this application embodiment, the initial climatological data of the target area is first obtained, which is the basis for subsequent processing. However, such data often contains invalid values ​​(missing values) due to observation limitations.

[0060] Step a2: Fill the invalid value region of the initial climatological data using the nearest neighbor difference method to obtain the target climatological data.

[0061] For example, the nearest neighbor interpolation method is used to fill each grid with invalid values ​​with the values ​​of its nearest valid data grid, quickly filling in the data gaps and finally obtaining spatially complete target climate data. This ensures that the subsequent downscaling process can proceed based on complete climate background data, thus preventing edge effects from occurring in subsequent filtering calculations.

[0062] Step S302: Construct a weather-scale climate trend surface with a spatial resolution of second resolution based on the first precipitation observation data.

[0063] Specifically, step S302 includes:

[0064] Step S3021: The first precipitation observation data is mapped to a grid at the target scale using a preset interpolation algorithm to obtain the second precipitation observation data. The spatial resolution of the second precipitation observation data is the second resolution.

[0065] For example, the preset interpolation algorithm may include, but is not limited to, the cubic convolution interpolation algorithm. In this embodiment, the cubic convolution interpolation algorithm is used to interpolate the first precipitation observation data. Resample to a finer-scale grid (e.g., a 1km grid).

[0066] Step S3022: Filter the second precipitation observation data to obtain a synoptic-scale climate trend surface.

[0067] For example, in this embodiment of the application, the second precipitation observation data is processed by Gaussian filtering, and the smoothing coefficient of the synoptic-scale Gaussian filter is... The value is set to 3.0 (corresponding to an effective influence diameter of approximately 18 km). This parameter is chosen based on the Nyquist sampling theorem and the principle of matching physical scale. The aim is to ensure that the smoothed scale after filtering is slightly larger than the physical resolution (10 km) of the original coarse grid, thereby completely eliminating artificial gradients at the grid boundaries and obtaining a synoptic-scale atmospheric trend surface that reflects only the macroscopic weather system (denoted as...). ).

[0068] Step S303: Calculate the dynamic enhancement factors for multiple grids within the target area based on the synoptic-scale climate trend surface. These dynamic enhancement factors characterize the magnitude of precipitation intensity. For details, please refer to [link to relevant documentation]. Figure 2 Step S203 of the illustrated embodiment will not be described again here.

[0069] Step S304: Based on the dynamic enhancement factors of each grid, a nonlinear transformation is performed on the target climatological data to obtain the spatially continuous weights of the target region. For details, please refer to [link to relevant documentation]. Figure 2 Step S204 of the illustrated embodiment will not be described again here.

[0070] Step S305: Process the first precipitation observation data using spatially continuous weighting to obtain precipitation field data. For details, please refer to [link to relevant documentation]. Figure 2 Step S205 of the illustrated embodiment will not be described again here.

[0071] Step S306: Correct the precipitation field data to obtain the target precipitation data. For details, please refer to [link to relevant documentation]. Figure 2 Step S206 of the illustrated embodiment will not be described again here.

[0072] This embodiment provides a precipitation data downscaling method, which can be used in the aforementioned server. Figure 4 This is a flowchart of a precipitation data downscaling method according to an embodiment of the present invention, as shown below. Figure 4 As shown, the process includes the following steps:

[0073] Step S401: Acquire the first precipitation observation data and the target climatological data for the target area. The spatial resolution of the first precipitation observation data is a first resolution, and the observation period is a first time period. The spatial resolution of the target climatological data is a second resolution, and the statistical period is a second time period. The second resolution is greater than the first resolution, and the second time period corresponds to the first time period. For details, please refer to [link to details]. Figure 3 Step S301 of the illustrated embodiment will not be described again here.

[0074] Step S402: Construct a synoptic-scale climate trend surface with a second spatial resolution based on the first precipitation observation data. For details, please refer to [link to relevant documentation]. Figure 3 Step S302 of the illustrated embodiment will not be described again here.

