A method for quickly calculating rain data of a rain radar partition

Abstract

The application discloses a kind of quick calculation methods of rain radar partition rainfall data, it is related to meteorological technical field.The method is by constructing multistage basic rule grid system and stratified lossless compression weight matrix, complex geometric operation is converted into static look-up table;In real-time calculation stage, analysis radar data dynamic parameter, based on rainfall gradient threshold triggers adaptive grid refinement, uses coarse grid and nearest neighbor interpolation to gentle area of rainfall, uses fine grid and high-precision interpolation to dramatic change area and boundary area;Finally, through stratified compression weight, realizes the quick accumulation of partition statistics, outputs each partition surface average rainfall.The application reduces dimension from look-up table and weighted accumulation to full-amount geometric operation in real-time calculation load, under the premise that guaranteeing boundary statistical error is less than 1%, the computing time efficiency is improved by more than 90%, memory occupation is reduced by 90%, and can automatically adapt to different scanning modes of radar, provides efficient and accurate data support for flood forecasting and waterlogging early warning.

Description

Technical Field

[0001] This invention relates to the field of meteorological technology, and more specifically to a rapid calculation method for regional rainfall data from a rain-measuring radar, which can be applied to meteorology, hydrology, water conservancy and hydropower engineering and other fields. Background Technology

[0002] Rainfall-measuring radars (such as X-band phased array radars) provide high spatiotemporal resolution rainfall field data, serving as a crucial data source for urban flooding early warning and watershed flood forecasting. Radar data is typically stored in NetCDF format, containing rainfall values ​​within a regular grid. In water conservancy operations, this grid data must be mapped to specific administrative divisions or watershed zones (polygon vector data) to calculate the areal average rainfall for each zone, aiding in the assessment of rainfall patterns in different zones. Currently, the commonly used method is the centroid assignment method, which has the following significant problems:

[0003] 1) Computational bottleneck and poor timeliness: Traditional "point-polygon" geometry involves high computational complexity. When the radar coverage area is wide (millions of grids) and the boundaries of the partitioned polygons are complex, the calculation is extremely time-consuming and it is difficult to complete the calculation within the data update cycle (e.g., 1 minute), resulting in data backlog.

[0004] 2) Difficulty adapting to radar mode switching: Rain measuring radar switches between different scanning modes according to weather conditions, causing changes in the grid size (resolution) and origin coordinates of the output data. This means that without a fixed "radar-partition" lookup table established in advance, spatial calculations must be performed again for each data update.

[0005] 3) Large statistical errors at boundaries: Current technologies typically assign a grid point simply to a specific zone. However, at zone boundaries, a single grid often spans multiple administrative regions. This simplistic "either / or" classification leads to significant deviations in rainfall statistics at boundary areas, known as the "sawtooth effect," especially when the radar grid is coarse, where the error cannot be ignored.

[0006] 4) Fixed grid division lacks flexibility: It is difficult to distinguish between areas with gradual rainfall changes and areas with drastic rainfall changes, making it difficult to achieve a good balance between accuracy and efficiency. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of the prior art and provide a fast calculation method for rainfall data from a rain-measuring radar zone. It solves the technical problems of low computational efficiency caused by dynamic changes in rain-measuring radar grid parameters and poor statistical accuracy at region boundaries in the traditional center point assignment method. It focuses on addressing the technical bottlenecks of memory overload in lookup tables and computational redundancy caused by global high-fidelity calculations when introducing a high-resolution base grid to improve boundary accuracy.

[0008] The objective of this invention is achieved through the following technical solution: a rapid calculation method for regional rainfall data from a water conservancy rainfall radar, characterized by comprising the following steps:

[0009] Step S1: Construct a multi-level basic rule grid system and a hierarchical lossless compression weight matrix.

[0010] Step S2: Analyze the real-time rainfall radar results data, receive the real-time pushed rainfall radar data, and extract the dynamic parameters of the data at the current moment;

[0011] Step S3: Dynamic grid triggering and mapping based on rainfall gradient.

