A multi-band weather radar data fusion method

Abstract

The application discloses a kind of multi-band weather radar data fusion methods, step 1, establish band constraint raindrop spectrum distribution model;Step 2, establish multi-band weather radar data inversion model;Step 3, meteorological monitoring;Step 4, correction radar parameter;Step 5, multi-band radar parameter data conversion, specifically: step 5-1, select setting conversion waveband;Step 5-2, determine the raindrop spectrum parameter of waveband to be converted;Step 5-3, conversion;Step 6, grid interpolation;Step 7, fusion.The application can fuse multi-band radar data, as far as possible reduce the difference due to multi-band radar detection same weather target itself exists, to facilitate user in the monitoring, early warning and forecast work of strong convective weather the application of multi-band weather radar data.

Description

Technical Field

[0001] This invention relates to the field of weather radar data fusion, and in particular to a multi-band weather radar data fusion method. Background Technology

[0002] Severe convective weather, including short-duration heavy rainfall, thunderstorms, strong winds, hail, lightning, and tornadoes, is a type of high-impact, highly localized, and destructive weather event. Weather radar, by emitting electromagnetic waves, can observe the internal three-dimensional structure and evolution characteristics of severe convective systems, making it the most effective means of monitoring severe convection. However, my country currently has numerous operational weather radar models, including three main categories: X-band, C-band, and S-band, with a particularly large-scale deployment of X-band weather radars in recent years. Therefore, how to integrate data from all types of weather radar has become one of the urgent problems to be solved in operational applications.

[0003] In recent years, my country has made rapid progress in the development of X-band weather radar. Currently, high-density weather radar networks coordinating X and S bands have been established in major urban clusters such as the Beijing-Tianjin-Hebei region, the Yangtze River Delta, and the Pearl River Delta, along with corresponding application software systems. Regarding application systems, the Beijing-Tianjin-Hebei region primarily uses the Beijing X-band Networked Radar Application System (BJ-Xnet), independently developed by the Beijing Urban Meteorological Research Institute; Shanghai in the Yangtze River Delta region mainly uses the X-band array weather radar network application system developed by the Joint Laboratory for Phased Array Weather Radar; Jiangsu mainly uses the X-band weather radar network application system provided by radar manufacturers; and Guangdong in the Pearl River Delta region mainly uses the X-band phased array weather radar network application system provided by radar manufacturers. However, none of these application systems integrate X and S band weather radars; the two systems operate independently, causing numerous inconveniences for operations and hindering the realization of their synergistic effects.

[0004] Recently, the Beijing Urban Meteorological Research Institute is upgrading BJ-Xnet to a multi-band networked weather radar application system (referred to as "BJ-SXnet"). When it comes to multi-band weather radar data fusion, direct fusion and fitting-corrected fusion are currently the most commonly used methods. Direct fusion does not consider the differences in how multi-band weather radars detect the same weather target. Linear correctioned fusion refers to calculating the corresponding X-band and S-band radar detection results separately using long-term raindrop spectrum observation data, obtaining the conversion relationship between the two through fitting, and then using this conversion relationship for conversion and fusion of different band radars. For strong convective systems, due to the different electromagnetic wave scattering characteristics of different band radars, the X-band and S-band weather radars detect the same weather target differently (e.g., the intensity detected by the X-band is 60 dBZ, and the intensity detected by the S-band is 55 dBZ). Direct fusion is clearly an unscientific approach. Furthermore, the linear correctioned fusion method uses mathematical statistics to fit single variables individually, without considering the relationship between variables of dual-polarization radars, and therefore lacks physical constraints. Therefore, both of the above methods have certain drawbacks and are not conducive to the monitoring, early warning and forecasting of severe convective weather. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to address the shortcomings of the prior art by providing a multi-band weather radar data fusion method. This multi-band weather radar data fusion method can access and process X / C / S band weather radar base data, perform data quality control and fusion processing, and finally generate a gridded fusion product.

[0006] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is as follows:

[0007] A method for fusing multi-band weather radar data includes the following steps.

[0008] Step 1: Establish a constrained raindrop spectral distribution model, which includes the following steps.

