Method for inverting raindrop spectrum in air by using characteristic parameters of dual-polarization radar

Abstract

The invention relates to the technical field of raindrop spectrum inversion, in particular to a method for inverting an aerial raindrop spectrum by using characteristic parameters of a dual-polarization radar, and the method comprises the steps of data collection, quality control, calculation and fusion. Modeling and counting; the statistics specifically comprises the statistics of a radar and rainfall type relationship and the statistics of a typing and raindrop spectrum type fitting relationship; and the business application process comprises a four-dimensional raindrop spectrum product and a rainfall estimation product. The research of the invention can bring significant improvements in three aspects: in the aspect of monitoring capability, a radar secondary product system can be enriched, and visual visualization of spatial precipitation particle distribution characteristics is realized; in the aspect of early warning and forecasting, the heavy rainfall falling area can be researched and judged in advance through the distribution characteristics of particles in the air, and the disastrous weather early warning timeliness is effectively prolonged; in the aspect of artificial influence weather, refined rainfall micro-physical structure information is provided, and reliable technical support is provided for operation condition identification and effect evaluation.

Description

Technical Field

[0001] This invention relates to the field of raindrop spectrum inversion technology, specifically a method for inverting the airborne raindrop spectrum using characteristic parameters of a dual-polarization radar. Background Technology

[0002] The atmospheric precipitation particle spectrum typically refers to the size distribution of precipitation particles such as raindrops, snowflakes, and ice pellets in the atmosphere. Changes in the atmospheric precipitation particle spectrum are the most direct description of the complex interactions between precipitation particles in clouds. Simultaneously, as the most direct observation target of weather radar, precipitation particles are subject to varying sensitivities for different sizes of raindrops due to limitations in radar detection principles. These differences in the distribution of the precipitation particle spectrum directly affect the accuracy of radar precipitation estimation. However, current technologies cannot accurately obtain inversion parameters of ground and atmospheric microphysical parameters, meaning they cannot understand the formation, development, and variation mechanisms of atmospheric cloud precipitation, thus hindering the accurate prediction and interpretation of atmospheric processes and meteorological disasters.

[0003] In existing technologies, there are published articles on inverting raindrop spectra using dual-polarization radar. However, these mainly involve using raindrop spectrometer parameters to invert dual-polarization radar parameters and calibrating the radar's observation parameters. There are very few articles on inverting raindrop spectra using dual-polarization radar parameters because the inversion process involves significant errors, making it difficult to implement in operational applications. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method for retrieving aerial raindrop spectra using the characteristic parameters of a dual-polarization radar. This method leverages the full-domain observation capability of dual-polarization radar and the ability of its characteristic parameters to reflect raindrop size and concentration to retrieve aerial raindrop spectra in real time.

[0005] To achieve the above objectives, the following technical solution is provided:

[0006] A method for inverting the airborne raindrop spectrum using characteristic parameters of a dual-polarization radar, characterized by comprising the following steps:

[0007] Step 1: Data collection, quality control, calculation, and fusion; specifically including the following steps:

[0008] Step 1: Data preparation; collect rainfall intensity statistics from national weather stations, with hourly resolution;

[0009] Step 2: Read raindrop spectrum data and calculate three basic parameters; collect raindrop spectrum baseline data corresponding to the precipitation process, perform three rounds of quality control on the data, and use the intercept method to calculate the three basic parameters of the raindrop spectrum (0 and 1 are respectively: N (intercept parameter), reflecting the concentration of small droplets; u (shape parameter), controlling the steepness of the spectrum shape; Λ (slope parameter), determining the attenuation rate of large droplets, which is related to the precipitation intensity); the resolution is 1 minute.