[0075] Step S403: Calculate the dynamic enhancement factors for multiple grids within the target area based on the synoptic-scale climate trend surface. These dynamic enhancement factors characterize the magnitude of precipitation intensity. For details, please refer to [link to relevant documentation]. Figure 3 Step S303 of the illustrated embodiment will not be described again here.

[0076] Step S404: Based on the dynamic enhancement factors of each grid, a nonlinear transformation is performed on the target climatological data to obtain the spatially continuous weights of the target region. For details, please refer to [link to relevant documentation]. Figure 3 Step S304 of the illustrated embodiment will not be described again here.

[0077] Step S405: Process the first precipitation observation data using spatially continuous weights to obtain precipitation field data. For details, please refer to [link to relevant documentation]. Figure 3 Step S305 of the illustrated embodiment will not be described again here.

[0078] Step S406: Correct the precipitation field data to obtain the target precipitation data.

[0079] Specifically, step S406 includes:

[0080] Step S4061: Based on the precipitation values ​​of multiple grids in the weather-scale climate trend surface and the preset trace precipitation threshold, at least one target grid is determined among the multiple grids. The target grid is the grid whose precipitation value is less than the trace precipitation threshold.

[0081] For example, in this embodiment of the social situation, the trace precipitation threshold may include, but is not limited to, 1.0 mm. When the precipitation value of a certain grid in the synoptic-scale climate trend surface is less than this threshold, it is determined that a trace precipitation event has occurred in that grid, and that grid is identified as the target grid.

[0082] Step S4062: Correct the climatological precipitation values ​​of each target grid in the precipitation field data to obtain the corrected precipitation field data.

[0083] For example, the embodiments of this application do not limit the specific modification process, as long as it is reasonable.

[0084] In some optional implementations, step S4062 above includes:

[0085] Step b1: Extract multiple non-zero values ​​from the target climatological data.

[0086] For example, in the embodiments of this application, the specific extraction process of multiple non-zero values ​​can be determined according to actual needs, and the embodiments of this application do not impose specific limitations.

[0087] Step b2: Sort the multiple non-zero values ​​to obtain the sorting result.

[0088] For example, in this embodiment of the application, multiple non-zero values ​​can be sorted in ascending order to obtain a sorting result.

[0089] Step b3: Determine the precipitation threshold based on the sorting results and the preset ratio.

[0090] For example, the precipitation values ​​at the corresponding proportions in the sorted numerical sequence are then found according to a pre-set ratio (e.g., 30%, which is usually determined in conjunction with regional climate characteristics or correction requirements), and these values ​​are used as the precipitation threshold. The purpose of this threshold is to delineate the target grid range that needs correction, providing an objective and accurate basis for determining the correction coefficients when substituted into the smoothing correction function, thus avoiding the problem of mismatch between traditional fixed thresholds and actual climate characteristics.

[0091] Step b4: Based on the precipitation critical value, the climatological precipitation value of each target grid in the precipitation field data, and the preset transition width coefficient, the pre-constructed smoothing correction function is input to solve for the correction coefficient of the corresponding target grid.

[0092] For example, in an embodiment of this application, the smoothing correction function can be as follows:

[0093]

[0094] in, This represents the climatological precipitation value of the i-th grid in the precipitation field data. This represents the correction factor for the i-th grid. Indicates the critical value for precipitation. This represents the transition zone coefficient. The function constructs a continuously and gradually changing probability transition zone near the drought threshold, such that the correction coefficient approaches 0 in arid regions and approaches 1 in humid regions. It can be set to 5.0, meaning it's near the drought threshold. Construct a continuously gradually changing probability transition zone within the range of mm.

[0095] Step b5: Use the correction coefficients of each target grid to correct the climatological precipitation values ​​of the corresponding target grid in the precipitation field data to obtain the corrected precipitation field data.

[0096] For example, in this embodiment of the application, the preliminary precipitation field is multiplied by the correction factor. This allows precipitation to naturally and smoothly decrease at the boundary between dry and wet conditions, eliminating artificially created jagged boundaries.

[0097] Step S4063: Correct the modified precipitation field data using a preset mass balance constraint algorithm to obtain the target precipitation data.

