Inversion method for detecting cloud particle physical characteristics by satellite-borne multi-wavelength laser radar

Abstract

The invention discloses a spaceborne multi-wavelength laser radar detection cloud particle physical property inversion method. The method comprises the following steps: solving the altitude height and the surface height of each point of an echo signal; de-noising the echo signal; correcting the pre-processed signal to obtain an attenuated backscattering signal and identifying a cloud boundary; establishing a molecular transmittance lookup table, and calculating transmittance and atmospheric molecular parameters; the back scattering coefficient of the cloud is obtained through inversion according to the attenuation back scattering signal, the transmittance lookup table, the aerosol and cloud particle transmittance and the atmospheric molecule parameters; the optical thickness, the extinction coefficient and the like of the cloud are calculated by using the attenuation backscattering signal, the cloud particle and atmospheric molecule transmittance and the atmospheric molecule parameters; calculating the depolarization ratio of the cloud; calculating an integral backscattering coefficient and an integral depolarization ratio of the cloud by using the cloud boundary information; and identifying a cloud particle phase state. The method is characterized in that cloud boundary identification is carried out on spaceborne laser radar echo signals, and cloud optical parameters are solved accurately and efficiently.

Description

Technical Field

[0001] The present invention belongs to the field of atmospheric environment satellite optical remote sensing, and specifically is an inversion method for detecting the physical characteristics of cloud particles using a satellite-borne multi-wavelength lidar. Background Art

[0002] Clouds cover approximately 75% of the Earth's surface and play a crucial role in Earth's radiation balance and global and regional water cycles. They are both the most crucial and the most uncertain component of weather forecasting and climate change. Clouds cool the Earth-atmosphere system by scattering and absorbing solar radiation, reducing the amount of shortwave radiation reaching the Earth. They also absorb longwave radiation emitted by the Earth's surface and the atmosphere beneath them, while simultaneously emitting longwave radiation downward, warming the Earth-atmosphere system. Research indicates that this radiative forcing exceeds the currently known radiative forcing from factors such as greenhouse gases and aerosols. The spatiotemporal variations in physical parameters such as cloud volume, geometric thickness, and cloud top height at the macroscopic level, and the phase state of cloud particles at the microscopic level, directly influence Earth's energy budget. Therefore, a deeper understanding of the macro- and microphysical properties of clouds, and the parameterization of cloud physical transmission and transformation processes, are crucial for studying climate evolution and improving weather forecasts.

[0003] LiDAR, an active remote sensing technology integrating optics, mechanics, and electronics, has experienced rapid development and widespread application in the past two decades. Its ability to detect vertical profiles, its all-day operation, high detection sensitivity, and high spatiotemporal resolution make it a unique advantage in long-term, continuous, three-dimensional observations of clouds, making it a crucial detection method in current research on cloud physical properties. Passive remote sensing has significant uncertainty in cloud height detection and makes it difficult to obtain the vertical structure of clouds. Existing airborne LiDARs are expensive and significantly affected by the observation environment. The observation time and available observation samples are limited, making them difficult to operate continuously. They are generally used as a scaled-down verification method for spaceborne LiDARs. Ground-based LiDARs can achieve continuous observations, but ground-based observation stations are limited in distribution and can only reveal the cloud physical properties of local areas. This is insufficient to support large-scale or even global research on the macro and micro physical properties of clouds.

[0004] Patent document CN112904308 discloses a method for detecting cloud phase and cloud water content using ground-based Raman lidar. However, this patented technology addresses the data processing and inversion of Raman lidar. First, it is not suitable for data inversion from spaceborne platforms, and the data inversion range is far lower than that of spaceborne lidar. Patent document CN119126054A discloses a method for inverting atmospheric particles using spaceborne hyperspectral detection lidar. However, this patent primarily addresses the preliminary inversion of cloud and aerosol optical parameters, without further addressing physical quantities such as cloud particle phase and cloud height information. Summary of the Invention

[0005] The purpose of the present invention is to provide an inversion method for detecting the physical properties of cloud particles using a spaceborne multi-wavelength lidar, so as to overcome the current technical defect of difficulty in conducting high-precision continuous observation of clouds throughout the day, and to achieve high-precision acquisition of optical parameters of clouds with vertical structures on a global scale, precise identification of cloud boundaries and cloud top heights, and accurate judgment of the phase state of cloud particles.

