A raindrop particle size inversion method based on doppler velocity spectrum and a light rain radar system

Abstract

The application discloses a raindrop particle diameter inversion method based on Doppler velocity spectrum and a light rain radar system, and particularly relates to the field of radar detection technology, and is used for solving the systematic deviation problem caused by environmental disturbance and phase state difference when the existing radar system inverts raindrop particle diameter; the Doppler velocity spectrum is obtained through vertical detection of a frequency-modulated continuous wave radar, the initial raindrop particle diameter distribution is obtained through inversion based on the physical relationship between the terminal falling speed of raindrops and the diameter, the atmospheric temperature and humidity profiles measured by a ground-based microwave radiometer and the cloud vertical structure information measured by a laser radar are synchronously obtained, the physical mutation point of the initial raindrop particle diameter distribution on the vertical profile is diagnosed based on the atmospheric temperature and humidity profiles and the cloud vertical structure information, the constraint weight of the corresponding auxiliary data is inhibited at the height layer where the physical mutation point is located, and the adjacent height layer information is mainly used for fusion correction to obtain the corrected raindrop particle diameter distribution, and finally the corrected raindrop particle diameter distribution is output, so that the inversion precision is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of radar detection, more particularly, the present application relates to a raindrop particle size inversion method based on Doppler velocity spectrum and a micro rain radar system. BACKGROUND

[0002] In the field of radar meteorological detection, vertical detection can be performed using a frequency-modulated continuous wave radar, and atmospheric parameters can be inverted by analyzing the Doppler velocity spectrum of the precipitation particle echo. In the prior art, such a radar system can achieve preliminary inversion of the raindrop spectrum distribution by establishing a theoretical relationship between the terminal falling velocity of the raindrop and the particle size.

[0003] However, the existing radar system has inherent technical bottlenecks in realizing high-precision raindrop spectrum inversion. Due to the influence of precipitation attenuation, environmental disturbance and the difference in scattering characteristics of different hydrometeors in the propagation path of the radar signal, systematic deviations exist between the Doppler velocity spectrum obtained based on the original echo data and the true physical parameters, which will further be transmitted to the inversion result of the raindrop particle size distribution, thereby reducing the reliability and accuracy of the detection data of the radar system under complex weather conditions. SUMMARY

[0004] In order to overcome the above-mentioned defects of the prior art, the present application provides a raindrop particle size inversion method based on Doppler velocity spectrum and a micro rain radar system to solve the problems raised in the background art.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0006] A raindrop particle size inversion method based on Doppler velocity spectrum, comprising the following steps:

[0007] S1, vertical detection of a precipitation area is performed by a frequency-modulated continuous wave radar to obtain the Doppler velocity spectrum of the precipitation particle;

[0008] S2, the Doppler velocity spectrum is inverted based on the physical relationship between the terminal falling velocity of the raindrop and the diameter to obtain an initial raindrop particle size distribution;

[0009] S3, atmospheric temperature and humidity profiles measured by a ground-based microwave radiometer and cloud vertical structure information measured by a laser radar are synchronously obtained;

[0010] S4, based on the atmospheric temperature and humidity profiles and the cloud vertical structure information, a physical mutation point of the initial raindrop particle size distribution in the vertical profile is diagnosed;

[0011] S5, the constraint weight of the corresponding auxiliary data is suppressed at the height layer where the physical mutation point is located, and the initial raindrop particle size distribution is fused and corrected based on the information of the adjacent height layer to obtain a corrected raindrop particle size distribution;

[0012] S6, output the corrected raindrop size distribution.

[0013] Further, the precipitation area is vertically detected by the frequency-modulated continuous wave radar to obtain a Doppler velocity spectrum of precipitation particles, including:

[0014] The frequency-modulated continuous wave radar transmits a continuous wave signal and receives a scattered echo signal of the precipitation particles;

[0015] The echo signal is mixed to obtain a difference frequency signal;

[0016] The difference frequency signal is subjected to spectrum analysis to generate a power spectral density distribution;

[0017] Based on the power spectral density distribution, signal intensity of the precipitation particles in different radial velocity intervals is determined to form the Doppler velocity spectrum.

[0018] Further, the Doppler velocity spectrum is inverted based on the physical relationship between the terminal falling velocity of the raindrop and the diameter to obtain an initial raindrop size distribution, including:

[0019] A mapping relationship between the Doppler velocity and the raindrop size is established according to the physical relationship between the terminal falling velocity of the raindrop and the diameter;

[0020] Based on the mapping relationship between the Doppler velocity and the raindrop size, the signal intensity in the Doppler velocity spectrum is converted into a number concentration corresponding to a raindrop size interval;

[0021] The number concentrations of all raindrop size intervals are integrated to form the initial raindrop size distribution.

[0022] Further, the atmospheric temperature and humidity profiles measured by the ground-based microwave radiometer and the cloud vertical structure information measured by the laser radar are synchronously obtained, including:

[0023] The temperature profile and the humidity profile output by the ground-based microwave radiometer are synchronously received through a data acquisition interface to form the atmospheric temperature and humidity profile;

[0024] The backscattering signal vertical profile measured by the laser radar is received through the data acquisition interface at the same time, and the cloud bottom height and the cloud layer thickness are identified based on the backscattering signal vertical profile to generate the cloud vertical structure information;

[0025] The atmospheric temperature and humidity profile and the cloud vertical structure information are time-synchronously associated.

[0026] Further, based on the atmospheric temperature and humidity profile and the cloud vertical structure information, a physical mutation point of the initial raindrop size distribution in the vertical profile is diagnosed, including:

[0027] The vertical distribution of liquid water content is calculated according to the atmospheric temperature and humidity profile;

[0028] Identify the height of the phase change layer in combination with the cloud vertical structure information;

[0029] Based on the vertical distribution of liquid water content and the height of the phase change layer, gradient analysis is performed on the initial raindrop size distribution in the vertical direction;

[0030] Through gradient analysis, the physical mutation point of the initial raindrop size distribution in the vertical profile is located.

[0031] Further, based on the vertical distribution of liquid water content and the height of the phase change layer, gradient analysis is performed on the initial raindrop size distribution in the vertical direction, including: calculating the vertical profile representing the total particle concentration based on the initial raindrop size distribution; comparing the vertical profile representing the total particle concentration with the vertical distribution of liquid water content near the phase change layer in terms of their cooperative change trend; and locating the physical mutation point of the initial raindrop size distribution in the vertical profile according to the significant deviation characteristics of the cooperative change trend.

[0032] Further, the constraint weight of the corresponding auxiliary data is suppressed at the height layer where the physical mutation point is located, and the initial raindrop size distribution is fused and corrected based on the information of adjacent height layers to obtain the corrected raindrop size distribution, including:

[0033] Determine the constraint weight coefficient of the atmospheric temperature and humidity profile and the cloud vertical structure information corresponding to the height layer where the physical mutation point is located;

[0034] Reduce the constraint weight coefficient of the height layer where the physical mutation point is located;

[0035] Based on the initial raindrop size distribution of adjacent height layers, an alternative distribution of the height layer where the physical mutation point is located is calculated by interpolation;

[0036] Fuse the alternative distribution with the atmospheric temperature and humidity profile and the cloud vertical structure information after weight adjustment to generate the corrected raindrop size distribution.