[0098] For example, in this embodiment of the application, a preset mass balance constraint algorithm is used to calibrate the total macroscopic precipitation of the locally corrected precipitation field so that it is consistent with the water volume characteristics of the original low-resolution observation data, and finally obtains target precipitation data with reasonable microscopic spatial texture and macroscopic water volume conservation.

[0099] In some continuing embodiments, step S4063 includes:

[0100] Step c1: Aggregate the corrected precipitation field data to obtain a simulated precipitation field, with the spatial resolution of the simulated precipitation field being the first resolution.

[0101] For example, in this embodiment of the application, the corrected precipitation field data is aggregated and averaged back to the original coarse-scale grid to obtain a simulated coarse-scale precipitation field, i.e., a simulated precipitation field.

[0102] Step c2: Determine the first correction coefficients for each grid in the target area at the first resolution based on the simulated precipitation field and the first precipitation observation data.

[0103] Exemplary, in the embodiments of this application, the original observations are calculated. The ratio of the simulated coarse-scale precipitation field to the first coarse-scale correction coefficient is obtained. .

[0104] Step c3: Using a preset interpolation algorithm, the first correction coefficients of multiple grids are mapped to the grid at the target scale to obtain the second correction coefficients of each grid in the target region at the second resolution.

[0105] For example, to avoid the quadratic mesh boundary effect caused by directly applying this coefficient, cubic convolution interpolation is used to... Mapping back to a fine-scale grid and applying a background-scale Gaussian filter yields a spatially continuous and smooth correction field. Correction field The second correction coefficient is used to characterize each grid in the target region at the second resolution.

[0106] Step c4: Correct the modified precipitation field data using the second correction coefficients of each grid in the target area at the second resolution to obtain the target precipitation data.

[0107] For example, in this embodiment of the application, the corrected precipitation field data is compared with the correction field. Multiply the data to output high-resolution target precipitation data.

[0108] The following specific embodiment illustrates the precipitation data downscaling method provided in this application.

[0109] Example:

[0110] like Figure 5As shown, this application proposes a precipitation data downscaling method based on a dynamic microclimate characteristic response mechanism and a Logistic probabilistic smoothing strategy. This method aims to transform low spatial resolution (coarse-scale) daily precipitation observation data (e.g., ERA5 reanalysis data, with an original resolution of approximately 10 km) into high-resolution (fine-scale) gridded precipitation data (e.g., 1 km resolution). While introducing refined local microclimate characteristics, it maintains the statistical conservation of macroscopic water volume and eliminates spatial artifacts found in traditional methods. The method includes the following steps:

[0111] Step 1: In this embodiment, a synoptic-scale climate trend surface is constructed. First, coarse-scale daily precipitation observation data (denoted as ) are obtained within the study area. For example, ERA5 data), coarse-scale daily precipitation observation data can be like... Figure 6 As shown. To eliminate the inherent "blocking effect" of coarse-grid data and extract the macroscopic atmospheric background field, a cubic convolution interpolation algorithm is used to... Resampling is performed to a target fine-scale grid (e.g., a 1km grid). Subsequently, a synoptic-scale Gaussian filter is applied to the resampled data. In this preferred embodiment, the smoothing coefficient of the synoptic-scale Gaussian filter is... The value is set to 3.0 (corresponding to an effective influence diameter of approximately 18 km). This parameter is chosen based on the Nyquist sampling theorem and the principle of matching physical scale. The aim is to ensure that the smoothed scale after filtering is slightly larger than the physical resolution (10 km) of the original coarse grid, thereby completely eliminating artificial gradients at the grid boundaries and obtaining a synoptic-scale atmospheric trend surface that reflects only the macroscopic weather system (denoted as...). );