[0006] The technical solution adopted by the present invention is a method for inverting the physical properties of cloud particles detected by a spaceborne multi-wavelength lidar, comprising:

[0007] S1. Use a spaceborne multi-wavelength lidar system to emit 532nm and 1572nm pulsed lasers to acquire backscattered echo signals from interactions with atmospheric molecules, aerosols, and cloud particles.

[0008] S2. Based on the satellite platform coordinates, the longitude and latitude of the laser pointing point, and the acquisition delay time of the measurement point, combined with the elevation data of the laser radar pointing point, the altitude corresponding to each measurement point in the echo signal is calculated;

[0009] S3. The echo signal is preprocessed by sliding average and wavelet denoising to obtain a preprocessed echo signal;

[0010] S4. Perform distance correction and system constant correction on the preprocessed echo signal to obtain the attenuated backscattering signals of the 532nm vertical channel, parallel channel, hyperspectral molecular channel, and 1572nm channel;

[0011] S5. From the attenuated backscatter signal of the 1572 nm channel, threshold screening and differential processing are used to identify the upper and lower boundaries of the cloud layer. The cloud boundary locations are then correlated with the calculated altitude corresponding to the measurement point to obtain cloud height information.

[0012] S6. Based on the temperature and pressure profiles provided by the numerical weather prediction model reanalysis meteorological dataset, the S6 atmospheric model, and the absorption spectrum of the iodine molecular filter, a molecular transmittance lookup table was established at 532 nm to calculate aerosol and cloud particle transmittance and atmospheric molecular parameters.

[0013] S7. Invert the cloud backscattering coefficient using the attenuated backscattering signal from the 532 nm vertical channel, parallel channel, and hyperspectral molecular channel, the molecular transmittance lookup table, the aerosol and cloud particle transmittance, and atmospheric molecular parameters.

[0014] S8. Calculate the cloud optical thickness using the attenuated backscatter signal from the 532 nm vertical channel, parallel channel, and hyperspectral molecular channel, aerosol and cloud particle transmittance, and atmospheric molecular parameters. Obtain the extinction coefficient by differentiating the optical thickness.

[0015] S9. Calculate the integrated backscattering coefficient and integrated depolarization ratio of the cloud layer using the cloud boundary information, the attenuated backscattering signals of the 532nm vertical and parallel channels, and the cloud backscattering coefficient;

[0016] S10. Identify the cloud particle phase based on the correlation between the integrated backscattering coefficient and the integrated depolarization ratio of the cloud layer.

[0017] Compared with the prior art, the present invention has the following beneficial effects:

[0018] This invention combines lidar technology with satellite platforms to acquire atmospheric satellite observation data, including cloud data, enabling large-scale, long-range, all-day, and high-precision cloud detection. The altitude of each lidar sampling point and the surface altitude of the pointing point are decomposed based on this satellite observation data, improving the accuracy of altitude calculation. Based on this satellite observation data, the invention uses 1572nm channel data combined with a differential method and area integral threshold screening to extract cloud signals and further identify cloud boundary heights (cloud top and cloud base). This allows for accurate differentiation of aerosol and cloud signals, significantly improving the cloud detection range and cloud top height detection accuracy compared to ground-based and airborne detection methods. The present invention obtains the optical thickness, backscattering coefficient, extinction coefficient and depolarization ratio of cloud particles by inverting the satellite observation data, calculates the integral depolarization ratio and the integral backscattering coefficient using the cloud layer identification result, and identifies the cloud particle phase state based on the correlation between the integral backscattering coefficient and the integral depolarization ratio of the cloud layer. Compared with the method of directly discriminating the cloud phase state based on the depolarization ratio, the method takes into account the multiple scattering effect of spherical cloud particles and the influence of horizontally oriented ice cloud particles on the depolarization ratio of the cloud, thereby improving the accuracy of cloud particle phase state discrimination. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a specific flow diagram of the present invention.