[0037] Further, based on the initial raindrop size distribution of adjacent height layers, an alternative distribution of the height layer where the physical mutation point is located is calculated by interpolation, including: selecting the two adjacent height layers above and below the height layer where the physical mutation point is located as reference layers; based on the initial raindrop size distribution data of the two reference layers, constructing a continuous physical change model of the particle size distribution in the vertical direction; and calculating the alternative distribution of the height layer where the physical mutation point is located through the continuous physical change model.

[0038] Further, the corrected raindrop size distribution is output, including:

[0039] Convert the corrected raindrop size distribution into a vertical distribution data product in a preset data format;

[0040] The vertical distribution data product is transmitted to a display device through a data interface to realize real-time graphical display, and the vertical distribution data product is stored in a non-volatile memory to form a historical data set.

[0041] In another aspect, the application provides a micro rain radar system, comprising:

[0042] A frequency-modulated continuous wave radar is used for vertical detection of a precipitation area to obtain a Doppler velocity spectrum of precipitation particles.

[0043] A processing device is connected with the frequency-modulated continuous wave radar, and is used for inverting the Doppler velocity spectrum to obtain an initial raindrop particle size distribution based on a physical relationship between terminal falling velocity and diameter of raindrops, synchronously obtaining an atmospheric temperature and humidity profile measured by a ground-based microwave radiometer and cloud vertical structure information measured by a laser radar, diagnosing a physical mutation point of the initial raindrop particle size distribution on a vertical profile based on the atmospheric temperature and humidity profile and the cloud vertical structure information, suppressing a constraint weight of corresponding auxiliary data at a height layer where the physical mutation point is located, and performing fusion correction on the initial raindrop particle size distribution based on information of adjacent height layers to obtain a corrected raindrop particle size distribution.

[0044] An output device is connected with the processing device, and is used for converting the corrected raindrop particle size distribution into a vertical distribution data product in a preset data format and outputting the vertical distribution data product.

[0045] The frequency-modulated continuous wave radar, the processing device and the output device are sequentially connected to form a data processing link.

[0046] Compared with the prior art, the application has the following beneficial effects:

[0047] 1. Through multi-source data fusion and vertical profile correction mechanism, the raindrop particle size distribution inversion precision is significantly improved in the field of radar meteorological detection. Firstly, the atmospheric temperature and humidity profile of the ground-based microwave radiometer and the cloud vertical structure information of the laser radar are synchronously obtained to construct a multi-dimensional atmospheric environmental parameter field, so that the radar inversion process can effectively fuse the physical characteristics of different height layers. The initial raindrop particle size distribution is diagnosed based on the atmospheric temperature and humidity profile and the cloud vertical structure information, which can accurately identify the physical mutation points caused by phase change, environmental disturbance and other factors, and break through the technical limitation of systematic deviation caused by the lack of environmental parameters in traditional single radar inversion.

[0048] 2. By implementing the constraint weight dynamic adjustment strategy at the height layer where the physical mutation point is located, and correcting based on the information of adjacent height layers, the influence of abnormal data on the inversion result is effectively suppressed. The weight self-adaptive mechanism based on physical characteristics not only retains the vertical resolution advantage of radar detection, but also overcomes the defects of unstable inversion results of traditional methods under complex weather conditions. The corrected raindrop size distribution not only maintains physical consistency in the vertical direction, but also significantly improves the reliability and accuracy of the radar system in the detection of precipitation microphysical processes. BRIEF DESCRIPTION OF DRAWINGS

[0049] Figure 1 A flowchart of a raindrop size inversion method based on a Doppler velocity spectrum according to the present application;

[0050] Figure 2 A structural schematic diagram of a light rain radar system according to the present application. DETAILED DESCRIPTION

[0051] The technical solutions in the embodiments of the present application will be described clearly and completely in combination with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0052] Embodiment 1: Figure 1 A raindrop size inversion method based on a Doppler velocity spectrum according to the present application is given, which comprises the following steps:

[0053] S1. Vertical detection of a precipitation area is performed by a frequency-modulated continuous wave radar to obtain a Doppler velocity spectrum of precipitation particles;

[0054] S2. The Doppler velocity spectrum is inverted based on the physical relationship between the terminal falling velocity of raindrops and the diameter to obtain an initial raindrop size distribution;

[0055] S3. Atmospheric temperature and humidity profiles measured by a ground-based microwave radiometer and cloud vertical structure information measured by a laser radar are synchronously obtained;

[0056] S4. Physical mutation points of the initial raindrop size distribution on the vertical profile are diagnosed based on the atmospheric temperature and humidity profiles and the cloud vertical structure information;

[0057] S5. The constraint weight of the corresponding auxiliary data is suppressed at the height layer where the physical mutation point is located, and the initial raindrop size distribution is corrected based on the information of adjacent height layers to obtain a corrected raindrop size distribution;

[0058] S6. The corrected raindrop size distribution is output.

[0059] S1, vertical detection of the precipitation area is performed by a frequency-modulated continuous wave radar, and a Doppler velocity spectrum of the precipitation particles is obtained, which is specifically implemented as:

[0060] In the specific implementation, when the precipitation area is vertically detected by the frequency-modulated continuous wave radar, the radar system transmits a continuous wave signal in a linear frequency modulation mode, for example, the transmission frequency is set near 94 GHz, and the frequency is selected based on the high resolution characteristics of the millimeter wave radar in precipitation detection. The transmission signal power is controlled in the range of 20 dBm to 30 dBm, which is determined according to the typical values of the radar detection distance and the scattering cross section of the precipitation particles, to ensure that the signal penetrates the precipitation layer while avoiding excessive attenuation. The beam width is set to 0.5 degrees to 2 degrees, for example, to balance the vertical resolution and signal coverage, and the beam width is adjusted based on the antenna aperture and detection height requirements. The radar receiver synchronously receives the echo signal scattered by the precipitation particles, and the receiver sensitivity is set to between -110 dBm and -90 dBm, for example, and the dynamic range is not less than 80 dB, which is set based on the typical range of precipitation echo intensity to ensure weak signal detection and strong signal unsaturation. The transmission signal and the echo signal are mixed in a mixer, the mixer uses a double balanced mixing structure, the local oscillator signal frequency is fixedly offset from the transmission signal frequency, for example, the offset is set in the range of 100 kHz to 1 MHz, and the output difference frequency signal frequency range after mixing is, for example, between 10 kHz and 10 MHz, which corresponds to the typical requirements of precipitation particle Doppler frequency shift and distance resolution.