[0112] Step 2: Obtain high-resolution historical climatological data for the study area and calculate scale-matched local climate weights (e.g., WorldClim v2.1 monthly average precipitation data from 1970 to 2000, denoted as...). This is used to characterize the local microclimate features formed by the combination of topography, land-sea location, and underlying surface properties. Before calculation, the... Invalid value regions (such as land-sea edges) are filled using nearest neighbor interpolation to prevent edge effects in subsequent filtering calculations. The core of this step is to construct a spatially continuous weight field with physical elasticity. First, based on the trend surface obtained in step 1... Calculate the intensity-dependent dynamic enhancement factor The calculation formula is: Sensitivity coefficient The preferred setting is 0.1. This formula makes it suitable for heavy rain scenarios. The value increases significantly, thus simulating the nonlinear sensitivity of severe convective weather to orographic forcing and uplift. Next, the local climate weights are calculated. Using a background-scale Gaussian filter to process the... The transformed fine-scale climatological data is processed to obtain the background climate field. In this embodiment, the background scale smoothing coefficient... The weight is set to 5.0 to simulate background climate characteristics at a coarse scale of 10 km. Determined by the ratio of fine-scale climatology to background climatology, this weight characterizes the gain factor of local microclimate features on precipitation relative to the coarse-scale background.

[0113] Step 3: Address the common micro-precipitation bias in reanalysis data using intermittent correction based on the Logistic function. First, set a micro-precipitation threshold (e.g., 1.0 mm). When the trend surface... If the precipitation is less than this threshold, it is considered a trace precipitation event. In this case, the drought cutoff threshold is calculated. (For example, taking the 30th percentile of non-zero climatological data within the region for the current month). Unlike traditional hard thresholding, this invention utilizes the Logistic function to construct a smooth correction coefficient. The transition width coefficient of this function. A setting of 5.0 is preferred, meaning it is near the drought threshold. A continuously varying probability transition zone is constructed within a range of mm. The initial precipitation field is multiplied by this correction factor. This allows precipitation to naturally and smoothly decrease at the boundary between dry and wet conditions, eliminating artificially created jagged boundaries.

[0114] Step 4: Smoothing Ratio Mass Balance Correction. To ensure that the downscaled data remains consistent with the original observations in terms of macroscopic water quantity, the corrected precipitation field obtained in Step 3 is first aggregated and averaged back to the original coarse-scale grid to obtain the simulated coarse-scale precipitation field. The original observations are then calculated. The ratio of the simulated coarse-scale precipitation field to the coarse-scale correction coefficient is used to obtain the coarse-scale correction coefficient. To avoid the quadratic mesh boundary effect caused by directly applying this coefficient, cubic convolution interpolation is used to... Mapping back to a fine-scale mesh and applying the same background-scale Gaussian filter as in step 2 yields a spatially continuous and smooth correction field. Finally, the precipitation field obtained in step 3 is compared with the smoothed correction field. Multiply them to output the final high-resolution precipitation data, such as... Figure 7 As shown. This step, at the cost of extremely small local numerical closure deviations, achieves significant accuracy and continuity in the spatial structure of precipitation, realizing a physical mass balance constraint.

[0115] This embodiment also provides a precipitation data downscaling device for implementing the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0116] This embodiment provides a precipitation data downscaling device, such as... Figure 8 As shown, it includes:

[0117] The acquisition module 801 is used to acquire the first precipitation observation data and the target climatological data of the target area. The spatial resolution of the first precipitation observation data is the first resolution, the observation period is the first time period, the spatial resolution of the target climatological data is the second resolution, the statistical period is the second time period, the second resolution is greater than the first resolution, and the second time period corresponds to the first time period.

[0118] Module 802 is used to construct a synoptic-scale climate trend surface with a spatial resolution of the second resolution based on the first precipitation observation data;

[0119] The calculation module 803 is used to calculate the dynamic enhancement factor of multiple grids within the target area based on the synoptic-scale climate trend surface. The dynamic enhancement factor is used to characterize the magnitude of precipitation intensity.

[0120] Transformation module 804 is used to perform nonlinear transformation on the target climatological data based on the dynamic enhancement factors of each grid to obtain the spatial continuous weights of the target region.

[0121] The processing module 805 is used to process the first precipitation observation data using spatial continuous weights to obtain precipitation field data;

[0122] The correction module 806 is used to correct the precipitation field data to obtain the target precipitation data.