[0020] Figure 2 Schematic diagram of the laser radar pointing point.

[0021] Figure 3 This is the relationship between the cloud integrated depolarization ratio and the integrated backscattering ratio. DETAILED DESCRIPTION

[0022] In order to more clearly illustrate the inversion method for detecting the physical characteristics of cloud particles using a spaceborne multi-wavelength lidar, the present invention is further described below with reference to examples and accompanying drawings. Figure 1 A schematic diagram of the specific flow of the algorithm of the present invention is given.

[0023] The specific embodiments of the present invention are as follows

[0024] (1) Obtaining spaceborne multi-wavelength lidar data

[0025] The satellite-borne multi-wavelength lidar system mainly consists of a laser transmitting system and a receiving system. The satellite-borne lidar transmits pulsed lasers with wavelengths of 532nm and 1572nm, which interact with atmospheric molecules, aerosols and cloud particles. The backscattered echo signal containing the detection target is detected at the receiving end. The echo signal obtained at the 532nm wavelength includes the vertical channel signal. Parallel channel signals and hyperspectral molecular channel signals The total echo signal P is obtained at a wavelength of 1572nm 1572 (r).

[0026] (2) Calculate the altitude corresponding to each measurement point in the echo signal and the surface height of the laser radar pointing point

[0027] Using the satellite origin information in the WGS-84 coordinate system, the latitude, longitude, and elevation information of the satellite-borne lidar pointing point, and the acquisition delay time information of each measurement point, the detection distance r from the satellite to the sampling point, the distance l from each pointing point to the satellite, and the altitude h of each measurement point are calculated. The calculation formula for the detection distance r from the satellite to the sampling point is as follows:

[0028] r=(T d -T a +N / DA)·c / 2#(1)

[0029] Among them, T d T is the delay time for collecting the measurement point signal. a is the inherent time difference between the trigger pulse of the lidar system and the laser emission, N is the acquisition point of the measurement point, DA is the echo sampling frequency, and c is the speed of light in vacuum;

[0030] The distance calculation formula from each pointing point to the satellite is:

[0031]

[0032] Among them, X, Y, Z are the pointing point positions, X1, Y1, Z1 are the satellite position coordinates in the WGS-84 coordinate system;

[0033] The DEM data provided by the satellite is used to calculate the elevation d of the laser radar pointing point and the distance l from the pointing point to the satellite. The altitude corresponding to each measurement point is further calculated. The specific relationship is as follows: Figure 2 The calculation formula for the altitude h of each measuring point is:

[0034] h=l+dr#(3)

[0035] Where d is the elevation of the pointing point.

[0036] (3) Preprocessing the echo signal

[0037] The echo signals are averaged and superimposed to extract the signals. The vertical average times of the 532nm echo signal are 2 times, the horizontal average times are 59 times, and the vertical average times of the 1572nm echo signal are 16 times, and the horizontal average times are 59 times. The echo signals with a vertical resolution of 48m and a horizontal resolution of 20km are obtained. The echo signals include all the echo signals described in step (1). The echo signals at the resolution are subjected to wavelet denoising. The 532nm echo signal adopts sym4 and the 1572nm echo signal adopts sym7 to further improve the signal-to-noise ratio.

[0038] (4) Perform distance and system constant correction on the preprocessed signal

[0039] Based on the pre-processed data, the distance correction and system constant correction of the signal are performed to obtain the 532nm vertical channel attenuated backscattered signal. Parallel channels attenuate backscattered signals and hyperspectral molecular channel attenuation backscattering signal 1572nm attenuated backscattered signal B 1572 (r), calculated as follows:

[0040]

[0041] Among them, r is the detection distance, E 532 and E 1572 represent the laser pulse energies of 532 nm and 1572 nm respectively, and C1, C2, C3, and C4 are the system constants of the 532 nm vertical channel, parallel channel, hyperspectral molecular channel, and 1572 nm detection channel in the lidar equation respectively.