[0012] Step S4: Adaptive dimensionality reduction interpolation and numerical resampling.

[0013] Step S5: Fast accumulation of partition statistics based on hierarchical compression weights.

[0014] Step S6: Calculate the average rainfall for each zone, iterate through all zones, calculate the average rainfall, output the statistical results, and enter the next frame of radar data loop.

[0015] Preferred, step S1 includes: defining a multi-level basic rule grid system A covering the monitoring area; acquiring partition polygon vector data for each grid level; calculating the intersection area ratio between the unit and each partition; and constructing a hierarchical lossless compression weight matrix.

[0016] Prior to this, the multi-level basic rule grid system A defining the monitoring area includes:

[0017] coarse grid level Its resolution is comparable to that of conventional radar systems in the past;

[0018] Multiple refinement levels Subdivision by level, with the highest level of refinement. The resolution is 1 / 2, 1 / 4, 1 / 8 of the minimum radar resolution, etc.

[0019] Preferably, the construction of the hierarchical lossless compression weight matrix includes: for a pure grid in any level that falls completely within a single partition, one-dimensional run-length encoding compression is used, and only the start and end indices of the continuous grids and the partition ID are recorded;

[0020] For hybrid grids where the grid overlaps with the partition boundaries and crosses the boundaries, an independent hash table is used to store the exact two-dimensional index and the corresponding multi-partition weight array.

[0021] Prior to this, the dynamic parameters of the data at the current moment in step S2 include: the resolution b of the result grid B, the coordinates s,t of the upper left corner, and the rainfall matrix Data_Matrix.

[0022] Preferred, step S3 includes: (1) Global gradient scan: First, the spatial gradient matrix Gradient_Matrix of the current frame rainfall field is quickly calculated by mapping the coarse resolution network A0 to the resulting grid B;

[0023] (2) Adaptive mesh refinement: Traverse the scan results and compare the local rainfall gradient. With preset threshold set The following processing is performed:

[0024] like This area is a region with moderate or no rainfall, and the coarse grid level A0 is maintained in this area;

[0025] like This region is located in an area of ​​drastic rainfall changes and requires high-precision calculations. Therefore, mesh splitting is triggered, and the region is automatically refined down to a matching high-precision mesh level based on the gradient magnitude. or ;

[0026] If a coarse grid cell is marked as a mixed grid that crosses administrative boundaries, it should be forcibly refined to the highest precision level An, regardless of the rainfall gradient.

[0027] Based on the above operations, a dynamic composite mesh mapping map of the current frame is generated.

[0028] Preferably, step S4 includes:

[0029] Traverse the active grid cells in the dynamic composite mesh mapping graph, assuming the center point of each grid cell... Calculate its floating-point index coordinates in the B coordinate system of the result grid. ;

[0030]

[0031] For flat areas using a coarse grid level, the nearest neighbor interpolation is directly downgraded to quickly extract rainfall results.

[0032] For regions with drastic changes or boundary areas where a finer mesh level is used, bilinear or bicubic interpolation is employed, utilizing the fractional part. Calculate high-fidelity rainfall results .

[0033] Preferably, the calculation method for bilinear interpolation is as follows:

[0034] (1) Based on floating-point index Find the four adjacent grid points in grid B that surround the given point. The values ​​for rainfall are located at the top left, top right, bottom left, and bottom right, respectively.

[0035] (2) Take The decimal part corresponding to the floating-point index After performing bilinear interpolation, the rainfall result for grid A at that point is:

[0036] .

[0037] Preferably, step S5 includes:

[0038] Based on the level to which the current computing grid belongs The hierarchical LUT matrix corresponding to dynamic addressing:

[0039] If the internal pure zone is hit, press Full accumulation, of which This represents the ratio of the grid level's area to the base area;

[0040] If the boundary mixing region is hit, extract the multi-partition weights. ,according to Perform weighted allocation;

[0041] Simultaneously accumulate the sum of the effective area weights of the corresponding partitions. .

