A method for measuring solid precipitation based on digital holography
By using digital holographic imaging technology and proprietary empirical formulas, the problems of inaccurate particle type differentiation and missing parameters in existing solid precipitation measurement technologies have been solved, achieving high-precision solid precipitation measurement and microscopic data acquisition.
Patent Information
- Application Number
- CN202611273037.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-21
- Publication Date
- 2026-11-17
AI Technical Summary
Existing solid precipitation measurement technologies cannot accurately distinguish between different types of solid precipitation particles and lack multi-dimensional microphysical parameters, resulting in large measurement errors and a lack of microscopic data support.
A digital holographic imaging device is used to acquire holograms of solid precipitation particles. Through preprocessing, particle reconstruction, classification and identification, and extraction of microphysical parameters, the particle mass is calculated using a proprietary empirical formula, enabling accurate measurement and acquisition of spectral distribution characteristics.
It enables accurate identification of different types of solid precipitation particles such as snow, sleet, and hail, and simultaneous extraction of multi-dimensional microphysical parameters, reducing calculation errors and providing high-precision numerical data and microscopic data support for solid precipitation.
Smart Images

Figure QLYQS_4 
Figure QLYQS_6
Abstract
Description
Technical Field
[0001] This invention belongs to the field of meteorology, weather modification, and disaster prevention and mitigation technology, specifically relating to a method for measuring hail momentum based on digital holography. Background Technology
[0002] In the field of meteorological observation, solid precipitation refers to the mass or depth of solid precipitation particles falling on a unit area per unit time after completely melting into water. It is a core meteorological parameter for analyzing solid precipitation weather processes, assessing precipitation intensity, and mitigating disasters. Among known technologies, traditional methods for measuring solid precipitation mainly include the weighing method, rain gauge conversion method, and laser particle spectrometer method. 1. The gravimetric method collects solid precipitation samples over a certain period of time, weighs them, and then converts the weight into meltwater depth. It is the benchmark method for measuring solid precipitation. However, this method relies on manual monitoring, cannot achieve continuous real-time observation, and is easily affected by factors such as sample evaporation and wind erosion. 2. The rain gauge method collects solid precipitation using a rain gauge and reads the melt depth after it melts naturally. This method has low measurement accuracy and is only applicable to uniform solid precipitation processes. It cannot reflect the microscopic distribution characteristics of precipitation particles. 3. Laser precipitation particle spectrometers record the horizontal size of particles by measuring the amplitude of the laser beam blocked by the particles. Combined with empirical formulas, the particle mass and precipitation amount can be calculated. This technology can achieve automatic observation. However, due to the limitations of the laser detection principle, it cannot distinguish between different types of solid precipitation particles (such as snow, graupel, and hail), and the particle size measurement range is limited. The measurement error is relatively large for fine snow crystals and large hailstones.
[0003] Meanwhile, it is well known in the field of meteorology that the mass-particle size and thickness-diameter of solid precipitation particles follow a power law relationship. The empirical regression coefficients of different types of solid precipitation particles are significantly different. Particle volume and effective density can be derived by combining their geometric parameters with empirical formulas. The calculation of solid precipitation needs to be based on the integration or summation of core microphysical parameters such as particle number concentration, mass, and terminal velocity.
[0004] Another method is optical imaging-based solid-state precipitation estimation: this method uses ordinary CCD imaging and laser scattering technology (such as laser precipitation particle spectrometers and two-dimensional raindrop meters) to photograph precipitation particles, extract particle size parameters, estimate particle mass using empirical formulas, and simultaneously obtain approximate particle velocity values through multi-device velocities or statistical algorithms. However, the particle size is an approximation and the velocity is a statistical estimate, neither of which is precise measurement data.
[0005] In addition, some existing technologies combine digital holography with particle spectral analysis to try to extract the particle size parameters of precipitation particles through holographic images and then calculate precipitation using empirical formulas. However, this technical solution only applies to a single type of solid precipitation particle (such as only snow crystals or only hail) and does not consider the differences in microphysical characteristics of different types of solid precipitation particles such as snow, graupel, and hail.