[0010] Step 3: Read the radar base data from Tianjin Baodi Radar Station (Z9025) and perform quality control; locate the positions of 13 national stations (corresponding latitude and longitude) on the radar map, extract the corresponding positions and surrounding 3*3=9 pixels, and perform REF (radar reflectivity factor) and Zdr (differential reflectivity factor) layer by layer, with a time resolution of 6 minutes; if the Zdr data quality is poor, it needs to be quality controlled: first, overall offset, the Zdr of individual precipitation cases should be greater than 0, drizzle is concentrated near 0, so the overall Zdr is increased by 1; second, range calibration, remove values ​​less than -2 and greater than 5; third, smoothing, smooth the Zdr in different ways according to different REF values;

[0011] Step 4: Correspondence between raindrop spectrum and radar data; make spatiotemporal correspondences between the raindrop spectrum three parameters and radar parameters corresponding to the previously output precipitation time period and national station location, and unify them into a large table. The table includes station, latitude and longitude, time, raindrop spectrum three parameters, radar REF and ZDR, and the time resolution is unified to 6 minutes.

[0012] Step 2, Modeling and Statistics; specifically including the following steps:

[0013] Step 1: Statistical analysis of the relationship between radar and precipitation type

[0014] The purpose of precipitation type is to determine the precipitation type based on radar products during operational applications, and then select an appropriate model to calculate the three basic parameters of the raindrop spectrum based on the precipitation type.

[0015] The way precipitation is classified should be related to the spectral shape of raindrops, so it cannot be classified solely based on the precipitation magnitude. To avoid the problem of many small drops and a few large drops having the same precipitation magnitude but inconsistent spectral shapes, both particle size and concentration should be considered when classifying precipitation. Therefore, precipitation is classified into 9 types based on different thresholds of REF and Zdr.

[0016] The REF nodes are 40 and 50, and the Zdr nodes are 1 and 2. The nodes are selected using a statistical method to avoid having too few cases within the interval.

[0017] Step 2: Statistical analysis of the fitting relationship between classification and raindrop spectral type

[0018] Based on the nine classifications, the fitting relationship between raindrop spectral parameters and radar parameters in different classifications was obtained using a large table.

[0019] The relationship between Λ and u, and the good fitting effect of Λ-10log(REF / N0);

[0020] The value of Λ is related to Zdr, so quality control of Zdr is particularly important. Only by obtaining an accurate Λ can the other two parameters be obtained accurately.

[0021] Λ has no type, the other two parameters are type 9;

[0022] Step 3, Business Application Process, including 4D raindrop spectrum products and precipitation estimation products; details are as follows:

[0023] In the four-dimensional raindrop spectrum product, the input source is real-time and past 2-hour radar parameters. Radar 0-2km corresponds to actual 1km, radar 1-3km corresponds to actual 2km, and radar 3-5km corresponds to actual 3km. First, find the radar parameters at the corresponding height position. Each grid point with high spatial resolution corresponds to 3 sets of REF and Zdr at 3 heights. For each height of each point, precipitation is classified using REF and Zdr. For different classifications, three raindrop spectrum parameters are obtained. Various raindrop spectrum parameters are calculated based on these three parameters, including median particle size, median particle size concentration, and mass-weighted diameter. In this way, each spatial point corresponds to the basic raindrop spectrum parameters and secondary products at three heights. Clicking on a location on the plane displays three unit volumes at 1km, 2km, and 3km, filled with particles. The particle size and distribution can be seen, with different particle sizes represented by different colors. Clicking on any volume displays the time curves of several raindrop spectrum products within that volume over the past 2 hours, showing the magnitude and changes.

[0024] In the precipitation estimation product, the input source is real-time radar parameters. The radar parameters corresponding to the 1km height position are found. Each grid point with high spatial resolution corresponds to a set of REF and Zdr. The three parameters of the raindrop spectrum (for verification) and the 1h precipitation estimation product (for verification and operation) are generated. Different levels are filled with color. The rainfall intensity is calculated based on the raindrop spectrum of each grid point at the 1km height. The color is filled with stability, non-short-term convective intensity and short-term intensity. Short-term intensity is determined by precipitation amount, and whether there is convection is determined by echo intensity.

[0025] The beneficial effects of this invention are as follows:

[0026] 1. The data source required by this invention is only dual-polarization radar. The data is singular and highly operable, and there is no need for dependencies between data, thus reducing errors.

[0027] 2. This invention fully leverages the advantage of dual-polarization radar in capturing particle shape characteristics. By establishing a quantitative relationship model between dual-polarization parameters and ground-based raindrop spectrometers, it achieves real-time inversion of three-dimensional raindrop spectral parameters, thereby obtaining key information such as median particle size, particle number concentration, and precipitation intensity.