[0123] In some alternative implementations, the construction module 802 includes:

[0124] The mapping submodule is used to map the first precipitation observation data to a grid at the target scale using a preset interpolation algorithm to obtain the second precipitation observation data. The spatial resolution of the second precipitation observation data is the second resolution.

[0125] The processing submodule is used to filter the second precipitation observation data to obtain a synoptic-scale climate trend surface.

[0126] In some alternative implementations, the target climatological data are determined through the following steps:

[0127] Acquire initial climatological data for the target area;

[0128] The nearest neighbor difference method is used to fill in invalid value regions in the initial climatological data to obtain the target climatological data.

[0129] In some alternative implementations, the correction module 806 includes:

[0130] The determination submodule is used to determine at least one target grid among multiple grids based on the precipitation values ​​of multiple grids in the weather-scale climate trend surface and a preset trace precipitation threshold. The target grid is the grid whose precipitation value is less than the trace precipitation threshold.

[0131] The correction submodule is used to correct the climatological precipitation values ​​of each target grid in the precipitation field data to obtain the corrected precipitation field data.

[0132] The correction submodule is used to correct the modified precipitation field data using a preset mass balance constraint algorithm to obtain the target precipitation data.

[0133] In some optional implementations, the correction submodule includes:

[0134] The extraction unit is used to extract multiple non-zero values ​​from the target climatological data;

[0135] A sorting unit is used to sort multiple non-zero values ​​to obtain a sorted result;

[0136] The first determining unit is used to determine the precipitation critical value based on the sorting results and a preset ratio;

[0137] The solving unit is used to solve the corresponding target grid's correction coefficient by inputting the precipitation critical value, the climatological precipitation value of each target grid in the precipitation field data, and the preset transition width coefficient into a pre-constructed smoothing correction function.

[0138] The correction unit is used to correct the climatological precipitation values ​​of the corresponding target grid in the precipitation field data using the correction coefficients of each target grid, so as to obtain the corrected precipitation field data.

[0139] In some optional implementations, the correction submodule includes:

[0140] The aggregation unit is used to aggregate the corrected precipitation field data to obtain a simulated precipitation field, the spatial resolution of which is the first resolution;

[0141] The second determining unit is used to determine the first correction coefficient of each grid in the target area at the first resolution based on the simulated precipitation field and the first precipitation observation data.

[0142] The mapping unit is used to map the first correction coefficients of multiple grids to the grid at the target scale using a preset interpolation algorithm, so as to obtain the second correction coefficients of each grid in the target area at the second resolution.

[0143] The correction unit is used to correct the corrected precipitation field data by using the second correction coefficient of each grid in the target area at the second resolution, so as to obtain the target precipitation data.

[0144] The precipitation data downscaling device provided in this embodiment of the invention can execute the precipitation data downscaling method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects for executing the method. Further functional descriptions of the various modules and units described above are the same as in the corresponding embodiments described above, and will not be repeated here.

[0145] Figure 9 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention.

[0146] The following is a detailed reference. Figure 9 This diagram illustrates a structural schematic suitable for implementing an electronic device according to embodiments of the present invention. The electronic device may include a processor (e.g., a central processing unit, graphics processor, etc.) 901, which can perform various appropriate actions and processes according to a program stored in read-only memory (ROM) 902 or a program loaded from memory 908 into random access memory (RAM) 903. The RAM 903 also stores various programs and data required for the operation of the electronic device. The processor 901, ROM 902, and RAM 903 are interconnected via a bus 904. An input / output (I / O) interface 905 is also connected to the bus 904.

[0147] Typically, the following devices can be connected to I / O interface 905: input devices 906 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 907 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; memory devices 908 including, for example, magnetic tapes, hard disks, etc.; and communication devices 909. Communication device 909 allows electronic devices to exchange data via wireless or wired communication with other devices. Although Figure 9 Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown, and more or fewer devices may be implemented or have instead.

[0148] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication device 909, or installed from a memory 908, or installed from a ROM 902. When the computer program is executed by a processor 901, it performs the functions defined in the precipitation data downscaling method of the embodiments of the present invention.