[0006] In summary, existing solid precipitation measurement technologies suffer from the following core technical defects, which directly lead to low measurement accuracy, limited data dimensions, and an inability to reflect the microscopic characteristics of solid precipitation: 1. Inability to accurately distinguish different types of solid precipitation particles, and poor adaptability of empirical formulas: Existing equipment such as laser particle spectrometers and holographic cloud combination detectors cannot accurately classify snow, sleet, and hail. Furthermore, the empirical regression coefficients of parameters such as mass-particle size, thickness-diameter, and effective density of different types of solid precipitation particles vary significantly. Using a single empirical formula to calculate particle mass and volume will produce large errors, ultimately leading to distorted precipitation measurement results. 2. Inability to simultaneously acquire multi-dimensional microphysical parameters of solid precipitation particles: Existing technologies can only extract basic parameters such as particle size, and cannot simultaneously acquire core parameters such as particle morphology (e.g., roundness, major and minor axes), terminal velocity, and effective density. Accurate calculation of solid precipitation depends on the coordinated data of particle number concentration, mass, and terminal velocity. The lack of parameters leads to insufficient theoretical basis for precipitation calculation. 3. The measurement results lack microscopic data support and cannot reflect the spatiotemporal distribution characteristics of precipitation: Existing technology can only output the overall value of solid precipitation, and cannot obtain the distribution characteristics of solid precipitation particles of different sizes and types per unit time and unit volume. It cannot provide microscopic data support for the mechanism analysis of solid precipitation weather processes and the prediction of precipitation intensity evolution. 4. Traditional holographic technology has insufficient accuracy in extracting parameters of solid precipitation particles: The image preprocessing algorithm of existing holographic detection equipment is not optimized for the characteristics of solid precipitation particles and is easily affected by external factors such as stray light and vibration, resulting in low particle recognition rate and large errors in particle size and morphology parameter extraction, which in turn affects the accuracy of subsequent precipitation calculation. Summary of the Invention
[0007] The purpose of this invention is to solve the technical problems of inaccurate particle type differentiation, incomplete acquisition of microphysical parameters, large errors in precipitation calculation, and lack of microscopic data support in existing solid precipitation measurement methods. It aims to achieve accurate identification of different types of solid precipitation particles such as snow, sleet, and hail, simultaneous extraction of multi-dimensional microphysical parameters, and thus derive high-precision solid precipitation values and obtain the spectral distribution characteristics of solid precipitation particles, providing comprehensive and accurate solid precipitation data for meteorological observation, weather modification, and disaster reduction and prevention.
[0008] To achieve the above objectives, this invention provides a method for measuring solid precipitation based on digital holography. This method acquires holograms of solid precipitation particles using a digital holographic imaging device. After preprocessing, particle reconstruction, classification and identification, and extraction of microphysical parameters, the particle mass is calculated using specific empirical formulas for different types of solid precipitation particles. Finally, the method achieves accurate measurement of precipitation through a solid precipitation calculation model, while also obtaining the spectral distribution characteristics of solid precipitation particles. Specifically, the following steps are included: Step S01: Collect and store holograms of solid precipitation particles during their descent. Step S02: Holographic preprocessing and particle 3D reconstruction; Step S03: Accurate identification and classification of solid precipitation particles to form a solid precipitation particle classification dataset; Step S04: Simultaneous extraction of multi-dimensional microphysical parameters to obtain the particle spectral distribution function N(D); Step S05: Calculate the individual mass of different types of particles; Step S06: Calculate solid precipitation; Step S07: Data integration and standardized output.
[0009] Furthermore, the specific process of step S01 is to place the holographic hail spectrometer at the designated location for solid precipitation observation, start the hologram acquisition program, and continuously capture holograms of solid precipitation particles falling according to the set parameters. The acquired hologram data is then stored in real time to the instrument's built-in storage module.
[0010] Furthermore, step S02 specifically includes: S021). Hologram preprocessing: Denoising, background subtraction, and contrast enhancement are performed on each frame of the acquired hologram. Median filtering algorithm is used to eliminate image noise, and background subtraction method is used to subtract the particle-free reference background to improve the clarity of the hologram and ensure that the particle outline can be accurately identified. S022). Particle 3D Reconstruction: Based on digital holographic imaging theory, the preprocessed hologram is numerically reconstructed, and the 3D spatial morphology of solid precipitation particles is calculated by the diffraction light field restoration algorithm to obtain the 2D projection image and 3D contour data of the particles, providing a basis for subsequent particle identification and parameter extraction.