[0028] 3. The research of this invention will bring about three significant improvements: In terms of monitoring capabilities, it can enrich the radar secondary product system and realize intuitive visualization of spatial precipitation particle distribution characteristics; in terms of early warning and forecasting, the distribution characteristics of airborne particles can be used to predict the location of heavy precipitation in advance, effectively extending the warning time of severe weather; in terms of artificial weather modification, it can provide refined precipitation microphysical structure information, providing reliable technical support for the identification of operational conditions and effect evaluation.

[0029] 4. This invention employs a 9-type classification method based on radar parameters every 6 minutes, corresponding to 9 different spectral types. Compared to traditional methods that classify based on precipitation or do not classify at all, this method avoids grouping large droplets with few large droplets into the same category, as their corresponding spectral types should be different. Furthermore, by using a 6-minute time unit for classification, it avoids the situation where traditional methods based on hourly rainfall intensity result in significant variations in precipitation particle concentration and size within an hour. The derived convective parameters are more accurate.

[0030] 5. The significance of calculating the precipitation estimation product in this invention is to compare it with the radar QPE. It should be better than QPE and objectively improve the precipitation estimation level. However, it is reasonable for there to be errors with the actual situation.

[0031] 6. This invention uses a four-dimensional visualization method to present the distribution of raindrop spectra in real time, which is equivalent to expanding the types of radar products. The product has been created from scratch, transforming the "abstract" cloud particles into the "concrete" and the "invisible" into the "visible". Combined with statistical tables, it improves the understanding and subjective prediction of the physical state of precipitation clouds and increases the lead time for forecasts and warnings. Attached Figure Description

[0032] Figure 1 This is a schematic diagram of the process of the present invention;

[0033] Figure 2 This is a comparison chart of the aerial quantitative precipitation estimation product based on dual-polarization radar inversion and the actual precipitation in this embodiment of the invention.

[0034] Figure 3 This is a graph showing the fitting relationship between the radar's basic reflectivity Ref at a 0.5° elevation angle and the basic parameter N0 of the precipitation droplet spectrum under different precipitation types in this embodiment of the invention.

[0035] Figure 4 This is a graph showing the fitting relationship between the basic reflectivity Ref of the radar at a 0.5° elevation angle and the basic parameter Λ of the precipitation droplet spectrum in an embodiment of the present invention.

[0036] Figure 5 This is a graph showing the fitting relationship between the basic parameters Λ and u of the raindrop spectrum under different precipitation types in an embodiment of the present invention.

[0037] Figure 6 This is a schematic diagram of radar echoes according to an embodiment of the present invention;

[0038] Figure 7 This is a schematic diagram of quantitative precipitation estimation according to an embodiment of the present invention;

[0039] Figure 8 This is a reflectivity diagram of dual-polarization radar combination between different height layers according to an embodiment of the present invention;

[0040] Figure 9 This is a visualization of the aerial three-dimensional raindrop spectrum based on dual-polarization radar, according to an embodiment of the present invention.

[0041] Figure 10 This is a detailed diagram of raindrops in an embodiment of the present invention. Detailed Implementation

[0042] 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 of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0043] The following detailed description of the construction process of the editing system of the present invention, with reference to the accompanying drawings, and further illustrates the usage method and experimental results of the editing system of the present invention through the embodiments.

[0044] Example 1

[0045] A method for inverting the airborne raindrop spectrum using characteristic parameters of dual-polarization radar, such as Figure 1 As shown, it includes the following steps:

[0046] Step 1: Data preprocessing, which includes the following steps:

[0047] Step 1: All periods of precipitation recorded at national weather stations in 2024.