[0009] Step 1-1: Select the raindrop spectrum distribution model: Use the Gamma distribution model to invert the raindrop spectrum N(D) of Doppler weather radar or dual polarization radar in different bands; where different bands include X-band, C-band and S-band; the raindrop spectrum N(D) of the Gamma distribution model includes the truncation parameter N0, the shape parameter μ and the slope parameter Λ.

[0010] Step 1-2: Collect raindrop spectrum observation data: For the area to be monitored, use a raindrop spectrometer to collect several sets of raindrop spectrum observation data N(D) for at least one year.

[0011] Steps 1-3: Establishing the μ-Λ constraint function: Substitute the several sets of raindrop spectrum observation data N(D) collected in Step 1-2 into the raindrop spectrum distribution model in Step 1-1 to obtain several sets of μ and Λ values; fit the several sets of μ and Λ values ​​to form a μ-Λ constraint function in which μ varies with Λ; substitute the μ-Λ constraint function into the raindrop spectrum distribution model in Step 1-1 to form a constrained raindrop spectrum distribution model.

[0012] Step 2: Establish a multi-band weather radar data inversion model: Based on the constrained raindrop spectral distribution model established in Step 1 and the complex refractive index m of precipitation particles, establish a multi-band weather radar data inversion model; the multi-band weather radar data inversion model includes one or two radar parameter models, namely the reflectivity model Z. H and / or differential reflectivity model Z DR The complex refractive index m is determined based on the radar wavelength.

[0013] Step 3, Meteorological Monitoring: At least two band Doppler weather radars or dual-polarization radars are used to conduct meteorological observations on the same area to be monitored; the reflectivity Z′ corresponding to each polar coordinate point is observed by each band Doppler weather radar or dual-polarization radar. H and / or differential reflectivity Z′ DR .

[0014] Step 4: Correct radar parameters: Correct the Z′ values ​​observed by each band of Doppler weather radar or dual-polarization radar in Step 3. H and / or Z′ DR The corrected radar parameters are obtained by performing corrections on each parameter, namely the corrected reflectivity Z′. H Z′ and / or modified differential reflectivity Z′ D ′ R .

[0015] Step 5: Multi-band radar parameter data conversion, which includes the following steps.

[0016] Step 5-1: Select and set conversion band A, where A is X-band, C-band, or S-band; if the band to be converted is the same as the set conversion band A, directly update the conversion radar parameters and jump to step 6; otherwise, proceed to step 5-2.

[0017] Step 5-2: Determine the raindrop spectral parameters of the band to be converted: Calculate the corrected reflectivity Z′ of the band to be converted. H Substituting the complex refractive index m corresponding to the band to be converted into the reflectivity model Z established in step 2, we can understand the relationship between the band and the complex refractive index m. H Equation 1 is obtained; simultaneously, the corrected differential reflectivity Z′ of the band to be converted is... D ′ RSubstituting the complex refractive index m corresponding to the band to be converted into the differential reflectivity model Z established in step 2, we can understand its meaning. DR Equation 2 is obtained from equation 1; by simultaneously solving equations 1, 2 and the μ-Λ constraint function, the raindrop spectral parameters corresponding to the band to be converted are obtained, which are the cutoff parameter N0, shape parameter μ and slope parameter Λ of the converted band.

[0018] Step 5-3, Conversion: Set the complex refractive index m corresponding to the conversion band A, and the raindrop spectral parameters of the band to be converted determined in Step 5-2; substitute them into the reflectivity model Z established in Step 2. H and / or differential reflectivity model Z DR Thus, the converted radar parameters are obtained, namely the converted reflectivity Z′. H "and / or converted differential reflectivity Z′ D ′ R ′.

[0019] Step 6, Grid Interpolation: Convert all polar coordinate points in each band to rectangular coordinate points; under the same three-dimensional grid of latitude, longitude and altitude, interpolate the converted radar parameters of all rectangular coordinate points in each band.

[0020] Step 7, Fusion: In the same latitude, longitude, and altitude 3D grid after interpolation in Step 6, when the same grid point has two or more converted reflectivities Z′ H "or converted differential reflectivity Z′ D ′ R When Z′ is used, a distance exponential weighting method is adopted to combine the converted reflectance Z′ of all grid points. H "or converted differential reflectivity Z′ D ′ R To merge.

[0021] In steps 1-2, the raindrop spectrometer collects ground raindrop spectrum observation data N(D) for 2 to 3 years.