[0011] Furthermore, step S03 specifically includes: S031). Particle Feature Extraction: Image analysis is performed on the reconstructed two-dimensional projection image of the particles to extract the morphological feature parameters of each particle, including roundness, axial ratio, number of contour edges, and equivalent diameter. Among them, snow crystals are hexagonal / dendritic, with low roundness, large difference in axial ratio, and many edges; graupel is spherical / conical, with roundness between 0.8 and 1.0 and axial ratio of about 0.9. S032). Particle classification and recognition: Based on the morphological differences between snow and sleet, a classification threshold is set: Particles with a roundness < 0.8, number of edges ≥ 6, and axial ratio > 1.2 are classified as snow crystals; particles with a roundness ≤ 0.8, axial ratio ≈ 0.9, and number of edges < 3 are classified as graupel. S033). Classification result storage: Store the corresponding number, classification result, and morphological feature parameters of each particle to form a solid precipitation particle classification dataset.
[0012] Furthermore, step S04 specifically includes: S041). Based on particle classification results and 3D reconstruction data, microphysical parameters were extracted for snow, sleet, and hail, and the equivalent particle diameter D, terminal velocity v, and number concentration N(D) were obtained simultaneously. All parameters were stored according to particle number and type, forming a complete particle microphysical parameter dataset. S042). Equivalent diameter extraction: Based on the particle's projected area, the equivalent diameter D of the particle is calculated using the circular equivalence method, that is, the particle's projected area is equal to the diameter of a circle with the same area; S043). Calculation of terminal velocity: Based on the single-exposure multiple sampling technique, the trajectory of the same particle in the hologram is extracted, the displacement and time difference of the particle between adjacent pulses are calculated, and the falling velocity of the particle is obtained. After multiple samplings and taking the average value, it is taken as the terminal velocity v of the particle. The velocity unit is mm / s. S044). Particle number concentration statistics: based on the effective sampling volume of the holographic hail spectrometer. The particle spectral distribution function N(D) is obtained by counting the number of particles of different types and particle size ranges per unit volume. .
[0013] Furthermore, step S05 specifically includes: To address the differences in the microphysical characteristics of snow, sleet, and hail, specific empirical formulas are used to calculate the mass m(D) of a single particle, in mg. The core principles are based on the power-law relationship between thickness and diameter of solid precipitation particles, the volume calculation formula, and the power-law relationship between mass and particle size, ensuring the accuracy of the mass calculation. S051). Snow Crystal Mass Calculation First, based on the snow crystal type, apply the corresponding thickness-diameter power-law relationship h = A· , where A and B are empirical regression coefficients, to calculate the snow crystal thickness h; A flat prism circumscribed volume model is adopted to calculate the snow crystal volume V≈Area×h, where Area is the projected area of the particle; Combined with the mass-particle size power-law relationship of snow crystals m=α· , where α=0.00739, β=2.45, 100μm<D≤3000μm, the mass of a single snow crystal is calculated; S052). Mass calculation of graupel A volume model with an axial ratio of 0.9 is adopted to calculate the volume of graupel V= ; Combined with the mass-particle size power-law relationship of graupel m=α· , where α=0.049, β=2.8, 500μm<D≤3000μm, the mass of a single graupel particle is calculated.
[0014] Further, the specific process of step S06 is: solid precipitation calculation: based on the experimental calculation formula of solid precipitation, the microphysical parameters of particles are substituted to calculate the solid precipitation intensity R, with the unit of mm / h; the formula is: Where: ρw is the density of water, the unit of water density is 1 mg / mm³, Vsample is the effective volume of holographic sampling, N is the product of sampling time and camera frame rate, k is the total number of observed effective solid precipitation particles, m(Di) is the mass of the i-th particle, and vi is the terminal falling velocity of the i-th particle; the calculated precipitation intensity R is accumulated according to the observation time, so as to obtain the total depth of solid precipitation within a specified period.
[0015] Further, step S07 specifically includes: S071). Data integration: uniformly integrate all data obtained in the above steps, and the core integrated data includes: solid precipitation intensity, with the unit of mm / h; total depth of solid precipitation in a specified period, with the unit of mm; the classified statistical quantity of snow / graupel / hail, equivalent diameter of particles, terminal falling velocity, number concentration spectrum, diameter-velocity spectrum, mass of a single particle, and holographic reconstruction image of particles; S072). Standardization processing: perform format standardization on the integrated data, unify data units and statistical dimensions, and eliminate data redundancy; S073). Data output and archiving: export the standardized data set into general data formats such as Excel and CSV, and generate a solid precipitation observation report at the same time, which contains contents such as precipitation values, particle spectrum distribution charts, and evolution characteristics of the precipitation process, so as to facilitate subsequent data query, analysis, sharing and secondary processing, and provide comprehensive data support for meteorological scientific research, weather modification, disaster reduction and prevention.