[0149] Figure 9 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0150] This invention also provides a computer-readable storage medium. The methods described above according to embodiments of the invention can be implemented in hardware or firmware, or implemented as computer code that can be recorded on a storage medium, or implemented as computer code downloaded via a network and originally stored on a remote storage medium or a non-transitory machine-readable storage medium and then stored on a local storage medium. Thus, the methods described herein can be processed by software stored on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. The storage medium can be a magnetic disk, optical disk, read-only memory, random access memory, flash memory, hard disk, or solid-state drive, etc.; further, the storage medium can also include combinations of the above types of memory. It is understood that computers, processors, microprocessor controllers, or programmable hardware include storage components capable of storing or receiving software or computer code. When the software or computer code is accessed and executed by the computer, processor, or hardware, the precipitation data downscaling method shown in the above embodiments is implemented.

[0151] A portion of this invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the invention through the operation of the computer. Those skilled in the art will understand that the forms in which computer program instructions exist in a computer-readable medium include, but are not limited to, source files, executable files, installation package files, etc. Correspondingly, the ways in which computer program instructions are executed by a computer include, but are not limited to: the computer directly executing the instructions, or the computer compiling the instructions and then executing the corresponding compiled program, or the computer reading and executing the instructions, or the computer reading and installing the instructions and then executing the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible to a computer.

[0152] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for downscaling precipitation data, characterized in that, The method includes: Acquire first precipitation observation data and target climatological data for the target area. The spatial resolution of the first precipitation observation data is a first resolution, and the observation period is a first time period. The spatial resolution of the target climatological data is a second resolution, and the statistical period is a second time period. The second resolution is greater than the first resolution, and the second time period corresponds to the first time period. A synoptic-scale climate trend surface with a second spatial resolution is constructed based on the first precipitation observation data; Based on the synoptic-scale climate trend surface, dynamic enhancement factors for multiple grids within the target area are calculated. These dynamic enhancement factors characterize the magnitude of precipitation intensity and are obtained using the following formula: in, This represents the enhancement factor of the i-th grid in the synoptic-scale climate trend surface; 1.0 represents the baseline value. The sensitivity coefficient is a calibrable empirical parameter used to control the intensity of the enhancement effect; it can be set to 0.

1. This represents the precipitation value on the weather-scale climate trend surface of the i-th grid. Based on the dynamic enhancement factors of each grid, a nonlinear transformation is performed on the target climatological data to obtain the spatially continuous weights of the target region. The nonlinear transformation of the target climatological data is then performed based on the following formula: in, This represents the weight of the i-th fine grid, where a fine grid refers to the grid at the second resolution; This represents the climatological data for the i-th fine grid. This indicates a background-scale Gaussian filtering operation; The smoothing coefficient for filtering; This indicates that for all fine meshes After performing background scale Gaussian filtering, the first The filtering result corresponding to each fine grid represents the average climatological response of the coarse-scale background to which the grid is located. Represents the regularization term; Indicates the dynamic enhancement factor; The first precipitation observation data is processed using the spatially continuous weights to obtain precipitation field data; The precipitation field data is corrected to obtain the target precipitation data.

2. The method according to claim 1, characterized in that, The steps for constructing a synoptic-scale climate trend surface with a second spatial resolution based on the first precipitation observation data include: The first precipitation observation data is mapped to a grid at the target scale using a preset interpolation algorithm to obtain the second precipitation observation data, and the spatial resolution of the second precipitation observation data is the second resolution; The second precipitation observation data is filtered to obtain a synoptic-scale climate trend surface.

3. The method according to claim 1 or 2, characterized in that, The target climatological data is determined through the following steps: Acquire initial climatological data for the target area; The initial climatological data is filled with invalid value regions using the nearest neighbor difference method to obtain the target climatological data.

