A quantitative rainfall estimation method based on dual-polarization multi-radar composite technique

Abstract

The application relates to the technical field of radar data processing, and specifically discloses a quantitative rainfall estimation method based on a dual-polarization multi-radar composite technology, which comprises the following steps: obtaining standardized data by preprocessing dual-polarization radar original observation data; calculating and fusing a multi-dimensional quality index: calculating a static environment index and a dynamic physical consistency index, and fusing to obtain a comprehensive quality index; adaptively selecting an optimal rainfall estimation algorithm according to the quality of radar dual-polarization observation data to generate an initial fused rainfall field; dividing multiple quality levels, independently calculating an average field bias correction factor for each quality level, and applying the average field bias correction factor to radar pixels of the corresponding quality level to obtain a final rainfall inversion result. The application realizes source identification and filtering of non-meteorological echoes and abnormal data, avoids the propagation of false information and the amplification of system errors, and improves the reliability and robustness of radar network composition.

Description

Technical Field

[0001] This invention relates to the field of radar data processing technology, specifically to a quantitative rainfall estimation method based on dual-polarization multi-radar composite technology. Background Technology

[0002] The spatiotemporal variations of rainfall and specific watershed characteristics profoundly influence the hydrological and geomorphological behavior of watersheds. The emergence of dual-polarization radar technology has brought revolutionary progress to high-precision rainfall monitoring. Compared with traditional single-polarization radar, dual-polarization radar, by transmitting and receiving electromagnetic waves with both horizontal and vertical polarization directions, can acquire multiple observational parameters such as horizontal reflectivity factor (Z), differential reflectivity (ZDR), differential phase ratio (KDP), and correlation coefficient (ρhv). These parameters can not only more accurately estimate rainfall intensity, but also effectively distinguish different precipitation types such as rain, snowfall, and hail, identify non-meteorological echoes such as ground clutter, and correct for radar signal attenuation, thereby significantly improving the physical basis and accuracy of rainfall estimation.

[0003] However, numerous technical challenges arise in applying dual-polarization radar networks for quantitative precipitation estimation (QPE). Especially when deploying radar stations in areas with significant topographical differences, beam blocking and ground clutter effects are unavoidable. Rainfall amounts are often underestimated when using a single weather radar due to beam blocking. Therefore, radar composite technology has emerged. By combining data from multiple radar networks in overlapping areas of weather radar coverage, radar network composite technology can effectively improve the accuracy of data in these overlapping areas. Traditional composite estimation algorithms are relatively simple, such as directly selecting the maximum available reflectivity at a pixel. However, this method cannot guarantee the quality of the composite product, tends to overestimate, and may propagate non-meteorological echoes from individual radars. Therefore, a more complex composite method is introduced. First, the observation quality of each radar is evaluated. Multiple representative quality factors are multiplicatively combined to form an overall quality index for each radar station. Then, a weighted average of the quality factors is used to calculate the radar reflectivity value of the overlapping area. However, the weighting of different quality factors in the composite estimation process faces challenges. Furthermore, traditional quality indices based on static environmental factors such as range and beam jamming rate cannot fully reflect the physical consistency and instantaneous reliability of dual-polarization radar observation data. Therefore, introducing a novel quality index that can fully consider the characteristics of dual-polarization observations and reflect the relative quality levels of each radar within the composite region is particularly important.

[0004] Furthermore, effective bias correction techniques during the composite process are crucial for improving the quality of radar rainfall estimation. Traditional bias correction methods adopt a globally uniform approach, adjusting a spatially uniform but temporally varying static bias factor. This method neither considers the uncertainty of radar observation quality nor the spatial variability of the bias factor, which may lead to significant errors in radar rainfall estimation in areas with signal obstruction. Summary of the Invention

[0005] The purpose of this invention is to provide a quantitative rainfall estimation method based on dual-polarization multi-radar composite technology to solve the following technical problems: (1) The problem of missing physical mechanism in the quality index construction process in radar network composite technology; (2) The problem of coarse granularity in the deviation correction of radar network composite rainfall assessment.