[0061] When the difference frequency signal obtained after mixing is subjected to spectrum analysis, the time domain difference frequency signal is converted into a frequency domain signal by using a fast Fourier transform algorithm, the number of points of the fast Fourier transform is set to 512 to 4096 points, for example, according to the detection distance resolution requirement, and the number of points is selected based on the balance between the distance library quantity and the calculation efficiency. The sampling frequency is set to 1 MHz to 10 MHz according to the maximum detection distance, for example, for a 10 km detection distance, the sampling frequency is set to 5 MHz to ensure sufficient time resolution. In the spectrum analysis process, a Hanning window function is applied to suppress spectrum leakage, and the window function length is consistent with the number of points of the fast Fourier transform, for example, for a 2048-point transform, the window length is also 2048 points. The power spectral density distribution is obtained by spectrum analysis, and the power spectral density value is calculated by squaring the signal amplitude, dividing by the system impedance of 50 ohms, and multiplying by the resolution bandwidth, which is determined according to the number of points of the fast Fourier transform and the sampling frequency, for example, for a 1024-point transform and a 2 MHz sampling frequency, the resolution bandwidth is about 1.95 kHz. The power spectral density value is output in dBm / Hz, which is convenient for subsequent analysis.

[0062] When determining the signal intensity of the precipitation particles in different radial velocity intervals based on the power spectral density distribution, first, the velocity resolution is calculated according to the radar operating frequency 94GHz and the signal processing parameters. The velocity resolution calculation formula is the radar wavelength divided by 2 and then multiplied by the fast Fourier transform sampling time interval. For example, the wavelength is about 3.19mm, and the sampling time interval is 0.5μs, and then the velocity resolution is about 0.16m / s. The power spectral density distribution is divided into several radial velocity intervals according to the calculated velocity resolution, and the width of each interval corresponds to a velocity range of 0.1m / s to 0.5m / s. The interval width is selected based on the dynamic range of the Doppler velocity of the precipitation particles. The power spectral density values in each radial velocity interval are integrated, and the integral is calculated by using the trapezoidal method or the rectangular method. The integral result is the signal intensity value corresponding to the radial velocity interval, and the signal intensity is expressed in voltage or power units. The signal intensity values of all radial velocity intervals together constitute the Doppler velocity spectrum, and the velocity range of the Doppler velocity spectrum covers-20m / s to +20m / s, which corresponds to the physical range of the falling terminal velocity of the precipitation particles, including the upward and downward motion.

[0063] In the process of forming the Doppler velocity spectrum, the signal intensity values need to be calibrated. The calibration process includes system noise deduction and path attenuation compensation. The system noise deduction is achieved by collecting the background noise power spectral density without precipitation, for example, collecting multiple measurements in clear weather and taking the average value as the noise reference. The background noise power spectral density is subtracted from the measured power spectral density. The path attenuation compensation is calculated according to the radar equation, considering the attenuation effect of electromagnetic waves in the precipitation medium. The attenuation coefficient uses an empirical value in the range of 0.01dB / km to 1dB / km, which is determined based on the attenuation model of different precipitation types such as drizzle or heavy rain. After calibration, the quality of the Doppler velocity spectrum is checked, for example, if the signal intensity of a certain velocity interval is lower than the noise threshold of 3dB, it is marked as invalid data to avoid false inversion. Finally, the Doppler velocity spectrum is used as the basis data for subsequent raindrop particle size inversion, ensuring the accuracy and reliability of the inversion results.

[0064] S2, based on the physical relationship between the terminal velocity of the raindrops and the diameter, the Doppler velocity spectrum is inverted to obtain the initial raindrop particle size distribution, which is specifically implemented as:

[0065] In the implementation process, when the mapping relationship between Doppler velocity and raindrop particle size is established according to the physical relationship between raindrop terminal falling velocity and diameter, first, a raindrop terminal falling velocity and diameter relationship model is obtained based on the principle of fluid dynamics. The model is verified by experimental observation data and forms an empirical relationship, for example, the diameter parameter covers the range of 0.1 millimeter to 5 millimeters, and the velocity parameter covers the range of 0 meters per second to 10 meters per second. The model parameters are determined by fitting historical raindrop spectrum observation data, for example, for stratiform cloud precipitation, the diameter and velocity relationship adopts a piecewise function form, when the diameter is less than 1 millimeter, the velocity increases linearly with the diameter, and when the diameter is greater than 1 millimeter, the velocity growth rate gradually slows down. When establishing the mapping relationship, the Doppler velocity range is divided into several discrete intervals, for example, each interval is set to 0.1 meters per second, and the characteristic diameter value corresponding to each velocity interval is calculated based on the physical relationship model to form a lookup table containing the corresponding relationship between velocity and diameter. The mapping relationship also introduces environmental parameter correction, for example, the air density correction coefficient is calculated according to the real-time obtained temperature and air pressure data, and the correction coefficient value range is for example 0.95 to 1.05, and the mapping relationship in the actual environment is obtained by multiplying the velocity value under standard conditions by the correction coefficient.

[0066] When the signal intensity in the Doppler velocity spectrum is converted into the number concentration of the corresponding raindrop particle size interval based on the mapping relationship between Doppler velocity and raindrop particle size, first, the signal intensity data of each velocity interval in the Doppler velocity spectrum is read, and the signal intensity is expressed in power units such as dBm. According to the pre-established mapping relationship, each velocity interval is mapped to the corresponding raindrop particle size interval, for example, the velocity interval -0.5 meters per second to 0 meters per second is mapped to the particle size interval 0.5 millimeter to 0.8 millimeter. In the conversion process, the radar equation is applied to convert the signal intensity into the radar reflectivity factor, and the radar reflectivity factor calculation needs to input radar system parameters including transmit power, antenna gain and wavelength, which are obtained from the radar system configuration file, for example, the transmit power value is 20 dBm to 30 dBm, and the antenna gain value is 25 dB to 35 dB. When the reflectivity factor is further converted into number concentration, a raindrop spectrum distribution model is used, and the distribution model type is selected according to the precipitation type, for example, an exponential distribution is used for stratiform precipitation, and a gamma distribution is used for convective precipitation. The distribution parameters are set based on climate statistics, for example, the slope parameter in the exponential distribution is set to 2 to 4.

[0067] In the process of integrating the number concentration of all raindrop size intervals to form the initial raindrop size distribution, first, the number concentration values calculated for all size intervals are summarized, and the size interval division and mapping relationship are consistent, for example, 50 equal-interval intervals are divided from 0.1 mm to 5 mm. The integration process includes data normalization processing and physical consistency checking. The normalization processing is achieved by calculating the proportion of the number concentration in each interval in the total particle number. The total particle number is the sum of the number concentrations of all intervals. The physical consistency checking is based on the characteristics of the precipitation particle distribution, for example, the number concentration value should generally decrease with the increase of the particle size. The abnormal data points are smoothed by using the weighted average method of adjacent intervals, and the weighting coefficient is set according to the particle size interval. The initial raindrop size distribution is output in the form of a discrete probability distribution, including the number concentration values of each size interval and the corresponding median particle size. At the same time, the data quality identifier is output, for example, when the relative error of the number concentration of a certain interval exceeds 20%, it is marked as low-reliability data. The final initial raindrop size distribution is used as the input data for subsequent correction processing, and the distribution statistical characteristics including the peak particle size and the distribution width parameter are recorded.

