A portable cloud parameter remote sensing system and method with multi-channel active and passive coordination
By employing a multi-channel active-passive collaborative cloud parameter remote sensing method, combined with near-infrared, thermal infrared, and microwave detection technologies, the problems of multiple solutions in cloud parameter remote sensing and system portability have been solved, enabling high-precision and portable observation of cloud parameters.
Patent Information
- Application Number
- CN202611626913.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-10-21
- Publication Date
- 2026-11-17
AI Technical Summary
Existing cloud parameter remote sensing technologies suffer from multiple solutions and uncertainties in inversion using a single detection method. Furthermore, existing integrated detection systems are bulky and difficult to deploy in a portable manner, failing to meet the needs of mobile field observation and refined detection of local areas.
A multi-channel, active-passive coordinated portable cloud parameter remote sensing method is adopted, combining near-infrared, thermal infrared, and microwave detection technologies. Near-infrared laser detection is used to determine the cloud base height, thermal infrared channel measures the cloud base temperature, microwave radiation detects the cloud liquid water content, and multi-objective optimization algorithm is used to invert cloud optical thickness and effective particle radius, achieving high-precision coordinated inversion of multiple parameters.
It effectively constrains the multiple solutions of inversion by a single detection method, and realizes high-precision joint inversion of cloud base height, cloud base temperature, cloud liquid water content, cloud optical thickness, cloud particle effective radius and cloud particle spectrum distribution. The system is portable and cost-effective, and is suitable for mobile field observation and emergency support scenarios.
Smart Images

Figure CN123113041A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of remote sensing technology, and in particular to a portable cloud parameter remote sensing system and method with multi-channel active and passive coordination. Background Technology
[0002] As one of the most important regulators in the Earth-atmosphere system, clouds profoundly influence global energy balance and water cycle processes through the absorption and scattering of radiation. Accurate detection of cloud parameters such as cloud base height, cloud base temperature, cloud liquid water content, cloud optical thickness, effective radius of cloud particles, and cloud particle spectral distribution is of significant scientific importance and practical value for weather forecasting, climate change research, and weather modification operations. However, current cloud observation still relies primarily on manual observation, which suffers from significant drawbacks such as high subjectivity, high labor intensity, and difficulties in nighttime observation, making it difficult to meet the requirements of automated, all-weather cloud observation.
[0003] Currently, cloud remote sensing systems based on spaceborne or ground-based platforms mainly employ single-band passive remote sensing technologies, such as infrared radiometers or microwave radiometers, to invert cloud microphysical parameters by detecting the thermal radiation information of the cloud. However, single-band passive detection has inherent limitations: the infrared band is sensitive to thin clouds but cannot penetrate thick clouds, while the microwave band can penetrate cloud layers but has low spatial resolution. Furthermore, neither of these methods can independently obtain information about the vertical structure of the cloud layer, leading to multiple solutions and uncertainties in cloud parameter inversion.
[0004] Existing lidar systems employing active detection methods detect clouds and aerosols via laser echo signals. However, these systems primarily rely on active detection, lacking the fusion of passive radiation information and failing to be optimized for portable applications. Their ability to jointly retrieve cloud parameters under complex weather conditions is also limited. Furthermore, while existing integrated active and passive detection systems can acquire vertical distribution profiles of various atmospheric parameters, as large-scale ground-based integrated detection systems, they are bulky, structurally complex, and costly, primarily deployed at fixed stations. This makes them unsuitable for portable applications in scenarios such as mobile field observation, detailed localized detection, and emergency support.
[0005] How to achieve the organic synergy between active laser detection and passive multi-band radiation detection, and effectively solve the inversion ambiguity and uncertainty of single detection methods through joint inversion of multi-wavelength detection information, while taking into account the portability of the system, is a technical problem that urgently needs to be solved in the field of cloud parameter remote sensing. Summary of the Invention
[0006] To address the issues of multiple solutions and uncertainties in cloud parameter inversion using single detection methods, and the large size and difficulty in portable deployment of existing integrated detection systems, this invention proposes a portable cloud parameter remote sensing system and method with multi-channel active and passive coordination.