[0042] This invention achieves a balance between efficient calculation of areas with gentle rainfall changes and high-precision calculation of areas with drastic rainfall changes and boundary areas by constructing a multi-level basic regular grid system and a lossless compressed spatial weight matrix, and by introducing a dynamic grid refinement and adaptive interpolation strategy based on the spatial gradient of rainfall in real-time calculation.

[0043] By using grids and pre-computing spatial weights, complex geometric operations are transformed into static lookup tables, and dynamic grid matching is transformed into algebraic mapping and weighted accumulation, thereby enabling rapid real-time calculation of regional rainfall data while ensuring high accuracy. Attached Figure Description

[0044] Figure 1 This is a flowchart of the present invention.

[0045] Figure 2 This is a diagram showing the relationship between the multi-level grid A and the partition boundary in this invention. Detailed Implementation

[0046] The following detailed description of the implementation of the present invention is in conjunction with the accompanying drawings. However, these descriptions do not constitute a limitation of the present invention and are merely examples. The advantages of the present invention will become clearer and easier to understand by explaining them.

[0047] The inventive concept of this invention is to achieve a balance between efficient calculation of areas with gentle rainfall changes and high-precision calculation of areas with drastic rainfall changes and boundary areas by constructing a multi-level basic regular grid system and a lossless compressed spatial weight matrix, and by introducing a dynamic grid refinement and adaptive interpolation strategy based on the spatial gradient of rainfall in real-time calculation.

[0048] By using grids and pre-computing spatial weights, complex geometric operations are transformed into static lookup tables, and dynamic grid matching is transformed into algebraic mapping and weighted accumulation, thereby enabling rapid real-time calculation of regional rainfall data while ensuring high accuracy.

[0049] See appendix Figure 1 , Figure 2 It can be seen that the present invention provides a rapid calculation method for regional rainfall data from a water conservancy rainfall measurement radar, comprising the following steps:

[0050] Step S1: Construct a multi-level basic rule grid system and a hierarchical lossless compression weight matrix.

[0051] Step S2: Analyze the real-time rainfall radar results data, receive the real-time pushed rainfall radar data, and extract the dynamic parameters of the data at the current moment;

[0052] Step S3: Dynamic grid triggering and mapping based on rainfall gradient.

[0053] Step S4: Adaptive dimensionality reduction interpolation and numerical resampling.

[0054] Step S5: Fast accumulation of partition statistics based on hierarchical compression weights.

[0055] Step S6: Calculate the average rainfall for each zone, iterate through all zones, calculate the average rainfall, output the statistical results, and enter the next frame of radar data loop.

[0056] This invention constructs a closed-loop technology system encompassing "static preprocessing - dynamic adaptation - efficient accumulation" across the entire process. Unlike existing technologies that require repeated geometric calculations for each data update, this method pre-processes and post-processes complex calculations, retaining only lightweight mapping and interpolation operations in the real-time processing stage. This ensures the system's real-time responsiveness within minute-level or even second-level data update cycles from an architectural perspective.

[0057] In the above technical solution, step S1 includes: defining a multi-level basic rule grid system A covering the monitoring area; acquiring the partition polygon vector data for each grid level; calculating the intersection area ratio between the unit and each partition; and constructing a hierarchical lossless compression weight matrix.

[0058] This invention defines the key preliminary steps for constructing a "multi-level basic rule grid system". This feature expands a single, fixed computational grid into a pyramid structure with different resolutions, providing a physical basis for dynamically matching computational accuracy based on the spatial characteristics of rainfall, and avoiding the computational waste caused by using a single high-resolution grid across the entire domain in traditional methods.