[0068] S3, synchronously acquiring the atmospheric temperature and humidity profile measured by the ground-based microwave radiometer and the cloud vertical structure information measured by the lidar, specifically implemented as:

[0069] In the specific implementation process, when the temperature profile and the humidity profile output by the ground-based microwave radiometer are synchronously received through the data acquisition interface, the data acquisition interface establishes a physical connection with the ground-based microwave radiometer using a standard communication protocol, for example, using an RS-232 serial interface or a TCP / IP network interface. The communication parameters are set according to the device specifications, for example, the baud rate is set to the range of 9600 bps to 115200 bps, which is determined based on the communication capabilities of common data acquisition devices. The temperature profile and the humidity profile data are transmitted in a specific format, for example, binary format or ASCII text format. The data frame structure includes a start identifier, a data payload, and an error check code. The error check uses a cyclic redundancy check algorithm to ensure data integrity. The received temperature profile contains a series of temperature values corresponding to height layers, with the temperature unit being Celsius, for example, the height range is from the ground to 10 kilometers, which is set based on typical tropospheric detection requirements, and the vertical resolution is set to 50 meters to 100 meters, considering the balance between data volume and accuracy. The humidity profile contains humidity values for the same height layers, with the humidity unit being relative humidity percentage or water vapor density grams per cubic meter, for example, the relative humidity range is from 0% to 100%. When forming the atmospheric temperature and humidity profile, the temperature profile and the humidity profile are matched according to the height coordinate, for example, by height index, the temperature value and the humidity value of each height layer are combined into a unified data structure, and the formed atmospheric temperature and humidity profile contains three data dimensions of height, temperature, and humidity, and additional acquisition time information, with the time stamp accuracy reaching seconds, and the time synchronization is based on the network time protocol or the GPS clock source.

[0070] Meanwhile, when receiving the backscattering signal vertical profile measured by the laser radar through the data acquisition interface, the data acquisition interface and the laser radar system are connected through a coaxial cable or an optical fiber, and the interface type is selected according to the signal characteristics, for example, an analog signal is collected by an analog-to-digital converter, and a digital signal is received by a direct memory access method. The backscattering signal vertical profile contains a sequence of signal intensities corresponding to consecutive height points, and the signal intensity is in units of voltage volts or power decibels milliwatts, for example, the voltage range is 0V to 5V, and the height range is from the ground to 15 kilometers, and the vertical resolution is set to 5 meters to 30 meters. The resolution is determined based on the pulse length and sampling rate of the laser radar. When identifying the cloud base height and cloud layer thickness based on the backscattering signal vertical profile, the original signal is first preprocessed, for example, a moving average filter is applied to reduce random noise, and the sliding window width is set to 3 to 5 data points. The value is determined based on the signal-to-noise ratio optimization experiment. The cloud base height is identified by detecting the signal intensity mutation point, and the cloud bottom detection threshold is set, which is calculated based on the statistical characteristics of the background signal measured during the cloudless period, for example, the average value + 3 × standard deviation of the background signal is taken as the threshold, and the background signal is obtained by continuously collecting multiple cloudless profiles. When the number of height points whose signal intensity continuously exceeds the threshold reaches a set minimum value, for example, 5 points, the first height point that exceeds the threshold is determined as the cloud base height. The cloud layer thickness is calculated by identifying the region from the cloud base height where the signal intensity is continuously higher than the threshold until the signal intensity drops to the background level, and the difference between the cloud top height and the cloud base height is taken as the cloud layer thickness. The cloud top height is determined by a similar method but focusing on the signal intensity falling edge. The generated cloud vertical structure information is stored in a structured data format, containing cloud base height, cloud top height and cloud layer thickness fields, and recording the laser radar measurement timestamp.

[0071] When time-synchronizing the atmospheric temperature and humidity profile and the cloud vertical structure information, a unified time reference source is used to ensure time consistency, such as using a GPS receiver to provide a coordinated universal time signal, or synchronizing the system clock through a network time protocol. The time synchronization accuracy requirement is better than 1 second. A precise time stamp is assigned to each atmospheric temperature and humidity profile data record and cloud vertical structure information data record. The time stamp format includes year, month, day, hour, minute, second, and millisecond fields. Time synchronization is achieved through a time matching algorithm, such as calculating the difference between the collection time of the atmospheric temperature and humidity profile and the measurement time of the cloud vertical structure information. When the absolute value of the time difference is less than a set threshold, such as 500 milliseconds, the two sets of data are considered synchronized. The threshold is determined based on the data collection period and the device response time. The associated data is organized in time sequence, such as each valid time point corresponding to a set of atmospheric temperature and humidity profile and cloud vertical structure information, forming a spatiotemporally aligned data set. An exception handling mechanism is implemented during synchronization, such as using linear interpolation to align data with similar times when the time deviation exceeds the threshold, or marking the data at that time point as out of sync and excluding it from use. The final time-synchronized association ensures the consistency of the atmospheric temperature and humidity profile and the cloud vertical structure information in the time dimension, providing a reliable data foundation for subsequent analysis.

[0072] S4, based on the atmospheric temperature and humidity profile and the cloud vertical structure information, diagnosing the physical mutation point of the initial raindrop particle size distribution in the vertical profile, specifically implemented as:

[0073] In the implementation process, when calculating the vertical distribution of liquid water content according to the atmospheric temperature and humidity profile, first, read the atmospheric temperature and humidity profile data obtained in step S3, which contains temperature and humidity values at different height layers from the ground to 10 kilometers in height, the temperature unit is Celsius, the humidity unit is relative humidity percentage, and the vertical resolution is set to 50-100 meters, which is determined based on the commonly used data collection interval in meteorological observation. When calculating the liquid water content, the air pressure parameter needs to be combined, and the air pressure value can be calculated according to the height by the standard atmospheric model, for example, the air pressure at 2 kilometers in height is about 795 hundred pascals, and the air pressure at 5 kilometers in height is about 540 hundred pascals. The calculation of liquid water content is based on the principle of saturated water vapor pressure, and the Magnus formula is used to calculate the saturated water vapor pressure value at each height layer, which calculates the saturated water vapor pressure through the temperature value, for example, the saturated water vapor pressure at 20 degrees Celsius is about 23.4 hundred pascals, and the saturated water vapor pressure at 0 degrees Celsius is about 6.11 hundred pascals. The actual water vapor pressure is obtained by multiplying the relative humidity value by the saturated water vapor pressure and then dividing by 100, when the actual water vapor pressure exceeds the saturated water vapor pressure, the excess part is the liquid water content. The liquid water content value is output in grams per cubic meter, forming the vertical distribution data of liquid water content, which contains the corresponding liquid water content value at each height layer, and records the temperature threshold and humidity threshold used in the calculation process, for example, when the relative humidity is less than 90%, the liquid water content is considered to be zero.