[0006] The objective of this invention can be achieved through the following technical solutions: A quantitative rainfall estimation method based on dual-polarization multi-radar composite technology includes the following steps: Step S1: Data preprocessing: The core parameters of reflectivity, differential reflectivity, and differential phase shift in the raw observation data of the dual-polarization radar are subjected to quality control, attenuation correction, and spatial gridding to obtain standardized data; Step S2: Calculation and fusion of multi-dimensional quality indices: Calculate the static environmental index, which includes distance quality index, beam jamming quality index, and height quality index; introduce the dynamic physical consistency index based on hydrophobic material classification and physical consistency check; and fuse the static environmental index and the dynamic physical consistency index to obtain the comprehensive quality index. Step S3: Adaptive multi-radar data fusion: For each radar, the optimal rainfall estimation algorithm is adaptively selected based on the quality of its dual-polarization observation data. In the radar overlap area, the optimal rainfall estimates of each radar are weighted and averaged using the comprehensive quality index as the weight to generate an initial fused rainfall field. Step S4: Spatial Stratification Bias Correction: Based on the comprehensive quality index, the radar-rain gauge data pair is divided into multiple quality levels. The mean field bias correction factor is calculated independently for each quality level and applied to the radar pixels of the corresponding quality level to obtain the final rainfall inversion result.

[0007] As a further aspect of the present invention: in step S2, the static environment index QI Static The calculation formula is as follows: QI static =QI D ×QI B ×QI H ; Among them, QI DThe distance quality index, QI B The beam jamming quality index, QI H This indicates a high quality index.

[0008] As a further aspect of the present invention: the distance quality index QI D The calculation formula is as follows: ; Where r is the straight-line distance between the point and the radar station; r min Indicates the minimum calculated distance; r max This indicates the radar's maximum effective observation radius.

[0009] As a further aspect of the present invention: in step S2, the beam blocking quality index QI B The calculation formula is as follows: QI B =1-B; Where B represents the beam blocking rate, with a value range of [0,1]. B=1 indicates complete blocking and zero quality.

[0010] As a further aspect of the present invention: in step S2, the height quality index QI H The calculation formula is as follows: ; Where h represents the ground elevation of the radar beam center at that pixel point, h max This indicates the maximum acceptable beam height threshold.

[0011] As a further aspect of the present invention: in step S2, the dynamic physical consistency index QI physical The calculation formula is as follows: ; Among them, f class This represents the quality factor based on hydrogel classification, with a value range of [0,1]; f consistency Indicates the physical consistency factor; The physical consistency factor f consistency The calculation formula is: ; Among them, Z observed Z is the measured reflectance factor. expected The desired reflectance is estimated based on KDP or ZDR, where k is a scaling parameter; Integrating the static environmental index QI static With the dynamic physical consistency index QI physical The overall quality index (QI) is obtained. Integrated =QIstatic ×QI physical .

[0012] As a further aspect of the present invention: in step S3, the fusion rainfall rate R of a certain pixel in the initial fused rainfall field fusion for: ; Where i represents the radar index, N represents the total number of radars, and R i This represents the optimal rainfall rate corresponding to radar i. This represents the overall quality index corresponding to radar i.

[0013] As a further aspect of the present invention: in step S4, the overall quality index QI of each pixel after fusion is determined... Integrated The radar-rain gauge data pairs are divided into several quality layers; For the k-th mass layer, at time step t, its independent mean field deviation factor MFB k，t for: ; Where j represents the index of the rain gauge, M k G represents the total number of rain gauges. j，t R represents the measurement value of rain gauge j in time period t. fusion，j,t This represents the estimated rainfall value of pixel j in time period t after fusion. Calculate the final corrected precipitation field R final : .

[0014] The beneficial effects of this invention are: 1) By utilizing dual-polarization particle classification and physical consistency checks, the source identification and filtering of non-meteorological echoes and anomalous data were achieved, avoiding the propagation of erroneous information and the amplification of system errors, and improving the reliability and robustness of radar network composites.

[0015] 2) The deviation correction process breaks away from the extensive model of "one correction factor applicable to the entire area" and realizes spatial hierarchical deviation correction based on data quality. It can provide more targeted and applicable correction factors for low-quality areas with severe beam obstruction, effectively solving the long-standing systematic underestimation problem in these areas.