[0048] 1450 individual hourly precipitation data points were collected from national weather stations from April to October 2024;

[0049] station_id datetime pre_1h type 0 54525 2024-8-9 17:00 106.3 Short and strong 1 54525 2024-7-30 10:00 42.9 Short and strong 2 54525 2024-8-23 11:00 42.6 Short and strong 3 54525 2024-7-24 14:00 31.0 Short and strong 4 54525 2024-7-21 21:00 24.3 Short and strong ... ... ... ... ... 1535 54428 2024-7-30 0:00 7.9 Stablize 1536 54525 2024-8-9 14:00 7.9 Stablize 1537 54526 2024-8-9 19:00 7.9 Stablize 1538 54528 2024-8-19 8:00 7.9 Stablize 1539 54645 2024-9-20 0:00 9.2 Stablize

[0050] Step 2: Extract the raindrop spectrum data for the corresponding time period and calculate the three basic parameters of the minute-by-minute raindrop spectrum;

[0051] The raindrop spectrum data was processed, and 12,094 valid data points per minute were extracted.

[0052] time station_id mu lambda N0 particle_counts 0 2024-05-03 00:33:00 54525 15 3.994337388911037 0.007162196349675804 11 1 2024-05-04 09:35:00 54528 15 7.0638120849675964 1964.92044258565795 13 2 2024-05-05 22:40:00 54527 -0.6707651067397445 1.4546010460205 975.2110501159922 844 3 2024-05-05 22:09:00 54528 3.8704232605751566 5.244107223209057 1657.0314932313852 1283 4 2024-05-05 22:12:00 54528 1.9084385413018723 2.7534689744643134 11037.030219826904 1932 ... ... ... ... ... ... ... 12089 2024-09-30 01:46:00 54526 3.1306859525534122 3.5425244008737106 32119.486120679623 1901 12090 2024-09-30 01:02:00 54527 -1.1220214795277892 791.834838960609 896 12091 2024-09-30 03:10:00 54627 0.3949582501574515 1.239426258972459 153.48177339854 836 12092 2024-09-30 03:40:00 54645 3.0310129804116829 2.4608320916301816 6026.5441778834875 1553 12093 2024-09-30 03:41:00 54645 1.94278506698207 3.2049473654437708 17821.763099495424 1299

[0053] Step 3: Download 5227 dual-polarization radar data points from station Z9025 within the past hour based on the individual case time.

[0054] Step 4: Download 1540 files of surface minute precipitation data from the past hour based on the individual case time. Each file contains 60 data points, for a total of 92400 data points.

[0055] Step 5: Merge the above data into 1-minute data sets, totaling 92,394 records. The radar ref and zdr extraction criteria are the maximum ref value and the corresponding zdr value within a 3x3 grid within a 1km radius above the station.

[0056] Step 2: Correspondence between radar parameters and precipitation types

[0057] Zdr\ref ref<40dBz ref40~50dBz ref>50dBz Hello<1 Type I Type II Type III 1<Zdr<2 Type IV Type V Type VI Zdr>2 Type VII Type VIII Type IX

[0058] like Figure 2 As shown, the origin represents the precipitation estimated by airborne quantitative precipitation based on dual-polarization radar, and the triangle represents the actual precipitation at the corresponding location. The good correspondence between the two indicates that the inversion accuracy of the airborne quantitative precipitation estimation product based on dual-polarization inversion is high.

[0059] Step 3: Correspondence between precipitation type and raindrop spectral parameters.

[0060] ① Calculate N0 (the relationship between the genotypic statistical reflectance REF and N0, where N0 is the genotypic reflectance). Figure 3 As shown;

[0061] lg(REF_N0)(lambda) = -0.0726 + -0.106lambda + -2.2812lambda**2

[0062] lg(REF_N0)(lambda) = -0.0971 + -0.1004lambda + -2.3962lambda**2

[0063] lg(REF_N0)(lambda) = -0.2142 + -0.1837lambda + -1.6004lambda**2

[0064] lg(REF_N0)(lambda) = -0.166 + -0.2197lambda + -2.3848lambda**2

[0065] lg(REF_N0)(lambda) = -0.2004 + -0.5533lambda + -2.7293lambda**2

[0066] lg(REF_N0)(lambda) = -0.1143 + -0.1829lambda + -2.1153lambda**2

[0067] lg(REF_N0)(lambda) = -0.0989 + -0.192lambda + -2.3612lambda**2

[0068] lg(REF_N0)(lambda) = -0.0985 + -0.4748lambda + -2.349lambda**2

[0069] ② Calculate Λ, such as Figure 4 As shown;