[0059] In the above technical solution, the multi-level basic rule grid system A defining the coverage monitoring area includes: a coarse grid level. Its resolution is comparable to that of conventional radar systems; multiple levels of refinement. The subdivision is progressive, with the highest level of refinement, An, having a resolution of 1 / 2, 1 / 4, 1 / 8, etc., of the radar's minimum resolution. Each level of subdivision simply requires that the size decreases as the subdivision decreases, and that the next level of mesh can be completely nested within the previous level. A1 does not necessarily start at 1 / 2, and adjacent meshes are not necessarily in a half-to-half relationship.

[0060] This invention further defines the hierarchical relationship of the multi-level grid, specifically by establishing a correlation between the highest refinement level and the minimum resolution of the radar. This design ensures that in critical areas such as areas with drastic rainfall changes or administrative boundaries, the sampling density of the computational grid can match or even exceed the accuracy of the original radar data, fundamentally eliminating the "jagged effect" in boundary statistics caused by excessively coarse grids.

[0061] In the above technical solution, the construction of the hierarchical lossless compression weight matrix includes: for a pure grid in any level that falls completely in a single partition, one-dimensional run-length encoding compression is used, and only the start and end indices of the continuous grid and the partition ID are recorded;

[0062] For hybrid grids where the grid overlaps with the partition boundaries and crosses the boundaries, an independent hash table is used to store the exact two-dimensional index and the corresponding multi-partition weight array. .

[0063] This invention introduces an innovative method for constructing a "layered lossless compressed weight matrix". By distinguishing between "pure mesh" and "hybrid mesh" and storing them using run-length encoding and hash tables respectively, this scheme not only significantly reduces the memory footprint of the high-precision weight matrix, but also enables fast addressing during real-time computation, overcoming the memory bottleneck and addressing latency issues of traditional methods in large-scale high-precision application scenarios.

[0064] In the above technical solution, the dynamic parameters of the current time data in step S2 include: the resolution b of the result grid B, the coordinates s,t of the upper left corner, and the rainfall matrix Data_Matrix.

[0065] This invention clarifies the specific dynamic parameters extracted during the real-time analysis phase. This feature enables the method to automatically capture changing parameters such as grid size and origin coordinates output by the radar under different scanning modes (such as switching elevation angles or scanning ranges), providing a data foundation for subsequent "reverse dynamic mapping" and achieving seamless compatibility with changes in radar operating modes.

[0066] In the above technical solution, step S3 includes: (1) Global gradient scanning: First, the spatial gradient matrix Gradient_Matrix of the current frame rainfall field is quickly calculated by mapping the coarse resolution mesh A0 to the resulting mesh B; (2) Adaptive mesh refinement: Traverse the scanning results and compare the local rainfall gradient. With preset threshold set The following processing is performed:

[0067] like This area is characterized by moderate or no rainfall, and a coarse grid layer is maintained in this region. ;

[0068] like 1. If the region is located in an area of ​​drastic rainfall changes and requires high-precision calculation, then grid splitting will be triggered. Based on the gradient magnitude, the region will be automatically refined down to a matching high-precision grid level, A1 or An.

[0069] For example: , Split

[0070] , Split to

[0071] , Split to

[0072] ...

[0073] And so on.

[0074] If a coarse grid cell is marked as a mixed grid that crosses administrative boundaries, it should be forcibly refined to the highest precision level An, regardless of the rainfall gradient.

[0075] Based on the above operations, a dynamic composite mesh mapping map of the current frame is generated.

[0076] The core mechanism of this invention achieves a balance between accuracy and performance. By introducing rainfall gradients as triggers for mesh refinement and combining them with boundary-forced refinement rules, this method achieves "on-demand allocation" of computing power: using coarse meshes to ensure efficiency in vast areas with gentle rainfall, and automatically upgrading to high-precision meshes to ensure quality in areas with dramatic changes such as fronts and administrative boundaries. This is more flexible and scientific than the "either / or" single mode in existing technologies.