[0042] (5) Obtain preliminary identification of cloud layer, cloud boundary and cloud height information from the attenuated backscatter signal

[0043] Using the attenuated backscattered signal B at 1572nm 1572 (r) Identify clouds and remove noise and aerosol signals using the area integral threshold method. The specific method is as follows:

[0044] ∫B 1572 (r)>2.0e -5 #(8)

[0045] Calculate the 1572nm attenuated backscattered signal B 1572 The integral of (r) is set to a threshold of 2.0e -5, the signal greater than this value is determined to be cloud, and the signal less than this value is determined to be aerosol; then the differential method is used to process the filtered 1572nm attenuated backscattered signal B 1572 (r), the calculation formula of the differential signal is as follows:

[0046]

[0047] Find the first zero-crossing point of the 1572nm differential signal from top to bottom and from bottom to top, as well as the global maximum point within the cloud layer, determine the cloud top, cloud bottom, and cloud peak positions, and obtain preliminary identification of the cloud boundary and cloud peak position information. Associate the cloud boundary position with the calculated altitude corresponding to the measurement point to obtain cloud height information.

[0048] (6) Establish a molecular transmittance lookup table in the 532nm band to calculate the aerosol and cloud particle transmittance and atmospheric molecular parameters

[0049] The temperature and pressure profile data provided by the numerical weather forecast model reanalysis meteorological data set and the S6 atmospheric model are used to calculate the Rayleigh scattering spectrum of molecules in the 532nm band. The transmittance r of the molecules in the 532nm band is obtained by convolution calculation with the absorption spectrum of the iodine molecular filter measured in the laboratory. mol The transmittance T of aerosols and cloud particles can be approximately equal to the result of the convolution of the laser's Gaussian linear spectrum and the iodine molecule filter's absorption spectrum; then, the atmospheric molecular parameters are calculated based on the Rayleigh scattering theory, including the atmospheric molecule's backscattering coefficient β. mol , extinction coefficient α mol and depolarization ratio δ mol .

[0050] (7) Calculation of backscatter coefficient

[0051] The cloud backscattering coefficient β(λ 532 ,r), the formula is as follows:

[0052]

[0053] Where, β mol (r) is the backscattering coefficient of atmospheric molecules at the detection distance r, r mol (r) is the transmittance of the molecular Rayleigh scattering signal at the detection distance r, r(λ 521 ,r) is the Mie scattering signal transmittance of clouds and aerosol particles at the detection distance r; δ mol is the depolarization ratio of atmospheric molecules, δ(λ 532r) is the total depolarization ratio at the detection distance r under the 532 nm wave band, indicating the ratio of the attenuated backscattering signals of the vertical channel and the parallel channel, and K(r) is the ratio of the attenuated backscattering signals of the parallel channel and the hyperspectral channel at the detection distance r under the 532 nm wave band.

[0054] (8) Calculation of optical thickness and extinction coefficient and other optical parameters

[0055] The optical thickness, extinction coefficient and other optical parameters of the cloud are retrieved by using the 532 nm attenuated backscattering signal, the transmittance of aerosol and cloud particles and atmospheric molecules and the atmospheric molecule parameters, and the backscattering coefficient of the cloud is calculated by using the atmospheric molecule parameters and the attenuated backscattering coefficient of the hyperspectral channel. The optical thickness τ(r) of the aerosol and cloud is calculated, and the formula is as follows:

[0056]

[0057] In the formula, β mol (r) is the backscattering coefficient of atmospheric molecules at the detection distance r, T mol (r) is the transmittance of the molecular Rayleigh scattering signal at the detection distance r, T(λ 532 (r) is the transmittance of the Mie scattering signal of the cloud and aerosol particles at the detection distance r; δ mol is the depolarization ratio of atmospheric molecules, and K(r) is the ratio of the attenuated backscattering signals of the parallel channel and the hyperspectral channel at the detection distance r under the 532 nm wave band, is the attenuated backscattering coefficient of the hyperspectral channel;

[0058] The extinction coefficient α(r) of the cloud particles is calculated according to the differential of the optical thickness and the atmospheric molecule extinction coefficient α mol (r), and the formula is as follows:

[0059]

[0060] (9) Calculation of integrated backscattering coefficient and integrated depolarization ratio

[0061] The integrated backscattering coefficient and the integrated depolarization ratio of the cloud layer are calculated by using the cloud layer cloud boundary information, the attenuated backscattering signals of the 532 nm vertical channel and parallel channel, and the backscattering coefficient of the cloud, and the depolarization ratio δ cloud (r) of the cloud particles is calculated from the attenuated backscattering signals of the vertical channel and the parallel channel, and the formula is as follows:

[0062]

[0063] The integrated backscattering coefficient Y' and the integrated depolarization ratio δ' of the cloud layer are calculated according to the extracted cloud layer boundary information, the backscattering coefficient of the cloud and the depolarization ratio of the cloud particles, and the calculation formula is as follows:

[0064]

[0065] (10) Identification of cloud particle phase

[0066] The phase state of cloud particles can be identified based on the correlation between the integrated backscattering coefficient and the integrated depolarization ratio of the cloud layer: Figure 3 As shown in Figure 2, the particles whose Υ′ decreases with the increase of δ′ are water cloud particles, and vice versa, they are ice cloud particles.

Claims

1. A method for inverting cloud particle physical properties detected by spaceborne multi-wavelength lidar, characterized in that: include: S1. Use a spaceborne multi-wavelength lidar system to emit 532nm and 1572nm pulsed lasers to acquire backscattered echo signals from interactions with atmospheric molecules, aerosols, and cloud particles. S2. Based on the satellite platform coordinates, the longitude and latitude of the laser pointing point, and the acquisition delay time of the measurement point, combined with the elevation data of the laser radar pointing point, the altitude corresponding to each measurement point in the echo signal is calculated; S3. The echo signal is preprocessed by sliding average and wavelet denoising to obtain a preprocessed echo signal; S4. Perform distance correction and system constant correction on the preprocessed echo signal to obtain the attenuated backscattering signals of the 532nm vertical channel, parallel channel, hyperspectral molecular channel, and 1572nm channel; S5. From the attenuated backscatter signal of the 1572 nm channel, use threshold screening and differential processing to identify the upper and lower boundaries of the cloud layer. Correlate the cloud boundary locations with the calculated altitude corresponding to the measurement point to obtain cloud height information. S6. Based on the temperature and pressure profiles provided by the numerical weather prediction model reanalysis meteorological dataset, the S6 atmospheric model, and the absorption spectrum of the iodine molecular filter, a molecular transmittance lookup table was established at 532 nm to calculate aerosol and cloud particle transmittance and atmospheric molecular parameters. S7. Invert the cloud backscattering coefficient using the attenuated backscattering signal from the 532 nm vertical channel, parallel channel, and hyperspectral molecular channel, the molecular transmittance lookup table, the aerosol and cloud particle transmittance, and atmospheric molecular parameters. S8. Calculate the cloud optical thickness using the attenuated backscattering signal from the 532 nm vertical channel, parallel channel, and hyperspectral molecular channel, aerosol and cloud particle transmittance, and atmospheric molecular parameters. Obtain the extinction coefficient by differentiating the optical thickness. S9. Calculate the integrated backscattering coefficient and integrated depolarization ratio of the cloud layer using the cloud boundary information, the attenuated backscattering signals of the 532nm vertical and parallel channels, and the cloud backscattering coefficient; S10. Identify the cloud particle phase based on the correlation between the integrated backscattering coefficient and the integrated depolarization ratio of the cloud layer.

2. The inversion method for detecting cloud particle physical properties using a spaceborne multi-wavelength lidar according to claim 1 is characterized in that: The step S2 of obtaining the altitude corresponding to each sampling point in the echo signal includes: S21. Using the acquisition delay time information of each measurement point, calculate the distance r from the satellite to the sampling point; S22. Using the satellite origin information in the WGS-84 coordinate system and the latitude and longitude of the satellite-borne lidar pointing point, obtain the distance l from each pointing point to the satellite. S23. Use the DEM data provided by the satellite to calculate the elevation d of the laser radar pointing point, and finally calculate the altitude h corresponding to each measurement point. The height h is calculated as follows: h=l+dr#(1).