[0074] When identifying the height of the phase change layer combined with the cloud vertical structure information, first read the cloud vertical structure information data obtained in step S3, which contains cloud base height and cloud layer thickness parameters, for example, the cloud base height ranges from 500 meters to 3000 meters, and the cloud layer thickness ranges from 100 meters to 2000 meters, which are obtained based on the analysis of laser radar backscattering signals. The identification of the phase change layer needs to combine the temperature data in the atmospheric temperature and humidity profile, and focus on analyzing the temperature distribution characteristics in the cloud layer. The identification process sets a phase change temperature threshold, for example, the 0 degree Celsius isotherm height is taken as an important reference height, which is obtained by linearly interpolating the temperature profile, when the temperature changes from positive to negative, the corresponding height is the 0 degree Celsius isotherm height. The determination of the height of the phase change layer also needs to consider the internal structure of the cloud layer, for example, in the region above the cloud base height and within the cloud layer thickness, when the temperature is below 0 degrees Celsius, the water-ice phase change may occur. The identification result outputs one or more height values of the phase change layer, for example, the main phase change layer height and the secondary phase change layer height, and records the corresponding temperature value and cloud layer position information of each phase change layer.

[0075] Based on the vertical distribution of liquid water content and the height of the phase change layer, the initial raindrop size distribution is analyzed in the vertical direction. First, the initial raindrop size distribution data obtained in step S2 is obtained, which contains the raindrop number concentration values of different height layers, distributed according to the particle size interval, the particle size interval is from 0.1 mm to 5 mm, and the number concentration unit is per cubic meter per millimeter. Based on the initial raindrop size distribution, the vertical profile representing the total particle concentration is calculated by summing the number concentration values of all particle size intervals at each height layer, for example, for a certain height layer, the number concentration values of all particle size intervals in the range of 0.1 mm to 5 mm are added to obtain the total particle concentration value of the height layer, and the total particle concentration unit is per cubic meter. When comparing the vertical profile of total particle concentration with the cooperative change trend of liquid water content near the phase change layer, first determine the analysis window range, for example, take the height of the phase change layer as the center, extend 200 meters to 500 meters upward and downward as the analysis area. The analysis of the cooperative change trend is realized by calculating the similarity of the vertical gradient of the total particle concentration and the vertical gradient of the liquid water content, and the vertical gradient is calculated using the central difference method, for example, the gradient value of a certain height point is equal to the difference between the upper adjacent height point value and the lower adjacent height point value divided by twice the height interval. According to the significant deviation characteristics of the cooperative change trend, the physical mutation point of the initial raindrop size distribution in the vertical profile is located, and the significant deviation characteristics are judged by setting a gradient difference threshold, for example, when the difference between the total particle concentration gradient value and the liquid water content gradient value exceeds the set threshold, it is considered that there is a significant deviation, and the threshold is determined based on historical observation data statistics, for example, taking 1.5 times to 2.5 times of the standard deviation of the gradient difference value sequence as the threshold range.

[0076] When locating the physical mutation point of the initial raindrop size distribution in the vertical profile by gradient analysis, the analysis results of multiple height layers need to be considered comprehensively to avoid misjudging isolated noise points as physical mutation points. The positioning process uses a sliding window test method, for example, requiring the physical mutation point to have significant gradient anomaly performance within the upper and lower 3 height layers. The strength of the physical mutation point is calculated by normalizing the gradient difference, for example, taking the ratio of the gradient difference to the historical maximum gradient difference as the mutation strength, and the mutation strength ranges from 0 to 1. The final output of the physical mutation point information includes the height of the physical mutation point, the strength of the physical mutation point and the type of the physical mutation point, for example, the physical mutation point is classified into strong mutation point, medium mutation point and weak mutation point, and the classification is determined according to the mutation strength threshold, for example, the mutation strength greater than 0.7 is a strong mutation point, the mutation strength between 0.3 and 0.7 is a medium mutation point, and the mutation strength less than 0.3 is a weak mutation point. All physical mutation point data are associated with corresponding liquid water content vertical distribution data and cloud vertical structure information data to form a complete diagnosis result, which provides a basis for subsequent correction processing.

[0077] S5, the constraint weight of the corresponding auxiliary data is inhibited at the height layer where the physical mutation point is located, and the initial raindrop size distribution is fused and corrected based on the information of the adjacent height layer to obtain the corrected raindrop size distribution, and the implementation is as follows:

[0078] In the implementation process, when determining the constraint weight coefficient of the atmospheric temperature and humidity profile and the cloud vertical structure information corresponding to the height layer where the physical mutation point is located, first, the physical mutation point information obtained in step S4 is read, which includes the height where the physical mutation point is located and the physical mutation point intensity parameter, for example, the height where the physical mutation point is located ranges from 1 km to 5 km, and the physical mutation point intensity value ranges from 0 to 1. The constraint weight coefficient is used to quantify the contribution of the atmospheric temperature and humidity profile and the cloud vertical structure information in data fusion. The initial value of the weight coefficient is set based on the physical mutation point intensity, for example, when the physical mutation point intensity is greater than 0.7, the initial weight coefficient is set to 0.8 to 1.0, when the physical mutation point intensity is between 0.3 and 0.7, the initial weight coefficient is set to 0.5 to 0.8, and when the physical mutation point intensity is less than 0.3, the initial weight coefficient is set to 0.2 to 0.5. The weight coefficient is also determined considering the position characteristics of the height layer, for example, the weight coefficient of the height layer inside the cloud layer is higher than that of the height layer outside the cloud layer. The specific value of the weight coefficient is obtained by looking up a table or using an empirical formula, for example, for the height layer near the cloud base height, the weight coefficient is increased by an adjustment amount of 0.1 to 0.2. Finally, the atmospheric temperature and humidity profile constraint weight coefficient and the cloud vertical structure information constraint weight coefficient corresponding to each height layer where the physical mutation point is located are assigned, and the setting basis and adjustment rules of the weight coefficient are recorded.

[0079] When reducing the constraint weight coefficient of the height layer where the physical mutation point is located, the determined constraint weight coefficient is adjusted based on the physical mutation point intensity. The adjustment amplitude is determined according to the corresponding relationship between the physical mutation point intensity and the adjustment coefficient, for example, when the adjustment coefficient is set to 0.5 to 0.8, the new value of the constraint weight coefficient is equal to the original constraint weight coefficient multiplied by the adjustment coefficient. The value of the adjustment coefficient is negatively correlated with the physical mutation point intensity, for example, when the physical mutation point intensity is 0.8, the adjustment coefficient is 0.5, and when the physical mutation point intensity is 0.4, the adjustment coefficient is 0.7. The reduced constraint weight coefficient needs to meet the minimum constraint, for example, the constraint weight coefficient is not less than 0.1, to avoid completely excluding the influence of auxiliary data. During the adjustment process, the continuity of the weight coefficients of adjacent height layers is also checked, for example, when the constraint weight coefficient of the height layer where the physical mutation point is located and the constraint weight coefficient of the adjacent height layer differ by more than 0.3, smoothing processing is performed, and the moving average of the constraint weight coefficients of the adjacent height layers is used as the adjustment reference. The reduced constraint weight coefficient is used for subsequent data fusion steps, and the numerical change before and after the adjustment is recorded.