[0016] Compared with existing solid precipitation measurement technologies, this invention has the following technical advantages: Advantage 1: This application sets classification thresholds based on the morphological characteristics (roundness, axial ratio, number of edges) of snow, sleet, and hail, and achieves accurate classification by combining the three-dimensional morphology of particles reconstructed by digital holography. It also adapts exclusive thickness-diameter, mass-particle-size power-law relationships and volume calculation formulas for different types of particles. This solves the problems of existing technologies being unable to distinguish between solid precipitation particle types and using empirical formulas uniformly, which leads to large calculation errors. The relative error of single particle mass calculation is reduced to within 10%, laying the foundation for accurate calculation of solid precipitation, realizing accurate classification of different types of solid precipitation particles, and significantly reducing particle mass calculation errors.
[0017] Advantage 2: This application combines digital holographic imaging technology and single-exposure multiple-sampling technology to simultaneously extract core microphysical parameters of particles, such as equivalent diameter, morphological characteristics, terminal velocity, and number concentration, achieving integrated, non-contact measurement of particle parameters. It solves the problem that existing technologies can only extract single particle size parameters, leading to insufficient basis for precipitation calculation due to parameter gaps. This provides complete parameter support for solid precipitation calculation, ensuring the theoretical scientific validity and accuracy of the calculation results. It has the advantage of simultaneously acquiring multi-dimensional microphysical parameters, achieving full parameter support for solid precipitation calculation.
[0018] Advantage 3: This application, while measuring the total amount of solid precipitation, also statistically analyzes the number concentration and terminal velocity distribution characteristics of particles of different sizes according to snow, graupel, and hail, and plots particle spectrum distribution curves. This solves the problem of existing technologies only outputting overall precipitation values and lacking microscopic data support. It not only enables the measurement of solid precipitation but also provides microscopic data support for the mechanism analysis of solid precipitation weather processes, the prediction of precipitation intensity evolution, and the evaluation of the effectiveness of artificial weather modification operations, thus enhancing the application value of solid precipitation observation. It has the advantage of simultaneously acquiring microscopic spectrum distribution data of solid precipitation, providing multidimensional support for weather process analysis.
[0019] Advantage 4: This application realizes the automatic collection, processing, calculation and output of solid precipitation particles based on digital holographic imaging equipment, without the need for manual operation; it solves the problems of traditional weighing method and rain gauge conversion algorithm relying on manual labor and unable to conduct continuous observation, and can realize 24-hour uninterrupted solid precipitation observation. It can accurately capture the spatiotemporal evolution characteristics of short-term and local solid precipitation processes, avoid the loss of key observation data, realize continuous real-time automatic observation of solid precipitation, and overcome the defects of manual dependence of traditional methods. Detailed Implementation
[0020] The present invention will be further described below with reference to the embodiments. However, those skilled in the art will understand that the following embodiments are only for illustrating the present invention and should not be regarded as limiting the scope of the present invention.
[0021] To achieve the objectives of this invention, a method for measuring solid precipitation based on digital holography is provided. The technical principle is as follows: its core technologies are supported by digital holographic imaging theory, single-exposure multiple-sampling technology, solid precipitation particle classification and recognition algorithms, and a precise solid precipitation calculation model. These principles work together to achieve high-precision measurement of solid precipitation, as specifically demonstrated below: 1. Digital holographic imaging theory: In a coaxial digital holographic system, after a plane wave irradiates solid precipitation particles, the diffracted light of the particles interferes with the undisturbed plane light to form a hologram. After being recorded by a CMOS camera, the three-dimensional spatial morphology of the particles is restored through numerical reconstruction, providing a basis for particle type identification and microphysical parameter extraction. 2. Single exposure multiple sampling technology: During a single exposure of the camera, multiple pulses are emitted by the pulse laser triggered by the rising edge, capturing the motion trajectory of the same particle at different moments in a hologram, and accurately calculating the final velocity of the particle's fall; 3. Solid precipitation particle classification and recognition algorithm: Based on the morphological characteristics (such as roundness, angularity, and axial ratio) and microphysical characteristics of snow, sleet, and hail, a classification threshold is set to accurately classify the particles after digital holographic reconstruction, providing a basis for subsequent adaptation to dedicated empirical formulas; 4. Precise Calculation Model for Solid Precipitation: Combining the thickness-diameter and mass-particle size power law relationships and effective density calculation formulas of different types of solid precipitation particles, the mass of a single particle is calculated. Then, based on the particle number concentration and the terminal velocity of the fall, the model is summed using the experimental calculation formula for solid precipitation to obtain a high-precision value of solid precipitation. At the same time, the distribution characteristics of different types and particle sizes of particles are statistically analyzed to form particle spectrum distribution data.