[0077] In the above technical solution, step S4 includes:

[0078] Traverse the active grid cells in the dynamic composite mesh mapping graph, assuming the center point of each grid cell... Calculate its floating-point index coordinates in the B coordinate system of the result grid. ;

[0079]

[0080] For flat areas using a coarse grid level, the nearest neighbor interpolation is directly downgraded to quickly extract rainfall results.

[0081] For regions with drastic changes or boundary areas where a finer mesh level is used, bilinear or bicubic interpolation is employed, utilizing the fractional part. Calculate high-fidelity rainfall results .

[0082] This invention proposes an "adaptive dimensionality reduction interpolation" strategy. Based on the grid level determined in step S3, this strategy dynamically selects the interpolation algorithm—using nearest-neighbor interpolation with extremely low computational cost in flat areas, and high-fidelity bilinear or bicubic interpolation in critical areas. This "site-specific" numerical resampling method ensures rainfall accuracy in critical areas while further saving computational resources in non-critical areas.

[0083] In the above technical solution, the calculation method for bilinear interpolation is as follows:

[0084] (1) Based on floating-point index Find the four adjacent grid points in grid B that surround the given point. The values ​​for rainfall are located at the top left, top right, bottom left, and bottom right, respectively.

[0085] (2) Take The decimal part corresponding to the floating-point index After performing bilinear interpolation, the rainfall result for grid A at that point is:

[0086] .

[0087] It should be noted that bilinear interpolation and bicubic interpolation are both commonly used interpolation algorithms in two-dimensional space. This explanation uses bilinear interpolation as an example; the corresponding bicubic interpolation and other interpolation methods can be easily derived by analogy.

[0088] This invention specifically discloses a bilinear interpolation method used in refined calculation regions and provides a clear mathematical formula. This feature combines the refined high-precision grid in step S3 with the high-fidelity interpolation algorithm in step S4, ensuring that in areas with drastic rainfall changes or complex boundaries, more continuous and accurate areal rainfall values ​​are calculated by fusing information from multiple surrounding radar grid points, avoiding errors caused by simplistic and crude "point assignment".

[0089] In the above technical solution, step S5 includes:

[0090] Based on the level to which the current computing grid belongs The hierarchical LUT matrix corresponding to dynamic addressing:

[0091] If the internal pure zone is hit, press Full accumulation, of which This represents the ratio of the grid level's area to the base area; The sum of all rainfall results for the zone numbered Zone_ID multiplied by the area ratio.

[0092] If the boundary mixing region is hit, extract the multi-partition weights. ,according to Perform weighted allocation;

[0093] Simultaneously accumulate the sum of the effective area weights of the corresponding partitions. .

[0094] This invention defines a method for zonal statistical accumulation based on hierarchical compression weights. This step efficiently summarizes all previous calculation results, distinguishes between the full accumulation of "pure zones" and the weighted allocation of "mixed zones," and combines the synchronous accumulation of area weights to ultimately achieve accurate calculation of the average rainfall of each zone, completing a seamless transformation from raw radar data to the hydrological information required for operations.

[0095] This invention proposes a complete "hierarchical-mapping-adaptive" technical framework, systematically solving the core contradiction of difficulty in achieving both accuracy and efficiency in rainfall calculation by rain-measuring radar. By pre-processing complex spatial geometric operations into a static multi-level weight matrix and introducing rainfall gradients as triggers for dynamic scheduling, this method reduces the real-time computational load from "full geometric operations" to "a small amount of table lookup and weighted accumulation" while ensuring statistical accuracy in the boundary area. This achieves configuration-free adaptive operation for arbitrary radar scanning modes, providing high-timeliness and high-precision data support for watershed flood forecasting and urban waterlogging early warning.

[0096] Example

[0097] 1. Scene initialization (static preprocessing stage)

[0098] Suppose the total area of ​​a certain watershed is divided into "upstream" "and downstream" "Two administrative regions or river basin hydrological zones, such as..." Figure 2 As shown.