4. The method according to claim 1 or 2, characterized in that, The precipitation field data is corrected to obtain the target precipitation data, including: Based on the precipitation values ​​of multiple grids in the weather-scale climate trend surface and the preset trace precipitation threshold, at least one target grid is determined among the multiple grids. The target grid is the grid whose precipitation value is less than the trace precipitation threshold. The climatological precipitation values ​​of each target grid in the precipitation field data are corrected to obtain the corrected precipitation field data; The target precipitation data is obtained by correcting the modified precipitation field data using a preset mass balance constraint algorithm.

5. The method according to claim 4, characterized in that, The climatological precipitation values ​​of each target grid in the precipitation field data are corrected to obtain the corrected precipitation field data, including: Extract multiple non-zero values ​​from the target climatological data; Sort the multiple non-zero values ​​to obtain the sorting result; The critical precipitation value is determined based on the sorting results and the preset ratio. Based on the precipitation threshold, the climatological precipitation values ​​of each target grid in the precipitation field data, and the preset transition width coefficient, the correction coefficients of the corresponding target grids are obtained by inputting them into a pre-constructed smoothing correction function. The climatological precipitation values ​​of the corresponding target grids in the precipitation field data are corrected using the correction coefficients of each target grid to obtain the corrected precipitation field data.

6. The method according to claim 4, characterized in that, The modified precipitation field data is corrected using a preset mass balance constraint algorithm to obtain the target precipitation data, including: The corrected precipitation field data are aggregated to obtain a simulated precipitation field, wherein the spatial resolution of the simulated precipitation field is a first resolution; Based on the simulated precipitation field and the first precipitation observation data, determine the first correction coefficient of each grid in the target area at the first resolution; The first correction coefficients of multiple grids are mapped to the grid at the target scale using a preset interpolation algorithm to obtain the second correction coefficients of each grid in the target region at the second resolution. The corrected precipitation field data is corrected using the second correction coefficients of each grid in the target area at the second resolution to obtain the target precipitation data.

7. A precipitation data downscaling device, characterized in that, The device includes: The acquisition module is used to acquire first precipitation observation data and target climatological data of the target area. The spatial resolution of the first precipitation observation data is a first resolution, and the observation period is a first time period. The spatial resolution of the target climatological data is a second resolution, and the statistical period is a second time period. The second resolution is greater than the first resolution, and the second time period corresponds to the first time period. The module is used to construct a weather-scale climate trend surface with a spatial resolution of the second resolution based on the first precipitation observation data. The calculation module is used to calculate the dynamic enhancement factor of multiple grids within the target area based on the synoptic-scale climate trend surface. The dynamic enhancement factor is used to characterize the magnitude of precipitation intensity and is calculated using the following formula: in, This represents the enhancement factor of the i-th grid in the synoptic-scale climate trend surface; 1.0 represents the baseline value. The sensitivity coefficient is a calibrable empirical parameter used to control the intensity of the enhancement effect; it can be set to 0.

1. This represents the precipitation value on the weather-scale climate trend surface of the i-th grid. The transformation module is used to perform a nonlinear transformation on the target climatological data based on the dynamic enhancement factors of each grid, to obtain the spatially continuous weights of the target region, and to perform a nonlinear transformation on the target climatological data based on the following formula: in, This represents the weight of the i-th fine grid, where a fine grid refers to the grid at the second resolution; This represents the climatological data for the i-th fine grid. This indicates a background-scale Gaussian filtering operation; The smoothing coefficient for filtering; This indicates that for all fine meshes After performing background scale Gaussian filtering, the first The filtering result corresponding to each fine grid represents the average climatological response of the coarse-scale background to which the grid is located. Represents the regularization term; Indicates the dynamic enhancement factor; The processing module is used to process the first precipitation observation data using the spatial continuous weights to obtain precipitation field data; The correction module is used to correct the precipitation field data to obtain the target precipitation data.

8. An electronic device, characterized in that, include: A memory and a processor are interconnected, the memory storing computer instructions, and the processor executing the computer instructions to perform the precipitation data downscaling method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing the computer to perform the precipitation data downscaling method according to any one of claims 1 to 6.

10. A computer program product, characterized in that, Includes computer instructions for causing a computer to perform the precipitation data downscaling method according to any one of claims 1 to 6.

Images