[0007] The technical solution adopted by this invention to solve its technical problem is: a portable cloud parameter remote sensing method with multi-channel active and passive coordination, comprising the following steps: Step 1: Transmit a near-infrared detection pulse into the upper cloud layer through a near-infrared transmitting antenna, and receive the echo formed by the reflection of the detection pulse through the cloud layer through a near-infrared receiving antenna. Determine the cloud base height based on the propagation time between the transmission time of the detection pulse and the reception time of the echo. Step 2: Receive atmospheric down-current radiation through the first and second thermal infrared channels of the thermal infrared receiving antenna to obtain the first brightness temperature corresponding to the first thermal infrared channel and the second brightness temperature corresponding to the second thermal infrared channel. Determine the cloud base temperature based on the first and second brightness temperatures using the split window algorithm. Step 3: Receive atmospheric microwave radiation through multiple K-band channels of the microwave receiving antenna to obtain the microwave brightness temperature of each K-band channel, and invert the cloud liquid water content based on the microwave brightness temperature. Step 4: Based on the cloud base height, cloud base temperature, and microwave brightness temperature, the cloud optical thickness and cloud particle effective radius are retrieved using a radiative transfer model and a multi-objective optimization algorithm. The multi-objective optimization algorithm uses the fitting error between the observed microwave brightness temperature and the simulated microwave brightness temperature, and the physical consistency error between each K-band channel as optimization objectives, and determines the cloud optical thickness and the cloud particle effective radius based on the Pareto dominance principle. Step 5: Based on the cloud liquid water content, cloud optical thickness, and cloud particle effective radius, solve the spectral parameters of the cloud particle spectrum distribution using the particle spectrum model, and integrate the cloud particle spectrum distribution over the entire particle size range to obtain the cloud droplet number concentration.
[0008] The above-mentioned multi-channel active-passive coordinated portable cloud parameter remote sensing method, specifically includes step 2 as follows: According to the formula Calculate cloud base temperature, where, The coefficients were calculated using the MODTRAN radiative transfer model. Different combinations of parameters from different atmospheric models and cloud base temperatures were input, and the optimal values were obtained through regression analysis. First brightness temperature The second brightness temperature; The center wavelength of the first thermal infrared channel is 11 μm, and the center wavelength of the second thermal infrared channel is 12 μm.
[0009] The above-mentioned multi-channel active-passive coordinated portable cloud parameter remote sensing method, specifically step 3, involves: using the Gauss-Newton iterative method to solve for the cloud water attenuation component in the total cloud and atmospheric attenuation based on the microwave brightness temperature and atmospheric microwave remote sensing equations for each K-band channel; wherein the total cloud and atmospheric attenuation is the sum of oxygen attenuation, water vapor attenuation, and cloud water attenuation, the cloud water attenuation component is determined based on the cloud water attenuation coefficient and the cloud water path, and the cloud water attenuation coefficient is determined by the cloud base temperature.
[0010] The aforementioned portable cloud parameter remote sensing method with multi-channel active and passive coordination specifically uses a multi-objective optimization algorithm, namely, a multi-objective entropy-weighted adaptive beetle whisker-quantum hybrid optimization algorithm.
[0011] The above-mentioned multi-channel active-passive coordinated portable cloud parameter remote sensing method, specifically the inversion process of the multi-objective entropy weight adaptive beetle whisker-quantum hybrid optimization algorithm, is as follows: Step 4.1: Construct a dual objective function for fitting error and physical consistency error, initialize the quantum longhorn beetle individual, and generate left and right probe solutions; Step 4.2: Calculate the information entropy for the target values of the fitting error and physical consistency error of the current iteration, calculate the dynamic weights of the fitting error and physical consistency error based on the information entropy, and obtain the comprehensive fitness based on the dynamic weights. Step 4.3: Compare the adaptability of the left and right whiskers to determine the direction of movement of the longhorn beetle whiskers; Step 4.4: Update the candidate solution using a quantum rotation gate. The quantum rotation angle corresponding to the j-th parameter to be inverted is: in, The fundamental angle represents the basic angular amplitude of quantum rotation and controls the overall rotation intensity; E j Let the information entropy of the j-th target be , The average entropy of all targets, For the direction symbol item, and These are the objective function values at the current solution and the historical best solution, respectively; the qubits corresponding to each parameter to be inverted are rotated and updated according to the quantum rotation angle, and the updated qubits are observed to obtain a new generation of candidate solutions; Step 4.5: Merge the contemporary solution with the historical Pareto elite library, perform fast non-dominated sorting and crowding calculation, and retain the optimal solution set with uniform front distribution; when the crowding is less than a preset lower limit, increase the quantum perturbation step size; when the crowding is greater than a preset upper limit, perform fine local search; terminate the iteration when the maximum number of iterations is reached or the change in the objective function is less than the convergence threshold, and select the solution with the smallest comprehensive error from the Pareto front as the final inversion result of the cloud optical thickness and the cloud particle effective radius.