[0070] Λ=2.175*1.075 38-ref1 +0.48

[0071] ③ Calculate u (the relationship between fractal statistics Λ and u, u), such as Figure 5 As shown;

[0072] lambda(mu) = 0.3683 + 0.396mu ​​+ 0.0046mu**2

[0073] lambda(mu) = 0.3357 + 1.3117mu + 1.9643mu**2

[0074] lambda(mu) = 0.244 + 0.2759mu + 0.0027mu**2

[0075] lambda(mu) = 0.6178 + 0.6588mu + 0.4181mu**2

[0076] lambda(mu) = 0.3872 + 0.4777mu + 0.2715mu**2

[0077] lambda(mu) = 0.186 + 1.2666mu + 1.4424mu**2

[0078] lambda(mu) = 0.3421 + 0.9135mu + 1.1377mu**2

[0079] lambda(mu) = 0.288 + 0.3139mu + -0.1737mu**2

[0080] 4. Calculation of secondary product parameters

[0081] Median particle size:

[0082]

[0083] concentration:

[0084] N(D) = N0D u e -ΛD

[0085] Mass-weighted diameter:

[0086]

[0087] Precipitation rate:

[0088]

[0089] Step 4, Business Application 1 – Precipitation Estimation Products

[0090] Selecting radar parameters at a height of 1km, and based on the precipitation classification, obtaining the basic parameters of the raindrop spectrum for each pixel, the precipitation estimation product is calculated. Figure 6 For radar echo, Figure 7 (for quantitative precipitation estimation)

[0091] Business Application 2 ---- Four-Dimensional Raindrop Spectrum Product

[0092] The four-dimensional raindrop spectrum is displayed in three-dimensional space and one-dimensional time.

[0093] The three-dimensional plane has a resolution of 250m and heights of 1km, 2km, and 3km. The time dimension is from the present to the past 2 hours; the maximum REF values ​​at heights of 0-2km, 1-3km, and 3-5km, and the corresponding ZDR inversion parameters at 1km, 2km, and 3km are selected respectively.

[0094] Real-time datasets can be provided to the company, including time, latitude and longitude, altitude, and raindrop spectra; such as Figure 8 As shown.

[0095] Regarding the business platform: It supports freely selecting point A on a plane; three cubes will pop up, representing the raindrop spectrum visualization per unit volume at 1km, 2km, and 3km above point A; the number of particles displayed can be experimentally determined, with default values ​​for the median, 25th, and 75th percentiles, which can be increased or decreased as needed; in terms of time, one of the three cubes can be selected to display the past 2 hours' changes in the median particle size, concentration, and mass-weighted diameter of the raindrop spectrum at the current location; for example... Figure 9 As shown.

[0096] Regarding the protection platform: For the rain phase recognition portion, different effects are displayed depending on the rain intensity; hovering the mouse over it brings up a visual display of raindrops; clicking again brings up the time series; for example... Figure 10 As shown.

[0097] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for inverting the airborne raindrop spectrum using characteristic parameters of a dual-polarization radar, characterized in that, Includes the following steps: Step 1: Data collection, quality control, calculation, and fusion; Step 2: Modeling and Statistics; specifically including statistics on the relationship between radar and precipitation type, and statistics on the fitting relationship between classification and raindrop spectral type; Step 3: Business application process, including four-dimensional raindrop spectrum products and precipitation estimation products.