[0022] The μ-Λ constraint function established in steps 1-3 is μ=aΛ 2 +bΛ+c; where a, b, and c are all fitting coefficients.

[0023] In step 2, the reflectivity model Z H and / or differential reflectivity model Z DR The expressions are as follows:

[0024]

[0025]

[0026] in:

[0027]

[0028] P = 4π - P′

[0029] In the formula, D is the diameter of the raindrop; P and P′ are both raindrop shape factors, both functions of D; and K is the dielectric constant of the raindrop.

[0030] The expression for the raindrop shape factor P is:

[0031]

[0032] in:

[0033] E 2 =1-r 2

[0034] In the formula, E is an intermediate calculation variable; r is the axial length ratio of the ellipsoidal raindrop, and 0 < r < 1.

[0035] The correction methods in step 4 include attenuation correction, system bias correction, and ground feature filtering.

[0036] In step 4, attenuation correction uses linear correction; system bias correction first determines the system bias using the micro-rain method and the external metal ball method, and then calculates the original reflectivity Z′. H or differential reflectivity Z′ DR Based on this, the determined system bias is directly subtracted; ground feature filtering mainly involves statistically analyzing the intensity texture, differential reflectivity standard deviation, differential propagation phase shift standard deviation, correlation coefficient standard deviation, correlation coefficient, and signal-to-noise ratio distribution of meteorological echoes and non-meteorological echoes, classifying them using fuzzy logic algorithms, and finally identifying and filtering out non-meteorological echoes.

[0037] In step 5-1, the conversion band A is set to the S band.

[0038] In step 7, suppose a certain grid point A has n radars, and each radar corresponds to a set of radar parameters to be fused; where the radar parameters to be fused are the converted reflectivity Z′. H "or converted differential reflectivity Z′ D ′ R Let f be the i-th radar parameter to be fused in grid point A before fusion. A (i), then the radar parameters after fusion of grid points A are denoted as f. A Then f A The expression is:

[0039]

[0040] in:

[0041]

[0042] In the formula, w i d represents the weight value of the i-th radar parameter to be fused at grid point A; i R is the distance from grid point A to the i-th radar; i This sets the maximum detection range from grid point A to the i-th radar.

[0043] R i Set the distance to 100.

[0044] The present invention has the following beneficial effects:

[0045] This invention can fuse multi-band radar data and display and apply it in a single system and map, minimizing the differences in the detection of the same weather target by multi-band radar. This facilitates the application of multi-band weather radar data in the monitoring, early warning and forecasting of severe convective weather, and better leverages the synergistic effect of multi-band radar. Attached Figure Description

[0046] Figure 1 A flowchart of a multi-band weather radar data fusion method according to the present invention is shown.

[0047] Figure 2 The fitted constraint relationship between the shape parameter (μ) and slope parameter (Λ) mentioned in this invention is shown.

[0048] Figure 3 The flowchart of the multi-band weather radar data inversion model mentioned in this invention is shown. Detailed Implementation

[0049] The present invention will now be described in further detail with reference to the accompanying drawings and specific preferred embodiments.

[0050] In the description of this invention, it should be understood that the terms "left side," "right side," "upper part," "lower part," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. "First," "second," etc., do not indicate the importance of the components, and therefore should not be construed as a limitation of this invention. The specific dimensions used in this embodiment are only for illustrating the technical solution and do not limit the scope of protection of this invention.

[0051] like Figure 1 As shown, a multi-band weather radar data fusion method includes the following steps.

[0052] Step 1: Establish a constrained raindrop spectral distribution model, which includes the following steps.

[0053] Step 1-1: Select the raindrop spectral distribution model

[0054] This invention employs a Gamma distribution model to invert the raindrop spectrum N(D) of Doppler weather radar or dual-polarization radar in different bands; wherein, the different bands include X-band, C-band, and S-band; the expression for the raindrop spectrum distribution model N(D) using the Gamma distribution model is as follows:

[0055] N(D) = N0D μ exp(-ΛD)

[0056] In the formula, N0 is the truncation parameter; μ is the shape parameter; Λ is the slope parameter; and D is the diameter of the raindrop particles.

[0057] Step 1-2: Collect raindrop spectrum observation data: For the area to be monitored, use a raindrop spectrometer to collect several sets of raindrop spectrum observation data N(D) for at least 1 year (preferably 2-3 years) on the ground.