[0080] When the initial raindrop size distribution of the height layer where the physical mutation point is located is calculated by interpolation based on the initial raindrop size distribution of the adjacent height layers, first, the two height layers adjacent to the height layer where the physical mutation point is located and not appearing the physical mutation point are selected as reference layers. The reference layer selection conditions include height proximity and data integrity, for example, the height difference between the reference layer and the height layer where the physical mutation point is located is not more than 200 meters, and the initial raindrop size distribution data of the reference layer is complete without missing. Based on the initial raindrop size distribution data of the two reference layers, a continuous physical change model of the particle size distribution in the vertical direction is constructed. The model construction adopts a physically constrained interpolation method, for example, it is assumed that the raindrop size distribution in the vertical direction decays exponentially or changes linearly. The model parameters are obtained by fitting the particle size distribution data of the two reference layers, for example, the decay coefficient in the exponential decay model is determined by least squares fitting. The alternative distribution of the height layer where the physical mutation point is located is calculated by the continuous physical change model. In the calculation process, the height layer where the physical mutation point is located is substituted into the continuous physical change model, and the number concentration value of each particle size interval at this height is calculated, for example, for the particle size interval of 0.5 mm to 1.0 mm, the number concentration value is calculated by the model and ensured to be dimensionally consistent, with a unit of per cubic meter per millimeter. The output format of the alternative distribution is consistent with the initial raindrop size distribution, including complete particle size intervals and number concentration values.

[0081] When the corrected raindrop size distribution is generated by fusing the alternative distribution with the weight-adjusted atmospheric temperature and humidity profile and cloud vertical structure information, first, the weight-adjusted atmospheric temperature and humidity profile and cloud vertical structure information are read. These data have been weighted according to the reduced constraint weight coefficients, for example, the temperature value and humidity value in the atmospheric temperature and humidity profile are multiplied by the corresponding constraint weight coefficient, and the cloud base height and cloud layer thickness parameters in the cloud vertical structure information are also adjusted by the corresponding weight. In the fusion process, a data assimilation method is adopted, for example, the alternative distribution is taken as the prior distribution, and the weighted atmospheric temperature and humidity profile and cloud vertical structure information are taken as the observation constraint. The corrected raindrop size distribution is obtained by Bayesian fusion or optimal interpolation method. In the fusion, the fusion weight is set to balance the contribution of the alternative distribution and the auxiliary data, for example, the fusion weight of the alternative distribution is set to 0.6 to 0.8, and the fusion weight of the auxiliary data is set to 0.2 to 0.4. The weight value is determined based on the data reliability evaluation. After the corrected raindrop size distribution is generated, a physical reasonableness check is performed, for example, the number concentration value should be non-negative and generally decreasing with the increase of particle size. Abnormal values are replaced or smoothed by the adjacent particle size interval value. The finally output corrected raindrop size distribution is taken as the input data of step S6, and the weight parameters and correction rules used in the fusion process are recorded.

[0082] In the specific implementation process, it is also necessary to consider the handling of various boundary conditions. For example, when the physical mutation point is located at the detection boundary height, the reference layer is selected using adjacent height layers on one side, and a continuous physical change model is constructed through extrapolation. When multiple physical mutation points are close to each other, the height layers where these physical mutation points are located are merged, and the constraint weight coefficients are adjusted uniformly. During the data fusion stage, if there is a significant conflict between the alternative distribution and the auxiliary data, a conflict resolution mechanism is activated, such as prioritizing the constraint effect of cloud vertical structure information or increasing the weight ratio of adjacent height layer information. The entire correction process is subject to quality monitoring, for example, by comparing the changes in the statistical characteristics of raindrop size distribution before and after correction to evaluate the correction effect. When the corrected distribution shows obvious unreasonable characteristics, the parameter settings of each step are checked back and the correction process is re-executed. These supplementary measures ensure that the corrected raindrop size distribution not only eliminates the influence of physical mutation points but also maintains physical consistency and data reliability in the vertical direction.

[0083] S6. Output the corrected raindrop size distribution, specifically implemented as follows:

[0084] In the specific implementation process, when converting the corrected raindrop size distribution into a vertical distribution data product in a preset data format, the corrected raindrop size distribution data generated in step S5 is first read. This data contains raindrop number concentration values ​​at different height levels, distributed according to size ranges. For example, the height range is from the ground to 10 kilometers, the vertical resolution is set to 50 meters to 100 meters, the size range is from 0.1 millimeters to 5 millimeters, and the number concentration unit is per cubic meter per millimeter. The preset data format adopts a commonly used scientific data format, such as NetCDF or HDF5. This format selection is based on the complexity of the data structure and the requirements for metadata support. The conversion process includes reorganizing the data structure and adding metadata. For example, height dimension variables, size dimension variables, and number concentration data variables are created in the data product. The height dimension variable stores the values ​​of each height level in meters, the size dimension variable stores the median values ​​of each size range in millimeters, and the number concentration data variable stores the number concentration values ​​corresponding to each height level and each size range in the form of a three-dimensional array. Metadata additions include acquisition time, geographic location, radar parameters, and calibration processing information. For example, the acquisition time is in UTC time format, the geographic location uses latitude and longitude coordinates, radar parameters include transmission frequency and antenna gain, and calibration processing information records the location of physical abrupt changes and weight adjustment parameters. The data product also includes a quality flag field to identify the reliability of each data point. For example, the quality level is set based on the reference layer data and fusion weights used in the calibration process, with three levels: high reliability, medium reliability, and low reliability. The final vertically distributed data product is output in file format, with the file size determined by the amount of data. For example, for 1 hour of observation data, the file size is approximately 10MB to 100MB. The file naming convention includes a timestamp and site identifier for easy management.

[0085] When transmitting the vertical distribution data product to the display device for real-time graphical display and storing the vertical distribution data product into the non-volatile memory to form a historical data set, the data interface adopts a standard communication protocol, such as an Ethernet interface or a USB interface, and the communication rate is set to a range of 100 Mbps to 1 Gbps, which is determined based on the size of the data product and the real-time requirement. Data compression and encryption are implemented during transmission, such as lossless compression using the ZIP algorithm to reduce the amount of transmitted data and the AES encryption algorithm to ensure data transmission security. Real-time graphical display is achieved through dedicated display software running on a display device such as a computer workstation or a touch screen display. The display interface includes a vertical distribution profile, a time series graph, and a statistical information panel. The vertical distribution profile has height as the vertical axis and particle size as the horizontal axis, with the number concentration value represented by a color gradient, such as a gradient color from blue to red representing a low to high number concentration. The graphical update frequency is consistent with the data acquisition frequency, such as updating the display every 1 minute. The time series graph displays the number concentration of a specific height layer or particle size interval over time, and the statistical information panel displays key parameters such as total particle concentration peak and distribution width. At the same time, the vertical distribution data product is stored in the non-volatile memory, which includes a solid state disk or a network attached storage device, with a storage capacity configured according to the historical data preservation requirements, such as 10 TB to 100 TB of storage space for saving historical data sets for 1 to 5 years. The historical data set is organized in chronological order, such as creating a directory structure by year, month, and day, and each data product file is stored in the corresponding time directory, and an index database is established for quick retrieval and query. Data backup and redundancy mechanisms are implemented during storage, such as periodically backing up data to offline storage devices or cloud storage platforms, with a backup frequency set to once a day to once a week to ensure data security and recoverability. The entire output process records operation logs, including data transmission time, display update time, and storage status, for monitoring system operation and troubleshooting.