2. The method for inverting airborne raindrop spectra using characteristic parameters of dual-polarization radar according to claim 1, characterized in that, In step one, the data collection, quality control, calculation, and fusion include the following steps: Step 1: Data preparation; collect rainfall intensity statistics from national weather stations, with hourly resolution; Step 2: Read raindrop spectrum data and calculate three basic parameters; collect raindrop spectrum baseline data corresponding to the precipitation process, perform three rounds of quality control on the data, and calculate the three basic parameters of the raindrop spectrum using the intercept method. The three basic parameters are: intercept parameter N, which reflects the concentration of small droplets; shape parameter u, which controls the steepness of the spectrum shape; and slope parameter Λ, which determines the attenuation rate of large droplets and is related to the precipitation intensity. The resolution is 1 minute. Step 3: Read the radar base data from the Tianjin Baodi radar station and perform quality control; locate the national station on the radar map, extract the layer-by-layer radar reflectivity factor (REF) and differential reflectivity factor (Zdr) of the corresponding location and surrounding 3*3=9 pixels, with a time resolution of 6 minutes; if the Zdr data quality is poor, it needs to be quality controlled: first, overall offset, the Zdr of individual precipitation cases should be greater than 0, and drizzle is concentrated near 0, so the overall Zdr is increased by 1; second, range calibration, removing values ​​less than -2 and greater than 5; third, smoothing, smoothing Zdr in different ways according to different REF values; Step 4: Correspondence between raindrop spectrum and radar data; The raindrop spectrum parameters and radar parameters corresponding to the previously output precipitation periods and national station locations are mapped in time and space into a single large table. The table includes station, latitude and longitude, time, raindrop spectrum parameters, radar REF and ZDR, with a time resolution of 6 minutes.

3. The method for inverting airborne raindrop spectra using characteristic parameters of dual-polarization radar according to claim 2, characterized in that, In step two, the purpose of the statistical analysis of the relationship between radar and precipitation type is to determine the precipitation type based on the radar product during operational applications, and then select an appropriate model to calculate the three basic parameters of the raindrop spectrum based on the precipitation type. The precipitation classification method should be related to the spectral shape of the raindrop spectrum. When classifying, particle size and concentration should be considered simultaneously. Therefore, precipitation is classified into 9 types based on different thresholds of REF and Zdr.

4. The method for inverting airborne raindrop spectra using characteristic parameters of dual-polarization radar according to claim 3, characterized in that, In the precipitation type, the REF nodes are 40 and 50, and the Zdr nodes are 1 and 2.

5. The method for inverting airborne raindrop spectra using characteristic parameters of dual-polarization radar according to claim 3, characterized in that, In step two, in the statistical analysis of the fitting relationship between the classification and the raindrop spectrum, based on the nine classifications, a large table is used to obtain the fitting relationship between the raindrop spectrum parameters and the radar parameters in different classifications: the relationship between Λ and u and the fitting effect of Λ-10log(REF / N0) are good; the value of Λ is related to Zdr, and only by obtaining an accurate Λ can the other two parameters be obtained accurately; Λ has no classification, and the other two parameters are of type nine.

6. The method for inverting airborne raindrop spectra using characteristic parameters of dual-polarization radar according to claim 5, characterized in that, In step three, the input source for the four-dimensional raindrop spectrum product is the radar parameters in real time and over the past 2 hours. Radar 0-2km corresponds to the actual 1km, radar 1-3km corresponds to the actual 2km, and radar 3-5km corresponds to the actual 3km. First, find the radar parameters at the corresponding height position. Each grid point with high spatial resolution corresponds to 3 sets of REF and Zdr at 3 heights. For each height of each point, use REF and Zdr to classify the precipitation. Obtain the three raindrop spectrum parameters for different classifications. Calculate various raindrop spectrum parameters based on the three parameters, including median particle size, median particle size concentration, and mass-weighted diameter. In this way, each spatial point corresponds to the basic raindrop spectrum parameters and secondary products at three heights. Clicking on a position on the plane displays three unit volumes at 1km, 2km, and 3km. The volumes are filled with particles, showing the particle size and distribution. Different particle sizes are represented by different colors. Clicking on any volume will display the time curves of several raindrop spectrum products within that volume over the past 2 hours, showing the magnitude and changes.

7. The method for inverting airborne raindrop spectra using characteristic parameters of dual-polarization radar according to claim 5, characterized in that, In step three, the input source for the precipitation estimation product is real-time radar parameters. The radar parameters corresponding to the 1km height location are found. Each grid point with high spatial resolution corresponds to a set of REF and Zdr. The three parameters of the raindrop spectrum and the 1h precipitation estimation product are generated and filled with different levels of color. The rainfall intensity is calculated based on the raindrop spectrum of each grid point at the 1km height. The color is filled with stability, non-short-term convective intensity, and short-term intensity. Short-term intensity is determined by precipitation amount, and whether there is convection is determined by echo intensity.

Images