[0058] Step 1-3: Establish μ-Λ constraint function: Substitute the several sets of raindrop spectrum observation data N(D) collected in Step 1-2 into the raindrop spectrum distribution model in Step 1-1 to obtain several sets of μ and Λ values.

[0059] By fitting several sets of μ and Λ values, a μ-Λ constraint function is formed, in which μ varies with Λ; wherein, the fitted curve is as follows: Figure 2 As shown; in this embodiment, the μ-Λ constraint function is preferably: μ=aΛ 2 +bΛ+c; where a, b, and c are fitting coefficients. Further, preferably:

[0060] μ = -0.0302Λ 2 +1.0537Λ-2.4526

[0061] That is, a = -0.0302, b = 1.0537, c = -2.4526.

[0062] Subsequently, the μ-Λ constraint function is substituted into the raindrop spectrum distribution model in step 1-1 to form a localized raindrop spectrum distribution model with constraints.

[0063] Step 2: Establish a multi-band weather radar data inversion model: Based on the constrained raindrop spectral distribution model established in Step 1 and the complex refractive index m of precipitation particles, establish a multi-band weather radar data inversion model; the multi-band weather radar data inversion model includes one or two radar parameter models, namely the reflectivity model Z. H and / or differential reflectivity model Z DR The complex refractive index m is determined based on the radar wavelength.

[0064] The above reflectivity model ZH and / or differential reflectivity model Z DR The preferred expressions are as follows:

[0065]

[0066]

[0067] in:

[0068]

[0069] P = 4π - P′

[0070] In the formula, P and P′ are both raindrop shape factors, both functions of D; K is the dielectric constant of the precipitation particles.

[0071] The expression for the raindrop shape factor P is:

[0072]

[0073] in:

[0074] E 2 =1-r 2

[0075] In the formula, E is an intermediate calculation variable; r is the axial length ratio of the ellipsoidal raindrop (i.e., the ratio of the major axis to the minor axis of the ellipsoid), and 0 < r < 1, which is a function of D. It is preferably obtained by fitting the raindrop spectrum observation data collected in steps 1-2. In this embodiment, the preferred function expression is:

[0076] r = 0.9951 + 0.02510D - 0.03644D 2 +0.005030D 3 -0.000249D 4

[0077] For formulas (2-1) and (2-2), since the complex refractive index m is related to the radar wavelength, the reflectivity model Z for the corresponding band can be obtained by substituting the complex refractive index m for different bands. H and / or differential reflectivity model Z DR .

[0078] Step 3, Meteorological Monitoring: At least two band Doppler weather radars or dual-polarization radars are used to conduct meteorological observations on the same area to be monitored; the reflectivity Z′ corresponding to each polar coordinate point is observed by each band Doppler weather radar or dual-polarization radar. H and / or differential reflectivity Z′ DR .

[0079] Step 4, Correcting Radar Parameters (also known as Quality Control): This involves adjusting the Z′ values ​​observed by each band of Doppler weather radar or dual-polarization radar in Step 3. H and / or Z′ DR The corrected radar parameters are obtained by performing corrections on each parameter, namely the corrected reflectivity Z′. H Z′ and / or modified differential reflectivity Z′ D ′ R .

[0080] The aforementioned correction methods preferably include attenuation correction, system bias correction, and ground feature filtering, so that they can reflect the actual weather system conditions.

[0081] The above attenuation correction uses linear correction, and the specific correction is as follows:

[0082] Z H (Corrected) = Z H (Before correction) +PHI DP × Attenuation coefficient (A) h )

[0083] Z DR (Corrected) = Z DR (Before correction) +PHI DP × Attenuation coefficient (A) dp )

[0084] The attenuation coefficient A mentioned above h and A dp All of these results were obtained by fitting the raindrop spectrum observation data collected in steps 1-2.

[0085] The above-mentioned systematic bias correction first determines the systematic bias using the micro-rain method and the external metal ball method, and then calculates the original reflectivity Z′. H or differential reflectivity Z′ DR The determined system deviation is directly subtracted from the base.