[0016] 3) This invention prioritizes the use of KDP parameters, which are insensitive to beam obstruction, for rainfall estimation. Combined with physical consistency constraints, it can effectively recover the real heavy rainfall information of areas obscured by terrain and solve the problem of data scarcity caused by complex mountainous terrain. Attached Figure Description

[0017] The invention will now be further described with reference to the accompanying drawings.

[0018] Figure 1 This is a flowchart illustrating a quantitative rainfall estimation method based on dual-polarization multi-radar composite technology according to the present invention. Detailed Implementation

[0019] 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, and 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.

[0020] The technical problem this invention aims to solve is to overcome the limitations of dual-polarization radar network composite rainfall assessment technology and deviation correction. The specific details are as follows: (1) Solving the problem of the lack of physical mechanism in the quality index construction process in radar network composite technology: Existing technologies rely on static, geometrically and environmentally based quality indices, which fail to reflect the instantaneous physical reliability of radar observation data. They cannot distinguish whether a high-reflectivity pixel originates from genuine heavy precipitation or is caused by ground clutter or non-meteorological targets such as hail, leading to the inclusion of low-quality or erroneous data in the fusion process and contaminating the final product. This invention leverages the advantage of dual-polarization radar's KDP parameters in distinguishing rainfall patterns to construct a comprehensive physical quality control index, thereby improving the accuracy of the quality indices.

[0021] (2) Solving the problem of coarse granularity in the bias correction of radar network composite rainfall assessment: Traditional mean field bias (MFB) correction calculates and applies a uniform correction factor to the entire region, ignoring the spatial variability of the bias factor caused by factors such as beam obstruction and range attenuation. This results in insufficient correction in areas severely affected by beam obstruction, while areas with good quality may be overcorrected, limiting the overall accuracy improvement. This invention, based on multi-radar fusion, achieves more flexible and refined bias correction through a hierarchical approach, improving the accuracy and reliability of multi-radar joint rainfall retrieval.

[0022] Please see Figure 1 As shown, this invention is a quantitative rainfall estimation method based on dual-polarization multi-radar composite technology, which mainly includes the following steps: Step S1: Data preprocessing: The core parameters of reflectivity, differential reflectivity, and differential phase shift in the raw observation data of the dual-polarization radar are subjected to quality control, attenuation correction, and spatial gridding to obtain standardized data; Step S2: Calculation and fusion of multi-dimensional quality indices: Calculate the static environmental index, which includes distance quality index, beam jamming quality index, and height quality index; introduce the dynamic physical consistency index based on hydrophobic material classification and physical consistency check; and fuse the static environmental index and the dynamic physical consistency index to obtain the comprehensive quality index. Step S3: Adaptive multi-radar data fusion: For each radar, the optimal rainfall estimation algorithm is adaptively selected based on the quality of its dual-polarization observation data. In the radar overlap area, the optimal rainfall estimates of each radar are weighted and averaged using the comprehensive quality index as the weight to generate an initial fused rainfall field. Step S4: Spatial Stratification Bias Correction: Based on the comprehensive quality index, the radar-rain gauge data pair is divided into multiple quality levels. The mean field bias correction factor is calculated independently for each quality level and applied to the radar pixels of the corresponding quality level to obtain the final rainfall inversion result.

[0023] In the calculation and fusion of the multi-dimensional quality index in step S2: Calculate a comprehensive quality index for each pixel of each dual-polarization radar in the network. This index is a combination of the static environment index and the dynamic physical index.

[0024] S2.1, Static Environmental Index (QI) static )calculate: This part of the index reflects long-term, stable changes in data quality caused by radar geometry and terrain.

[0025] Distance to Quality Index (QI) D : ; r is the straight-line distance (km) between the point and the radar station; min This represents the minimum calculation distance, usually 0 km; r max Indicates the radar's maximum effective observation radius (km); The physical meaning of this formula is that the greater the distance, the lower the quality due to beam broadening, volume averaging effect, and increased altitude.

[0026] Beamblocking Quality Index (QI) B : QI B =1-B; Where B represents the beam blocking rate, calculated using a digital elevation model and a radar beam propagation model, and its value ranges from [0,1]. This formula represents the degree to which terrain obstructs the radar beam; B=1 indicates complete obstruction, with a quality of zero.