[0086] Embodiment 2: Figure 2 A structural diagram of a micro rain radar system is given, a micro rain radar system, comprising:

[0087] A frequency-modulated continuous wave radar is used to vertically detect the precipitation area and obtain the Doppler velocity spectrum of the precipitation particles.

[0088] The processing device is connected with the frequency-modulated continuous wave radar, and is used for obtaining an initial raindrop particle size distribution by inverting a Doppler velocity spectrum based on a physical relationship between a raindrop terminal falling speed and a diameter, synchronously acquiring an atmospheric temperature and humidity profile measured by a ground-based microwave radiometer and cloud vertical structure information measured by a laser radar, diagnosing a physical mutation point of the initial raindrop particle size distribution on a vertical profile based on the atmospheric temperature and humidity profile and the cloud vertical structure information, inhibiting a constraint weight of corresponding auxiliary data at a height layer where the physical mutation point is located, and correcting the initial raindrop particle size distribution based on information of adjacent height layers to obtain a corrected raindrop particle size distribution.

[0089] The output device is connected with the processing device, and is used for converting the corrected raindrop particle size distribution into vertical distribution data products in a preset data format and outputting.

[0090] The frequency-modulated continuous wave radar, the processing device and the output device are sequentially connected to form a data processing link.

[0091] In the implementation process, when converting the corrected raindrop size distribution into a vertical distribution data product in a preset data format and outputting, first, the corrected raindrop size distribution data generated by the processing device is read, which contains the raindrop number concentration values of different height layers, distributed by particle size interval, for example, the height range is from the ground to 10 kilometers, the vertical resolution is set to 50 meters to 100 meters, the particle size interval is from 0.1 millimeter to 5 millimeters, and the number concentration unit is per cubic meter per millimeter. The preset data format adopts a scientific data universal format, such as the NetCDF format or the HDF5 format, which is determined based on the complexity of the data structure and the metadata support requirements, for example, the NetCDF format supports multi-dimensional data arrays and attribute metadata, which is convenient for storage and exchange. The conversion process includes reorganization of the data structure and addition of metadata, for example, creating height dimension variables, particle size dimension variables, and number concentration data variables in the data product, the height dimension variables store the values of each height layer in meters, the particle size dimension variables store the values of each particle size interval in millimeters, and the number concentration data variables store the corresponding number concentration values of each height layer and each particle size interval in the form of a three-dimensional array. The metadata addition includes collection time, geographic location, radar parameters, and correction processing information, for example, the collection time uses the UTC time format, the geographic location uses the latitude and longitude coordinates, the radar parameters include the transmission frequency and the antenna gain, and the correction processing information records the position of the physical mutation point and the weight adjustment parameter. The data product also contains a quality flag field for identifying the reliability of each data point, for example, according to the reference layer data and the fusion weight used in the correction process, the quality level is set, and the quality level is divided into three levels: high reliability, medium reliability, and low reliability. The output process is realized through a data interface, for example, an Ethernet interface or a USB interface, the vertical distribution data product is transmitted to a display device for real-time graphical display or stored in a non-volatile memory, data compression and encryption processing is implemented during transmission, for example, lossless compression is performed using the ZIP algorithm to reduce data volume, and the AES encryption algorithm is used to ensure data transmission security. Real-time graphical display is realized through special display software, for example, a visualization program running on a computer workstation, the display interface contains a vertical distribution profile and a statistical information panel, the graphical update frequency is consistent with the data acquisition frequency, for example, the display is updated every 1 minute. At the same time, the vertical distribution data product is stored in a non-volatile memory, such as a solid state disk or a network attached storage device, to form a historical data set, the storage capacity is configured according to the historical data preservation requirements, for example, 10TB to 100TB of storage space is configured to save 1 year to 5 years of historical data, the historical data set is organized in chronological order and an index database is established for fast retrieval.

[0092] In the light rain radar system, the connection mode of each device is realized through standard physical interface and communication protocol. The connection between the frequency-modulated continuous wave radar and the processing device adopts cable or optical fiber physical medium, such as using coaxial cable to transmit radio frequency signal, or using optical fiber to transmit digital signal, the connection interface type includes BNC interface or SMA interface, the communication protocol adopts self-defined protocol or standard protocol such as UDP protocol, which ensures high-speed data transmission, for example, the data transmission rate is set to 100Mbps to 1Gbps to meet the real-time processing requirement. The connection between the processing device and the output device is realized through internal bus or external interface, such as using PCI Express bus for internal data transmission, or using Ethernet interface for external connection, the communication protocol adopts TCP / IP protocol to ensure reliable transmission, and error detection and retransmission mechanism is implemented during data transmission, such as using cyclic redundancy check to check data integrity. In the overall system, the frequency-modulated continuous wave radar, the processing device and the output device are connected in turn to form a data processing link, the frequency-modulated continuous wave radar is responsible for raw data acquisition, the processing device is responsible for data inversion and fusion correction, and the output device is responsible for data format conversion and output. The power supply and signal connection between devices are realized through unified backplane or distributed power supply module, such as using 24V DC power supply to ensure stable operation of the system. In addition, the system also contains a synchronous clock module, such as a GPS receiver or a network time protocol client, which is used to coordinate the time synchronization of each device, and the time synchronization accuracy is required to be better than 1ms to ensure the consistency of data timestamp. All connection lines adopt shielding design to reduce electromagnetic interference, and grounding protection is implemented to ensure system safety.

[0093] The calculations involved in the embodiments are all de-dimensioned numerical calculations. The preset parameters and threshold values in the calculations are set by a person skilled in the art according to actual conditions.

[0094] The above embodiments can be realized wholly or partially by software, hardware, firmware or any other combination. When realized by software, the above embodiments can be realized wholly or partially in the form of a computer program product.

[0095] Those skilled in the art can realize that the modules and algorithm steps of the examples described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized by hardware or software depends on the specific application of the technical solution and the constraints of the invention. A person skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0096] In addition, each functional module in each embodiment of the present application can be integrated in one processing module, or each module can exist physically separately, or two or more modules can be integrated in one module.

[0097] In several embodiments provided in the present application, it should be understood that the disclosed system, device and method can be implemented in other ways. For example, the device embodiments described above are merely illustrative, for example, the division of the modules is only a logical function division, and actual implementation can have another division manner, for example, a plurality of modules or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the displayed or discussed each other can be indirect coupling or communication connection through some interfaces, devices or modules, which can be electrical, mechanical or other forms.