[0086] The above-mentioned ground feature filtering mainly involves statistically analyzing the intensity texture, differential reflectivity standard deviation, differential propagation phase shift standard deviation, correlation coefficient standard deviation, correlation coefficient, and signal-to-noise ratio distribution of meteorological echoes and non-meteorological echoes, classifying them using a fuzzy logic algorithm, and finally identifying and filtering out non-meteorological echoes.

[0087] Step 5: Multi-band radar parameter data conversion, such as... Figure 3 As shown, the specific steps include the following.

[0088] Step 5-1: Select and set conversion band A, where A is X-band, C-band, or S-band; if the band to be converted is the same as the set conversion band A, directly update the conversion radar parameters and jump to step 6; otherwise, proceed to step 5-2.

[0089] Currently, the majority of data used in our operations is from S-band weather radar. Therefore, the main task is to convert X / C-band weather radar data to S-band weather radar, that is, to set the preferred conversion band A to S-band. Alternatively, data from any two bands of weather radar can be converted to another band of weather radar data, depending on the user's needs.

[0090] Step 5-2: Determine the raindrop spectral parameters of the band to be converted: Calculate the corrected reflectivity Z′ of the band to be converted. H Substituting the complex refractive index m corresponding to the band to be converted into the reflectivity model Z established in step 2, we can understand the relationship between the band and the complex refractive index m. H Equation 1 is obtained; simultaneously, the corrected differential reflectivity Z′ of the band to be converted is... D ′ R Substituting the complex refractive index m corresponding to the band to be converted into the differential reflectivity model Z established in step 2, we can understand its meaning. DR Equation 2 is obtained from equation 1; by simultaneously solving equations 1, 2 and the μ-Λ constraint function, the raindrop spectral parameters corresponding to the band to be converted are obtained, which are the cutoff parameter N0, shape parameter μ and slope parameter Λ of the converted band.

[0091] The preferred method for solving the cutoff parameter N0, shape parameter μ, and slope parameter Λ of the aforementioned converted band is as follows:

[0092] 1. Obtain the expression for Λ(μ) from the μ-Λ constraint function.

[0093] 2. Substitute the expression for Λ(μ) into the differential reflectance Z. DR In the expression, Z is obtained. DR The relationship between Λ and Λ(Z) DR ).

[0094] 3. Through Λ(Z) DR The relationship is based on the observed Z. DR Find the values ​​of Λ and μ.

[0095] 4. Substitute the calculated values ​​of Λ and μ into the reflectivity factor Z. H In the expression, the value of N0 is obtained.

[0096] Step 5-3, Conversion: Set the complex refractive index m corresponding to the conversion band A, and the raindrop spectral parameters of the band to be converted determined in Step 5-2; substitute them into the reflectivity model Z established in Step 2. H and / or differential reflectivity model Z DR Thus, the converted radar parameters are obtained, namely the converted reflectivity Z′. H " and converted differential reflectivity Z′ D ′ R ′.

[0097] Taking S-band and X-band as examples, if the intensity Z of the weather target detected by the S-band weather radar is... H and / or differential reflectivity Z DR If Z is 50 dB and Z is 2 dB respectively, then the intensity Z of the weather target detected by the X-band weather radar after transformation by the raindrop spectral distribution model in step 1 is... H and / or differential reflectivity Z DR The values ​​are 55 dBZ and 3.2 dB respectively. Without conversion, using the fused results for applications such as radar quantitative precipitation estimation will result in significant errors, particularly in terms of intensity Z. H Taking a strong convection of 50 dBZ as an example, intensity Z H An error of 1 dBZ will introduce approximately 20% bias into the given precipitation estimate.

[0098] Step 6: Grid Interpolation

[0099] A. Coordinate Transformation

[0100] Weather radar scan data is stored in polar coordinates. To obtain spatial grid information, coordinate transformation is required. In this invention, all polar coordinate points in each band are converted to rectangular coordinate points. The three-dimensional grid of the rectangular coordinate points represents longitude, latitude, and altitude.

[0101] When performing coordinate transformation of weather radar echo data, it is necessary to set the spatial resolution. To ensure that the data is not distorted, the spatial resolution is set to an integer multiple of the weather radar range resolution, with the minimum resolution being one detection range unit of the weather radar.

[0102] B. Interpolation: Under the same three-dimensional grid of latitude, longitude and altitude, the converted radar parameters of all rectangular coordinate points in each band are interpolated.