[0027] High Quality Index (QI) H : ; h represents the altitude (km) of the radar beam center at that pixel. max This represents the maximum acceptable beam height threshold (km). The formula indicates that the higher the beam is above the ground, the worse the correlation between its measured reflectivity and ground precipitation.

[0028] Static Environment Index (QI) Static Fusion: QI static =QI D ×QI B ×QI H .

[0029] S2.2, Dynamic Physical Consistency Index (QI) physical )calculate: This index uses dual polarization parameters to evaluate the physical reliability of each pixel's observation in real time.

[0030] ; Quality factor (f) based on hydrocondensate classification class ): Fuzzy logic or decision tree hydrogel classification is performed using dual polarization parameters (Z, ZDR, ρhv, KDP), and a preset quality factor is assigned to each category.

[0031] f class =1.0: (Liquid) Rainfall; f class =0.8: Dry snow / ice crystals; f class =0.6: Wet snow / mixed phase; f class =0.3: Hail; f class =0: Ground clutter, biological echoes, non-meteorological echoes; Physical consistency factor (f consistency ): ; Among them, Z observed Z is the measured reflectance factor (dBZ). expected To estimate the desired reflectance based on KDP or ZDR, Z is estimated using empirical relationships. expected =a×KDP b An estimation is performed, where a and b are empirical coefficients. k is a scaling parameter used to control the strictness of the consistency check, and is set to 10 dBZ; This formula indicates that if the measured reflectance differs significantly from the reflectance estimated by KDP, which is insensitive to attenuation and shading, the data is considered unreliable, and its quality score is reduced.

[0032] Integrated Quality Index (QI) Integrated ): QI Integrated =QI static ×QI physical .

[0033] In step S3, adaptive multi-radar data fusion, adaptive fusion is performed for each pixel in the overlapping area of ​​the radar network, specifically including: S3.1 Selection of the optimal rainfall estimation algorithm: For each radar i, the optimal rainfall rate R is dynamically selected based on its physical observation conditions. i : Condition 1: If the KDP data quality is reliable (high signal-to-noise ratio and QI) physical If the threshold is greater than T1, then the following method will be used first: R i =a×KDP b ; Where a and b are empirical coefficients, a=40.5 and b=0.85.

[0034] Condition 2: If Condition 1 is not met, but ZDR is reliable and the classification is rainfall, then use: R i =c×Z d ×ZDR e ; Where c, d, and e are empirical coefficients, and Z represents the horizontal reflectivity factor.

[0035] Condition 3: If none of the above conditions are met, then use: R i =f×Z g ; Where f and g are the traditional ZR relationship coefficients.

[0036] S3.2, Quality-Weighted Fusion: The final fusion rainfall rate R of this pixel fusion for: ; A weighted sum is calculated for all N radars that are validly observed at this pixel.

[0037] In step S4, spatial stratification deviation correction, the specific steps include: S4.1, Quality Stratification: Based on the QI of each pixel in the fused productIntegrated The values ​​are used to divide all radar-rain gauge data pairs into N quality tiers (N≥2). The thresholds are determined through historical data optimization to ensure a relatively balanced data volume between the two tiers and the highest stability of the bias factor.

[0038] S4.2, Calculation of Stratification Deviation: For each layer k, at time step t, calculate its independent mean field bias factor MFB. k，t : ; For all valid M within layer k k The summation is performed using multiple rain gauges.

[0039] Among them, G j，t This represents the measurement value of rain gauge j during time period t. R fusion，j,t This represents the estimated rainfall value of the fused product at the corresponding pixel point of rain gauge j.

[0040] S4.3, Rainfall Inversion: The calculated layering deviation factor is applied to all radar pixels of the corresponding quality layer to obtain the final corrected precipitation field R. final : .

[0041] The pixel belongs to the k-th layer.

[0042] The foregoing has provided a detailed description of one embodiment of the present invention, but this description is merely a preferred embodiment and should not be construed as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the present invention should still fall within the scope of the present invention.