[0099] 1) First, construct a multi-level basic grid.

[0100] coarse grid level Resolution set to (Approximately 200 meters). Fine mesh layer.

[0101] Resolution set to

[0102] (Approximately 50 meters). Area ratio That is, 1 16 grids Grid.

[0103] Then a hierarchical LUT dictionary is generated.

[0104] (1) Case A (pure area): coarse mesh In the index The location is entirely within the upstream region. Record the run-length encoding. .

[0105] (2) Case B (Boundary Region): Coarse Mesh In the index This location crosses the upstream and downstream boundaries. The system forcibly marks it as a mixed zone and only... High-precision weights are generated hierarchically, such as within a fine mesh. The weight is .

[0106] 2. Real-time data parsing and adaptive scheduling (dynamic calculation stage)

[0107] Assuming the latest frame of radar data is received, the radar grid... The resolution is 200 meters. Set the rainfall gradient threshold. .

[0108] 1) Calculation and simulation for areas with moderate rainfall.

[0109] This triggers the coarse mesh and nearest-zero algorithm for coarse meshes. Mapped to radar data The local area. The radar values ​​for the four adjacent points are respectively... First, calculate the gradient:

[0110] Since the gradient value is less than the threshold, mesh subdivision is not triggered. In this case, use... Coarse grid, meaning it does not perform bilinear interpolation and directly takes the nearest neighbor value. Weighted accumulation: Query the run-length encoding and directly accumulate the rainfall upstream (ID=1). (Multiplying by a factor of 16 represents the area weight of 16 finer grids).

[0111] 2) Calculation and simulation for drastic change zones / boundary zones.

[0112] At this point, fine meshing is triggered, and bilinear interpolation is used. For coarse meshing... (Boundary area), the radar values ​​for the four adjacent points are respectively .

[0113] First, calculate the gradient:

[0114] (And this area itself is a boundary region), at this point, mesh subdivision is triggered, traversing the 16 meshes under this region. Fine mesh and high-precision interpolation methods are used. (This is for fine meshes.) Calculate the fractional part of its floating-point index. Substituting the values ​​into the bilinear interpolation formula, the high-precision rainfall data can be calculated. Accumulate boundary weights: Query ,

[0115] Perform weighted splitting and accumulation:

[0116] .

[0117] 3. Effect Comparison

[0118] Through the above mechanism, the vast majority of areas (flat or pure regions) only require a very small number of table lookups and addition operations; while only a few frontal gradient regions or administrative boundaries are subjected to high-density floating-point multiplication and addition operations. This achieves a balance between accuracy and performance.

[0119] To further demonstrate the technical effectiveness of this invention, the following table compares the differences in technical indicators between this invention and currently common solutions in the industry:

[0120]

[0121] All other parts not described in detail are existing technologies.

Claims

1. A rapid calculation method for regional rainfall data from a water conservancy rainfall measurement radar, characterized in that, Includes the following steps: Step S1: Construct a multi-level basic rule grid system and a hierarchical lossless compression weight matrix. Step S2: Analyze the real-time rainfall radar results data, receive the real-time pushed rainfall radar data, and extract the dynamic parameters of the data at the current moment; Step S3: Dynamic grid triggering and mapping based on rainfall gradient. Step S4: Adaptive dimensionality reduction interpolation and numerical resampling. Step S5: Fast accumulation of partition statistics based on hierarchical compression weights. Step S6: Calculate the average rainfall for each zone, iterate through all zones, calculate the average rainfall, output the statistical results, and enter the next frame of radar data loop.

2. The method for rapid calculation of regional rainfall data by hydrological rainfall radar according to claim 1, characterized in that, Step S1 includes: defining a multi-level basic rule grid system A covering the monitoring area; acquiring the polygon vector data of each partition for each grid level; calculating the intersection area ratio between the unit and each partition; and constructing a hierarchical lossless compression weight matrix.