3. The inversion method for detecting cloud particle physical properties using a spaceborne multi-wavelength lidar according to claim 1 is characterized in that: In step S4, the system constant calibration uses independent system constants C1, C2, C3, and C4 for different channels, corresponding to 532nm vertical channel system constant C1, 532nm parallel channel system constant C2, 532nm hyperspectral molecular channel system constant C3, 1572nm channel system constant C4.

4. The inversion method for detecting cloud particle physical properties using a spaceborne multi-wavelength lidar according to claim 1 is characterized in that: The cloud boundary identification method in step S5 includes: 1) Perform area integration on the attenuated backscattered signal of the 1572 nm channel and set the threshold to 2.0 × 10 -5 , the signal greater than the threshold is determined as a cloud signal; 2) Perform differential processing on the filtered signal and determine the cloud top height and cloud base height by finding the zero-crossing point position of the differential signal.

5. The inversion method for detecting cloud particle physical properties using a spaceborne multi-wavelength lidar according to claim 1 is characterized in that: The method for constructing the molecular transmittance lookup table in step S6 includes: 1) Calculate the molecular Rayleigh scattering spectrum at 532 nm based on the temperature and pressure profile data provided by the numerical weather forecast model reanalysis meteorological dataset and the S6 atmospheric model; 2) Perform convolution calculation based on the iodine molecule filter absorption spectrum measured in the laboratory.

6. The inversion method for detecting cloud particle physical properties using a spaceborne multi-wavelength lidar according to claim 1 is characterized in that: The backscattering coefficient β(λ 532 ,r), the formula is as follows: Where, β mol (r) is the backscattering coefficient of atmospheric molecules at the detection distance r, T mol (r) is the transmittance of the molecular Rayleigh scattering signal at the detection distance r, T(λ 532 ,r) is the Mie scattering signal transmittance of clouds and aerosol particles at the detection distance r; δ mol is the depolarization ratio of atmospheric molecules, δ(λ 532 ,r) is the total depolarization ratio at the detection distance r in the 532 nm band, which represents the ratio of the attenuated backscattered signals of the perpendicular channel and the parallel channel. K(r) is the ratio of the attenuated backscattered signals of the parallel channel and the hyperspectral channel at the detection distance r in the 532 nm band.

7. The inversion method for detecting cloud particle physical properties using a spaceborne multi-wavelength lidar according to claim 1 is characterized in that: The calculation formula of the optical thickness τ(r) in step S8 is as follows: Where, β mol (r) is the backscattering coefficient of atmospheric molecules at the detection distance r, T mol (r) is the transmittance of the molecular Rayleigh scattering signal at the detection distance r, T(λ 532 ,r) is the Mie scattering signal transmittance of clouds and aerosol particles at the detection distance r; δ mol is the depolarization ratio of atmospheric molecules, K(r) is the ratio of the attenuated backscattered signals of the parallel channel and the hyperspectral channel at the detection distance r in the 532 nm band, is the attenuated backscattering coefficient of the hyperspectral channel.

8. The inversion method for detecting cloud particle physical properties using a spaceborne multi-wavelength lidar according to claim 1 is characterized in that: The step S9 calculates the integrated backscattering coefficient and the integrated depolarization ratio of the cloud layer, including: S91. Calculation of the depolarization ratio δ of cloud particles cloud (r), the formula is as follows: S92. Calculate the cloud layer's integrated backscattering coefficient Y' and integrated depolarization ratio δ' based on the extracted cloud layer boundary information, the cloud backscattering coefficient, and the cloud particle depolarization ratio. The calculation formula is as follows:

9. The inversion method for detecting cloud particle physical properties using a spaceborne multi-wavelength lidar according to claim 1, characterized in that: The cloud particle phase identification method in step S10 is specifically: When Υ′ decreases with the increase of δ′, it is determined to be a water cloud particle; When Υ′ increases with the increase of δ′, it is determined to be ice cloud particles.

Images