[0098] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0099] Finally: the above is only the preferred embodiment of the present application, and is not used to limit the present application, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A raindrop size retrieval method based on Doppler velocity spectrum, characterized in that, The method comprises the following steps: S1, vertically detecting a precipitation area by a frequency-modulated continuous wave radar to obtain a Doppler velocity spectrum of precipitation particles; S2, inversely calculating the Doppler velocity spectrum based on a physical relationship between terminal falling velocity of raindrops and diameter to obtain an initial raindrop particle size distribution; S3, synchronously obtaining an atmospheric temperature and humidity profile measured by a ground-based microwave radiometer and cloud vertical structure information measured by a laser radar; S4, diagnosing a physical mutation point of the initial raindrop particle size distribution on a vertical profile based on the atmospheric temperature and humidity profile and the cloud vertical structure information; S5, suppressing a constraint weight of corresponding auxiliary data at a height layer where the physical mutation point is located, and correcting the initial raindrop particle size distribution based on information of adjacent height layers to obtain a corrected raindrop particle size distribution; S6, outputting the corrected raindrop particle size distribution.

2. The Doppler velocity spectrum based raindrop size retrieval method according to claim 1, characterized in that, The method for vertically detecting a precipitation area by a frequency-modulated continuous wave radar to obtain a Doppler velocity spectrum of precipitation particles comprises the following steps: a frequency-modulated continuous wave radar transmits a continuous wave signal and receives an echo signal scattered by precipitation particles; the echo signal is mixed to obtain a difference frequency signal; the difference frequency signal is subjected to frequency spectrum analysis to generate a power spectrum density distribution; the signal intensity of the precipitation particles in different radial velocity intervals is determined based on the power spectrum density distribution to form a Doppler velocity spectrum.

3. The Doppler velocity spectrum based raindrop size retrieval method according to claim 1, wherein, The method for inversely calculating the Doppler velocity spectrum based on a physical relationship between terminal falling velocity of raindrops and diameter to obtain an initial raindrop particle size distribution comprises the following steps: a mapping relationship between Doppler velocity and raindrop particle size is established according to the physical relationship between terminal falling velocity of raindrops and diameter; the signal intensity in the Doppler velocity spectrum is converted into number concentration of corresponding raindrop particle size intervals based on the mapping relationship between Doppler velocity and raindrop particle size; the number concentrations of all raindrop particle size intervals are integrated to form the initial raindrop particle size distribution.

4. The Doppler velocity spectrum based raindrop size retrieval method according to claim 1, wherein, The method for synchronously obtaining an atmospheric temperature and humidity profile measured by a ground-based microwave radiometer and cloud vertical structure information measured by a laser radar comprises the following steps: temperature and humidity profiles output by the ground-based microwave radiometer are synchronously received through a data acquisition interface to form the atmospheric temperature and humidity profile; a backscattering signal vertical profile measured by the laser radar is received through the data acquisition interface, and cloud bottom height and cloud thickness are identified based on the backscattering signal vertical profile to generate the cloud vertical structure information; the atmospheric temperature and humidity profile and the cloud vertical structure information are time-synchronously associated.

5. The Doppler velocity spectrum based raindrop size retrieval method according to claim 1, wherein, The method for diagnosing a physical mutation point of the initial raindrop particle size distribution on a vertical profile based on the atmospheric temperature and humidity profile and the cloud vertical structure information comprises the following steps: a vertical distribution of liquid water content is calculated according to the atmospheric temperature and humidity profile; a phase change layer is identified according to the cloud vertical structure information; gradient analysis is performed on the initial raindrop particle size distribution in a vertical direction based on the vertical distribution of liquid water content and the height of the phase change layer; the physical mutation point of the initial raindrop particle size distribution on the vertical profile is located through the gradient analysis.

6. The Doppler velocity spectrum based raindrop size retrieval method according to claim 5, wherein, The gradient analysis on the initial raindrop size distribution in the vertical direction includes: calculating a vertical profile representing total particle concentration based on the initial raindrop size distribution; comparing the vertical profile representing total particle concentration with the vertical distribution of liquid water content near the phase change layer; and locating a physical mutation point on the vertical profile of the initial raindrop size distribution according to a significant deviation of the cooperative change trend.

7. The Doppler velocity spectrum based raindrop size retrieval method according to claim 1, wherein, The constraint weight of the corresponding auxiliary data is suppressed at the height layer where the physical mutation point is located, and the initial raindrop size distribution is fused and corrected based on the information of adjacent height layers to obtain a corrected raindrop size distribution, including: determining the constraint weight coefficients of atmospheric temperature and humidity profiles and cloud vertical structure information corresponding to the height layer where the physical mutation point is located; reducing the constraint weight coefficients of the height layer where the physical mutation point is located; calculating an alternative distribution of the height layer where the physical mutation point is located based on the initial raindrop size distribution of adjacent height layers by interpolation; fusing the alternative distribution with the atmospheric temperature and humidity profiles and the cloud vertical structure information after weight adjustment to generate the corrected raindrop size distribution.

8. The Doppler velocity spectrum based raindrop size retrieval method according to claim 7, wherein, The calculation of the alternative distribution of the height layer where the physical mutation point is located based on the initial raindrop size distribution of adjacent height layers by interpolation includes: selecting two adjacent height layers above and below the height layer where the physical mutation point is located as reference layers; constructing a continuous physical change model of the particle size distribution in the vertical direction based on the initial raindrop size distribution data of the two reference layers; and calculating the alternative distribution of the height layer where the physical mutation point is located through the continuous physical change model.

9. The Doppler velocity spectrum based raindrop size retrieval method according to claim 1, wherein, The output of the corrected raindrop size distribution includes: converting the corrected raindrop size distribution into vertical distribution data products in a preset data format; transmitting the vertical distribution data products to a display device through a data interface for real-time graphical display, and storing the vertical distribution data products in a non-volatile memory to form a historical data set.

10. A micro rain radar system for implementing a Doppler velocity spectrum based raindrop size retrieval method according to any one of claims 1 to 9, characterized in that, including: a frequency-modulated continuous wave radar for vertical detection of a precipitation area to obtain a Doppler velocity spectrum of precipitation particles; a processing device connected to the frequency-modulated continuous wave radar, configured to obtain an initial raindrop size distribution by inverting the Doppler velocity spectrum based on the physical relationship between the terminal falling velocity and the diameter of the raindrop, and simultaneously obtain atmospheric temperature and humidity profiles measured by a ground-based microwave radiometer and cloud vertical structure information measured by a laser radar, diagnose a physical mutation point on the vertical profile of the initial raindrop size distribution based on the atmospheric temperature and humidity profiles and the cloud vertical structure information, suppress the constraint weight of the corresponding auxiliary data at the height layer where the physical mutation point is located, and fuse and correct the initial raindrop size distribution based on the information of adjacent height layers to obtain a corrected raindrop size distribution; an output device connected to the processing device, configured to convert the corrected raindrop size distribution into vertical distribution data products in a preset data format and output; wherein the frequency-modulated continuous wave radar, the processing device and the output device are connected in sequence to form a data processing link.

Images