[0103] Step 7, Fusion: In the same latitude, longitude, and altitude 3D grid after interpolation in Step 6, when the same grid point has two or more converted reflectivities Z′ H "or converted differential reflectivity Z′ D ′ R When Z′ is used, a distance exponential weighting method is adopted to combine the converted reflectance Z′ of all grid points. H "or converted differential reflectivity Z′ D ′ R To merge.

[0104] Suppose a grid point A has n radars, each radar corresponding to a set of radar parameters to be fused; where the radar parameters to be fused are the converted reflectivity Z′. H "or converted differential reflectivity Z′ D ′ RLet f be the i-th radar parameter to be fused in grid point A before fusion. A (i), then the radar parameters after fusion of grid points A are denoted as f. A Then f A The expression is:

[0105]

[0106] in:

[0107]

[0108] In the formula, w i d represents the weight value of the i-th radar parameter to be fused at grid point A; i R is the distance from grid point A to the i-th radar; i In this embodiment, to set the maximum detection range from the i-th radar to the grid point A, R i The preferred setting distance is 100.

[0109] This invention is applicable not only to the intensity Z of traditional Doppler weather radar H Fusion is also applicable to the differential reflectivity Z of dual-polarization weather radar. DR Fusion of equal polarization parameters.

[0110] The preferred embodiments of the present invention have been described in detail above. However, the present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention, various equivalent transformations can be made to the technical solutions of the present invention, and these equivalent transformations all fall within the protection scope of the present invention.

Claims

1. A method for fusing multi-band weather radar data, characterized in that: Includes the following steps: Step 1: Establish a constrained raindrop spectral distribution model, which includes the following steps: Step 1-1: Selecting the raindrop spectrum distribution model: The Gamma distribution model is used to invert the raindrop spectrum N(D) of Doppler weather radar or dual polarization radar in different bands; where different bands include X-band, C-band and S-band; the raindrop spectrum N(D) of the Gamma distribution model includes the truncation parameter N0, the shape parameter μ and the slope parameter Λ. Step 1-2: Collect raindrop spectrum observation data: For the area to be monitored, use a raindrop spectrometer to collect several sets of raindrop spectrum observation data N(D) for at least one year. Step 1-3: Establish the μ-Λ constraint function: Substitute the several sets of raindrop spectrum observation data N(D) collected in Step 1-2 into the raindrop spectrum distribution model in Step 1-1 to obtain several sets of μ and Λ values; fit the several sets of μ and Λ values ​​to form a μ-Λ constraint function in which μ varies with Λ; substitute the μ-Λ constraint function into the raindrop spectrum distribution model in Step 1-1 to form a constrained raindrop spectrum distribution model; Step 2: Establish a multi-band weather radar data inversion model: Based on the constrained raindrop spectral distribution model established in Step 1 and the complex refractive index m of precipitation particles, establish a multi-band weather radar data inversion model; the multi-band weather radar data inversion model includes one or two radar parameter models, namely the reflectivity model Z. H and / or differential reflectivity model Z DR The complex refractive index m is determined based on the radar wavelength. Step 3, Meteorological Monitoring: At least two band Doppler weather radars or dual-polarization radars are used to conduct meteorological observations on the same area to be monitored; the reflectivity Z′ corresponding to each polar coordinate point is observed by each band Doppler weather radar or dual-polarization radar. H and / or differential reflectivity Z′ DR ; Step 4: Correct radar parameters: Correct the Z′ values ​​observed by each band of Doppler weather radar or dual-polarization radar in Step 3. H and / or Z′ DR The corrected radar parameters are obtained by performing corrections on each parameter, namely the corrected reflectivity Z′. H Z′ and / or modified differential reflectivity Z′ D ′ R ; Step 5: Multi-band radar parameter data conversion, which includes the following steps: Step 5-1: Select and set conversion band A, where A is X-band, C-band, or S-band; if the band to be converted is the same as the set conversion band A, directly update the conversion radar parameters and jump to step 6; otherwise, proceed to step 5-2. Step 5-2: Determine the raindrop spectral parameters of the band to be converted: Calculate the corrected reflectivity Z′ of the band to be converted. H Substituting the complex refractive index m corresponding to the band to be converted into the reflectivity model Z established in step 2, we can understand the relationship between the band and the complex refractive index m. H Equation 1 is obtained; simultaneously, the corrected differential reflectivity Z′ of the band to be converted is... D ′ R Substituting the complex refractive index m corresponding to the band to be converted into the differential reflectivity model Z established in step 2, we can understand its meaning. DR In the process, we obtain Equation 2; by simultaneously solving Equations 1, Equation 2 and the μ-Λ constraint function, we obtain the raindrop spectral parameters corresponding to the band to be converted, which are the cutoff parameter N0, shape parameter μ and slope parameter Λ of the converted band, respectively. Step 5-3, Conversion: Set the complex refractive index m corresponding to the conversion band A, and the raindrop spectral parameters of the band to be converted determined in Step 5-2; substitute them into the reflectivity model Z established in Step 2. H and / or differential reflectivity model Z DR Thus, the converted radar parameters are obtained, namely the converted reflectivity Z′. H "and / or converted differential reflectivity Z′ D ′ R ′; Step 6, Grid Interpolation: Convert all polar coordinate points in each band to rectangular coordinate points; under the same three-dimensional grid of latitude, longitude and altitude, interpolate the converted radar parameters of all rectangular coordinate points in each band; Step 7, Fusion: In the same latitude, longitude, and altitude 3D grid after interpolation in Step 6, when the same grid point has two or more converted reflectivities Z′ H "or converted differential reflectivity Z′ D ′ R When Z′ is used, a distance exponential weighting method is adopted to combine the converted reflectance Z′ of all grid points. H "or converted differential reflectivity Z′ D ′ R To merge.