[0022] Here is an example: This embodiment proposes a method for measuring solid precipitation based on digital holography. The method acquires holograms of solid precipitation particles using a digital holographic imaging device. After preprocessing, particle reconstruction, classification and identification, and extraction of microphysical parameters, the particle mass is calculated using empirical formulas specific to different types of solid precipitation particles. Finally, the method achieves accurate measurement of precipitation through a solid precipitation calculation model, while also obtaining the spectral distribution characteristics of solid precipitation particles.
[0023] Specifically, the following steps are included: Step S01: Deployment and parameter configuration of holographic imaging equipment 1. Equipment deployment: Place the holographic hail spectrometer (HHM) at the designated location for solid precipitation observation, ensuring that the instrument is securely installed, the lens is unobstructed and pointed at the core area where solid precipitation particles fall, and avoid interference from external factors such as stray light, buildings, and vegetation. 2. Equipment Start-up and Parameter Configuration: Connect the power supply to the holographic hail spectrometer, start the instrument control system and CMOS camera acquisition function, and preset parameters such as camera frame rate, exposure time, pulse laser pulse frequency and wavelength according to the observation requirements of solid precipitation. The pulse frequency is set according to the typical falling velocity of solid precipitation particles. For example, set the camera frame rate to 30 frames / second, the exposure time to 3100μs, the pulse laser wavelength to 638nm, the pulse width to 150ns, the number of pulses to 4, and the pulse interval to 1000μs. The pulse frequency is set according to the typical falling velocity of solid precipitation particles (for example, the falling velocity of hail is usually in the range of 10m / s to 35m / s, so the pulse interval is set to 1000μs; the falling velocity of sleet is 1m / s to 5m / s, so the pulse interval is set to 685μs) to ensure that the effective motion trajectory of the particles can be captured in a single exposure. 3. Data Acquisition and Storage: Start the hologram acquisition program. The CMOS camera continuously captures holograms of solid precipitation particles during their fall according to the set parameters. The acquired hologram data is stored in real time to the instrument's built-in storage module.
[0024] Step S02: Holographic preprocessing and particle 3D reconstruction 1. Hologram Preprocessing: Each acquired hologram frame undergoes denoising, background subtraction, and contrast enhancement. A median filtering algorithm (a classic non-linear denoising algorithm proposed by Tukey in 1971 and included in "Digital Image Processing (Third Edition)," a mature existing image processing technique; algorithm implementation: selecting all pixel grayscale values within a 3×3 neighborhood of the target pixel, sorting them from smallest to largest, and replacing the center pixel's grayscale value with the sequence median; this scheme is used to filter out salt-and-pepper pulse noise from the CMOS sensor during holographic imaging, removing isolated noise points while preserving the edge contours of solid precipitation particles to the greatest extent) is employed to eliminate image noise, through background... The difference method (a classic algorithm for motion foreground extraction in computer vision, whose basic principles are fully described in the 2006 paper "Moving Target Detection Method Based on Background Difference and Inter-Frame Difference" in the Journal of Instrumentation) removes the reference background without particles, improving the clarity of the hologram and ensuring that the particle outline can be accurately identified. This method is implemented by taking four consecutively acquired images as a group and subtracting two images by skipping one image (e.g., subtracting the third image from the first, and the fourth image from the second). A fixed grayscale threshold is set; areas exceeding the threshold are identified as particle foreground, and the remaining pixels are set to zero to remove the background. This is a common preprocessing method in the field of optical particle detection. 2. Particle 3D Reconstruction: Based on digital holographic imaging theory, numerical reconstruction is performed on the preprocessed hologram using a diffraction field restoration algorithm (this algorithm is a mature and widely used existing reconstruction technique in the field of digital holography, theoretically based on the Huygens-Fresnel diffraction principle, first proposed by Goodman & Lawrence in 1967). The digital holographic numerical reconstruction approach was proposed in 2008. The core principle of the algorithm is that holographic fringes acquired by CMOS record the amplitude information of the diffraction light field of solid precipitation particles. Based on Fresnel diffraction integrals and Fast Fourier Transform (FFT), the three-dimensional light field distribution of the object is inversely deduced, converting the two-dimensional holographic interference data into complex amplitudes at various positions in particle space. In practical application, according to the instrument's laser wavelength (638nm in this device) and the recorded diffraction distance, multi-layer numerical diffraction iterative calculations are performed in segments to reconstruct the light field of different depth sections of the particle layer by layer, stitching together a complete three-dimensional outline. This algorithm has been widely applied to the holographic detection of aerosols and water droplets' microscopic morphology. It calculates the three-dimensional spatial morphology of solid precipitation particles, obtaining two-dimensional projection images and three-dimensional outline data of the particles, providing a foundation for subsequent particle identification and parameter extraction.