Claims

1. A quantitative rainfall estimation method based on dual-polarization multi-radar composite technology, characterized in that, Includes the following steps: Step S1: Data preprocessing: The core parameters of reflectivity, differential reflectivity, and differential phase shift in the raw observation data of the dual-polarization radar are subjected to quality control, attenuation correction, and spatial gridding to obtain standardized data; Step S2: Calculation and fusion of multi-dimensional quality indices: Calculate the static environmental index, which includes distance quality index, beam jamming quality index, and height quality index; introduce the dynamic physical consistency index based on hydrophobic material classification and physical consistency check; and fuse the static environmental index and the dynamic physical consistency index to obtain the comprehensive quality index. Step S3: Adaptive multi-radar data fusion: For each radar, the optimal rainfall estimation algorithm is adaptively selected based on the quality of its dual-polarization observation data. In the radar overlap area, the optimal rainfall estimates of each radar are weighted and averaged using the comprehensive quality index as the weight to generate an initial fused rainfall field. Step S4: Spatial Stratification Bias Correction: Based on the comprehensive quality index, the radar-rain gauge data pair is divided into multiple quality levels. The mean field bias correction factor is calculated independently for each quality level and applied to the radar pixels of the corresponding quality level to obtain the final rainfall inversion result.

2. The quantitative rainfall estimation method based on dual-polarization multi-radar composite technology according to claim 1, characterized in that, In step S2, the static environment index QI Static The calculation formula is as follows: QI static =QI D ×QI B ×QI H ; Among them, QI D The distance quality index, QI B The beam jamming quality index, QI H This indicates a high quality index.

3. The quantitative rainfall estimation method based on dual-polarization multi-radar composite technology according to claim 2, characterized in that, The distance quality index QI D The calculation formula is as follows: ； Where r is the straight-line distance between the point and the radar station; r min Indicates the minimum calculated distance; r max This indicates the radar's maximum effective observation radius.

4. The quantitative rainfall estimation method based on dual-polarization multi-radar composite technology according to claim 2, characterized in that, In step S2, the beam blocking quality index QI B The calculation formula is as follows: QI B =1-B; Where B represents the beam blocking rate, with a value range of [0,1]. B=1 indicates complete blocking and zero quality.

5. A quantitative rainfall estimation method based on dual-polarization multi-radar composite technology according to claim 2, characterized in that, In step S2, the height quality index QI H The calculation formula is as follows: ； Where h represents the ground elevation of the radar beam center at that pixel point, h max This indicates the maximum acceptable beam height threshold.

6. The quantitative rainfall estimation method based on dual-polarization multi-radar composite technology according to claim 2, characterized in that, In step S2, the dynamic physical consistency index QI physical The calculation formula is as follows: ； Among them, f class This represents the quality factor based on hydrogel classification, with a value range of [0,1]; f consistency Indicates the physical consistency factor; The physical consistency factor f consistency The calculation formula is: ； Among them, Z observed Z is the measured reflectance factor. expected The desired reflectance is estimated based on KDP or ZDR, where k is a scaling parameter; Integrating the static environmental index QI static With the dynamic physical consistency index QI physical The overall quality index (QI) is obtained. Integrated =QI static ×QI physical .

7. A quantitative rainfall estimation method based on dual-polarization multi-radar composite technology according to claim 6, characterized in that, In step S3, the fusion rainfall rate R of a certain pixel in the initial fused rainfall field fusion for: ； Where i represents the radar index, N represents the total number of radars, and R i This represents the optimal rainfall rate corresponding to radar i. This represents the overall quality index corresponding to radar i.

8. A quantitative rainfall estimation method based on dual-polarization multi-radar composite technology according to claim 6, characterized in that, In step S4, the overall quality index (QI) of each pixel after fusion is determined. Integrated The radar-rain gauge data pairs are divided into several quality layers; For the k-th mass layer, at time step t, its independent mean field deviation factor MFB k，t for: ； Where j represents the index of the rain gauge, M k G represents the total number of rain gauges. j，t R represents the measurement value of rain gauge j in time period t. fusion，j,t This represents the estimated rainfall value of pixel j in time period t after fusion. Calculate the final corrected precipitation field R final : 。

Images