2. The multi-band weather radar data fusion method according to claim 1, characterized in that: In steps 1-2, the raindrop spectrometer collects ground raindrop spectrum observation data N(D) for 2 to 3 years.

3. The multi-band weather radar data fusion method according to claim 1, characterized in that: The μ-Λ constraint function established in steps 1-3 is μ=aΛ 2 +bΛ+c; where a, b, and c are all fitting coefficients.

4. The multi-band weather radar data fusion method according to claim 1, characterized in that: In step 2, the reflectivity model Z H Sum of differential reflectivity model Z DR The expressions are as follows: in: P = 4π - P′ In the formula, D is the diameter of the raindrop; P and P′ are both raindrop shape factors, both functions of D; and K is the dielectric constant of the raindrop.

5. The multi-band weather radar data fusion method according to claim 4, characterized in that: The expression for the raindrop shape factor P is: in: E 2 ＝1-r 2 In the formula, E is an intermediate calculation variable; r is the axial length ratio of the ellipsoidal raindrop, and 0 < r < 1.

6. The multi-band weather radar data fusion method according to claim 1, characterized in that: The correction methods in step 4 include attenuation correction, system bias correction, and ground feature filtering.

7. The multi-band weather radar data fusion method according to claim 6, characterized in that: In step 4, attenuation correction uses linear correction; system bias correction first determines the system bias using the micro-rain method and the external metal ball method, and then calculates the original reflectivity Z′. H or differential reflectivity Z′ DR Based on this, the determined system bias is directly subtracted; ground feature filtering mainly involves statistically analyzing the intensity texture, differential reflectivity standard deviation, differential propagation phase shift standard deviation, correlation coefficient standard deviation, correlation coefficient, and signal-to-noise ratio distribution of meteorological echoes and non-meteorological echoes, classifying them using fuzzy logic algorithms, and finally identifying and filtering out non-meteorological echoes.

8. The multi-band weather radar data fusion method according to claim 1, characterized in that: In step 5-1, the conversion band A is set to the S band.

9. The multi-band weather radar data fusion method according to claim 1, characterized in that: In step 7, suppose a certain grid point A has n radars, and each radar corresponds to a set of radar parameters to be fused; where the radar parameters to be fused are the converted reflectivity Z′. H "or converted differential reflectivity Z′ D ′ R Let f be the i-th radar parameter to be fused in grid point A before fusion. A (i), then the radar parameters after fusion of grid points A are denoted as f. A Then f A The expression is: in: In the formula, w i d represents the weight value of the i-th radar parameter to be fused at grid point A; i R is the distance from grid point A to the i-th radar; i This sets the maximum detection range from grid point A to the i-th radar.

10. The multi-band weather radar data fusion method according to claim 9, characterized in that: R i Set the distance to 100.

Images