[0025] Step S03: Precise Identification and Classification of Solid Precipitation Particles 1. Particle Feature Extraction: Image analysis was performed on the reconstructed two-dimensional projection image of the particles to extract the morphological feature parameters of each particle, including roundness, axial ratio (short axis to long axis ratio), number of contour edges, and equivalent diameter. Among them, snow crystals are hexagonal / dendritic, with low roundness, large axial ratio differences, and many edges; graupel is spherical / conical, with moderate roundness (0.8-1.0) and an axial ratio of about 0.9. 2. Particle Classification and Recognition: Based on the morphological differences between snow and sleet, a classification threshold is set. Particles with a roundness < 0.8, number of edges ≥ 6, and axial ratio > 1.2 are classified as snow crystals (including hexagonal, star-shaped, and dendritic snow crystals); particles with a roundness ≤ 0.8, axial ratio ≈ 0.9, and number of edges < 3 are classified as graupel. 3. Classification result storage: Store the corresponding number, classification result, and morphological feature parameters of each particle to form a solid precipitation particle classification dataset.
[0026] Step S04: Synchronous extraction of multi-dimensional microphysical parameters 1. Based on particle classification results and 3D reconstruction data, microphysical parameters are extracted for snow, sleet, and hail, and core parameters such as particle equivalent diameter D, terminal velocity v, and number concentration N(D) are obtained simultaneously. All parameters are stored according to particle number and type, forming a complete particle microphysical parameter dataset. 2. Equivalent diameter extraction: Based on the particle's projected area, the equivalent diameter D of the particle is calculated using the circular equivalence method, that is, the particle's projected area is equal to the diameter of a circle with the same area; 3. Calculation of terminal falling velocity: Based on the single-exposure multi-sampling technology, the motion trajectory of the same particle in the hologram is extracted, the displacement and time difference of the particle between adjacent pulses are calculated to obtain the falling velocity of the particle, and the average value obtained after multiple samplings is taken as the terminal falling velocity v of the particle (unit: mm / s); 4. Statistics of particle number concentration: According to the effective sampling volume of the holographic hail spectrometer , the number of particles of different types and in different particle size intervals per unit volume is counted to obtain the particle size distribution function N (D) (unit: ).