3. The rapid calculation method for regional rainfall data from a water conservancy rainfall measurement radar according to claim 2, characterized in that, The multi-level basic rule grid system A that defines the monitoring area includes: coarse grid level Its resolution is comparable to that of conventional radar systems in the past; Multiple refinement levels The resolution is subdivided step by step, with the highest level of subdivision, An, having a resolution of 1 / 2, 1 / 4, 1 / 8, etc., of the minimum radar resolution.

4. The method for rapid calculation of regional rainfall data by water conservancy rainfall measurement radar according to claim 2, characterized in that, The construction of the hierarchical lossless compression weight matrix includes: for pure grids in any level that fall completely within a single partition, one-dimensional run-length encoding compression is used, and only the start and end indices of continuous grids and partition IDs are recorded; For hybrid grids where the grid overlaps with the partition boundaries and crosses the boundaries, an independent hash table is used to store the exact two-dimensional index and the corresponding multi-partition weight array. .

5. The rapid calculation method for regional rainfall data from a water conservancy rainfall measurement radar according to claim 1, characterized in that, The dynamic parameters of the current time data in step S2 include: the resolution b of the result grid B, the coordinates s,t of the upper left corner, and the rainfall matrix Data_Matrix.

6. The method for rapid calculation of regional rainfall data by hydrological rainfall radar according to claim 1, characterized in that, Step S3 includes: (1) Global gradient scanning: First, the coarse resolution network A0 is used to map the resulting grid B, and the spatial gradient matrix Gradient_Matrix of the current frame rainfall field is quickly calculated. (2) Adaptive mesh refinement: Traverse the scan results and compare the local rainfall gradient. local With preset threshold set The following processing is performed: like This area is a region with moderate or no rainfall, and the coarse grid level A0 is maintained in this area; like If the region is located in an area of ​​drastic rainfall changes and requires high-precision calculations, then grid splitting is triggered. Based on the gradient magnitude, the region is automatically refined down to a matching high-precision grid level, A1 or An. If a coarse grid cell is marked as a mixed grid that crosses administrative boundaries, it should be forcibly refined to the highest precision level An, regardless of the rainfall gradient. Based on the above operations, a dynamic composite mesh mapping map of the current frame is generated.

7. The rapid calculation method for regional rainfall data from a water conservancy rainfall measurement radar according to claim 1, characterized in that, Step S4 includes: Traverse the active grid cells in the dynamic composite mesh mapping graph, assuming the center point of each grid cell... Calculate its floating-point index coordinates in the B coordinate system of the result grid. ; ； For flat areas using a coarse grid level, the nearest neighbor interpolation is directly downgraded to quickly extract rainfall results. For regions with drastic changes or boundary areas where a finer mesh level is used, bilinear or bicubic interpolation is employed, utilizing the fractional part. Calculate high-fidelity rainfall results .

8. A rapid calculation method for regional rainfall data from a water conservancy rainfall measurement radar according to claim 7, characterized in that, The calculation method for bilinear interpolation is as follows: (1) Based on floating-point index Find the four adjacent grid points in grid B that surround the given point. The values ​​for rainfall are located at the top left, top right, bottom left, and bottom right, respectively. (2) Take The decimal part corresponding to the floating-point index After performing bilinear interpolation, the rainfall result for grid A at that point is: 。 9. A rapid calculation method for regional rainfall data from a water conservancy rainfall measurement radar according to claim 1, characterized in that, Step S5 includes: Based on the level to which the current computing grid belongs The hierarchical LUT matrix corresponding to dynamic addressing: If the internal pure zone is hit, press Full accumulation, of which This represents the ratio of the grid level's area to the base area; If the boundary mixing region is hit, extract the multi-partition weights. ,according to Perform weighted allocation; Simultaneously accumulate the sum of the effective area weights of the corresponding partitions. .

Images