[0027] Step S05: Accurate calculation of individual mass of different types of particles Aiming at the differences in microphysical characteristics of snow, graupel and hail, exclusive empirical formulas are respectively adopted to calculate the mass m(D) (unit: mg) of a single particle, which is mainly based on the thickness-diameter power law relationship, volume calculation formula and mass-particle size power law relationship of solid precipitation particles, so as to ensure the accuracy of mass calculation: 1. Calculation of snow crystal mass First, according to the type of snow crystal, the corresponding thickness-diameter power law relationship h = A· (A and B are empirical regression coefficients, for example, A=2.020 and B=0.449 for hexagonal plates), the snow crystal thickness h is calculated; The volume model of the circumscribed flat prism is adopted to calculate the snow crystal volume V≈Area×h (Area is the projected area of the particle); combined with the mass-particle size power law relationship of snow crystals m=α· (α=0.00739, β=2.45, 100μm<D≤3000μm), the mass of a single snow crystal is calculated; 2. Calculation of graupel mass A volume model with an axial ratio of 0.9 is adopted to calculate the graupel volume V= ; Combined with the mass-particle size power law relationship of graupel m=α· (α=0.049, β=2.8, 500μm<D≤3000μm), the mass of a single graupel particle is calculated; Step S06: Accurate calculation of solid precipitation Calculation of solid precipitation: Based on the experimental calculation formula of solid precipitation, the microphysical parameters of particles are substituted to calculate the solid precipitation intensity R (unit: mm / h), and the formula is as follows: Where: ρw is the density of water (1 mg / mm³), Vsample is the effective volume of holographic sampling (known as 134640 mm³), N is the sampling time × camera frame rate (known as 1800), k is the total number of observed effective solid precipitation particles, m(Di) is the mass of the i-th particle, and vi is the final velocity of the i-th particle; by summing the calculated precipitation intensity R over the observation time, the total depth of solid precipitation within the specified time period can be obtained; Step S07: Data Integration and Standardized Output 1. Data Integration: All data obtained in the above steps are integrated in a unified manner. The core data of the integration include: solid precipitation intensity (mm / h), total solid precipitation depth (mm) in a specified period, statistical number of snow / graupel / hail classifications, particle equivalent diameter, terminal velocity, number concentration spectrum, diameter-velocity spectrum, mass of individual particles, and particle holographic reconstruction image. 2. Standardization Processing: Standardize the format of the integrated data, unify data units and statistical dimensions, and eliminate data redundancy; 3. Data Output and Archiving: The standardized dataset is exported to common data formats such as Excel and CSV, and a solid precipitation observation report is generated, including precipitation values, particle spectrum distribution charts, precipitation process evolution characteristics, etc., which facilitates subsequent data query, analysis, sharing and secondary processing, and provides comprehensive data support for meteorological scientific research, weather modification, and disaster reduction and prevention.
[0028] The preferred embodiments of the present invention disclosed above are only for illustrating the present invention. These preferred embodiments do not describe all details exhaustively, nor do they limit the invention to specific implementations. Obviously, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention.
Claims
1. A method for measuring solid precipitation based on digital holography. The method acquires holograms of solid precipitation particles using a digital holographic imaging device. After preprocessing, particle reconstruction, classification and identification, and extraction of microphysical parameters, the particle mass is calculated using empirical formulas specific to different types of solid precipitation particles. Finally, the method achieves accurate measurement of precipitation through a solid precipitation calculation model, while also obtaining the spectral distribution characteristics of solid precipitation particles. Specifically, the following steps are included: Step S01: Collect and store holograms of solid precipitation particles during their descent. Step S02: Holographic preprocessing and particle 3D reconstruction; Step S03: Accurate identification and classification of solid precipitation particles to form a solid precipitation particle classification dataset; Step S04: Simultaneous extraction of multi-dimensional microphysical parameters to obtain the particle spectral distribution function N(D); Step S05: Calculate the individual mass of different types of particles; Step S06: Calculate solid precipitation; Step S07: Data integration and standardized output.
2. The method of claim 1, wherein, The specific process of step S01 is to place the holographic hail spectrometer at the designated location for solid precipitation observation, start the hologram acquisition program, and continuously capture holograms of solid precipitation particles falling according to the set parameters. The acquired hologram data is then stored in real time to the instrument's built-in storage module.
3. The method of claim 2, wherein, Step S02 specifically includes: S021). Hologram preprocessing: Denoising, background subtraction, and contrast enhancement are performed on each frame of the acquired hologram. Median filtering algorithm is used to eliminate image noise, and background subtraction method is used to subtract the particle-free reference background to improve the clarity of the hologram and ensure that the particle outline can be accurately identified. S022). Particle 3D Reconstruction: Based on digital holographic imaging theory, the preprocessed hologram is numerically reconstructed, and the 3D spatial morphology of solid precipitation particles is calculated by the diffraction light field restoration algorithm to obtain the 2D projection image and 3D contour data of the particles, providing a basis for subsequent particle identification and parameter extraction.
4. The method for measuring solid precipitation according to claim 3, characterized in that, Step S03 specifically includes: S031). Particle Feature Extraction: Image analysis is performed on the reconstructed two-dimensional projection image of the particles to extract the morphological feature parameters of each particle, including roundness, axial ratio, number of contour edges, and equivalent diameter. Among them, snow crystals are hexagonal / dendritic, with low roundness, large difference in axial ratio, and many edges; graupel is spherical / conical, with roundness between 0.8 and 1.0 and axial ratio of about 0.
9. S032). Particle classification and recognition: Based on the morphological differences between snow and sleet, a classification threshold is set: Particles with a roundness < 0.8, number of edges ≥ 6, and axial ratio > 1.2 are classified as snow crystals; particles with a roundness ≤ 0.8, axial ratio ≈ 0.9, and number of edges < 3 are classified as graupel. S033). Classification result storage: Store the corresponding number, classification result, and morphological feature parameters of each particle to form a solid precipitation particle classification dataset.
5. The method for measuring solid precipitation according to claim 4, characterized in that, Step S04 specifically includes: S041). Based on particle classification results and 3D reconstruction data, microphysical parameters were extracted for snow, sleet, and hail, and the equivalent particle diameter D, terminal velocity v, and number concentration N(D) were obtained simultaneously. All parameters were stored according to particle number and type, forming a complete particle microphysical parameter dataset. S042). Equivalent diameter extraction: Based on the particle's projected area, the equivalent diameter D of the particle is calculated using the circular equivalence method, that is, the particle's projected area is equal to the diameter of a circle with the same area; S043). Calculation of terminal velocity: Based on the single-exposure multiple sampling technique, the trajectory of the same particle in the hologram is extracted, the displacement and time difference of the particle between adjacent pulses are calculated, and the falling velocity of the particle is obtained. After multiple samplings and taking the average value, it is taken as the terminal velocity v of the particle. The velocity unit is mm / s. S044). Particle number concentration statistics: based on the effective sampling volume of the holographic hail spectrometer. The particle spectral distribution function N(D) is obtained by counting the number of particles of different types and particle size ranges per unit volume. .
6. The method for measuring solid precipitation according to claim 5, characterized in that, Step S05 specifically includes: To address the differences in the microphysical characteristics of snow, sleet, and hail, specific empirical formulas are used to calculate the mass m(D) of a single particle, in mg. The core principles are based on the power-law relationship between thickness and diameter of solid precipitation particles, the volume calculation formula, and the power-law relationship between mass and particle size, ensuring the accuracy of the mass calculation. S051). Snow Crystal Mass Calculation First, based on the snow crystal type, apply the corresponding thickness-diameter power-law relationship h = A· A and B are empirical regression coefficients used to calculate the snow crystal thickness h. Using a flat prism outer volume model, the snow crystal volume V≈Area×h is calculated, where Area is the particle's projected area. Combining the mass-particle size power law relationship m=α· of snow crystals , where α=0.00739, β=2.45, and 100μm<D≤3000μm, the mass of a single snow crystal is calculated; S052). Calculation of the mass of graupel. Using a volumetric model with an axial ratio of 0.9, the volume V of the graupel is calculated. ; Combined with the mass-particle size power law relationship of graupel m=α· , where α=0.049, β=2.8, 500μm<D≤3000μm, the mass of a single graupel particle is calculated.
7. The method for measuring solid precipitation according to claim 6, characterized in that, The specific process of step S06 is as follows: Calculation of solid precipitation: Based on the experimental calculation formula for solid precipitation, substitute the particle microphysical parameters to calculate the solid precipitation intensity R, unit: mm / h; the formula is: Where: ρw is the density of water, with the unit of water density being 1 mg / mm³, Vsample is the effective volume of holographic sampling, N is the sampling time × camera frame rate, k is the total number of observed effective solid precipitation particles, m(Di) is the mass of the i-th particle, and vi is the final velocity of the i-th particle; by accumulating the calculated precipitation intensity R over the observation time, the total depth of solid precipitation within the specified time period can be obtained.
8. The method for measuring solid precipitation according to claim 7, characterized in that, Step S07 specifically includes: S071). Data integration: All data obtained in the above steps are integrated in a unified manner. The core data of the integration include: solid precipitation intensity, in mm / h; total solid precipitation depth in a specified period, in mm; classification statistics of snow / graupel / hail, particle equivalent diameter, terminal velocity, number concentration spectrum, diameter-velocity spectrum, mass of a single particle, and particle holographic reconstruction image. S072). Standardization processing: Standardize the format of the integrated data, unify data units and statistical dimensions, and eliminate data redundancy; S073). Data output and archiving: Export the standardized dataset to common data formats such as Excel and CSV, and generate a solid precipitation observation report, including precipitation values, particle spectrum distribution charts, and precipitation process evolution characteristics.