[0012] The objective function for the fitting error in the aforementioned multi-channel active-passive coordinated portable cloud parameter remote sensing method is: in, For the frequency of channel i, For cloud optical thickness, The effective radius of the cloud particle. Number of channels To observe brightness temperature, To calculate brightness temperature; The objective function for the physical consistency error is: .
[0013] In the above-mentioned multi-channel active-passive collaborative portable cloud parameter remote sensing method, in step 4.4, boundary constraints are applied to the new generation of candidate solutions. The value range of the cloud optical thickness is 2 to 50, and the value range of the cloud particle effective radius is 4 μm to 30 μm. When the updated candidate solution crosses the boundary, the out-of-bounds component is reflected back into the feasible region using a mirror bounce method.
[0014] The above-mentioned multi-channel active-passive coordinated portable cloud parameter remote sensing method, in step 5, cloud droplet number concentration... for: in, For the intercept parameter, For gamma function, For spectral parameters, This is the slope parameter.
[0015] A multi-channel active-passive coordinated portable cloud parameter remote sensing system is provided for implementing the multi-channel active-passive coordinated portable cloud parameter remote sensing method described above. The system includes a near-infrared detection module, a thermal infrared detection module, a microwave radiation detection module, an optical path separation system, an inversion processing module, and a power management module integrated within a protective housing. The near-infrared detection module includes a near-infrared transmitting antenna and a near-infrared receiving antenna, which are used to transmit near-infrared detection pulses to the upper cloud layer and receive the echoes formed by the reflection of the cloud layer, and determine the cloud base height based on the propagation time. The thermal infrared detection module includes a thermal infrared receiving antenna with a first thermal infrared channel and a second thermal infrared channel, used to receive downward atmospheric radiation and output the brightness temperature corresponding to each of the two thermal infrared channels. The microwave radiation detection module includes a microwave receiving antenna with multiple K-band channels, used to receive atmospheric microwave radiation and output the microwave brightness temperature of each K-band channel. The optical path separation system is used to distinguish, guide and isolate the detection optical paths of the near-infrared band and the thermal infrared band according to wavelength and propagation direction, so that signals of different bands are transmitted to the corresponding receiving antennas without interference. The inversion processing module is used to determine the cloud base temperature based on the brightness temperature of the two thermal infrared channels using a split window algorithm, invert the cloud liquid water content based on the microwave brightness temperature, and invert the cloud optical thickness and cloud particle effective radius based on the cloud base height, cloud base temperature, microwave brightness temperature, radiative transfer model and multi-objective optimization algorithm. Then, based on the particle spectrum model, it solves the cloud particle spectrum distribution parameters and cloud droplet number concentration. The power management unit is used to manage the power supply to the system; The optical path separation system adopts a coaxial optical structure: the transmitting optical path and the receiving optical path share the same main lens, the receiving reflector is placed at an angle of 45° and is provided with a transmission area and a reflection area, the transmitted light is emitted through the transmission area, and the echo signal enters the receiving optical path through the reflection area. The near-infrared detection module, the thermal infrared detection module, and the microwave radiation detection module are integrated into a common aperture optical structure.
[0016] The aforementioned multi-channel active-passive coordinated portable cloud parameter remote sensing system further includes a radiometric calibration module and a mode switching module. The radiometric calibration module includes a first blackbody radiation source and a second blackbody radiation source respectively set on both sides of the optical system, with a temperature difference of not less than 20°C between the two blackbodies. Before each observation, the thermal infrared detection module sequentially collects the infrared radiation values of the two blackbodies and obtains the real-time calibration coefficients by solving the calibration model. The mode switching module is used to automatically switch the working mode according to the ambient brightness and the probability of cloud presence: when it is determined to be a cloudless state, the near-infrared detection module is turned off, and only the thermal infrared detection module and the microwave radiation detection module are kept working; when it is determined that there are clouds based on passive detection data, the near-infrared detection module is activated to perform active detection.
[0017] The beneficial effects of this invention are that it organically coordinates active near-infrared laser detection with passive thermal infrared and K-band microwave radiation detection. Active detection provides information on cloud geometry, while passive detection provides information on thermal characteristics and microphysical radiation. The two types of information constrain and verify each other in the inversion process, effectively constraining the multiple solutions of inversion by a single detection method. It overcomes the inherent defects of traditional single-band passive remote sensing, which is difficult to balance the sensitivity of thin clouds with the penetration ability of thick clouds and cannot obtain vertical structure information. It achieves coordinated and high-precision joint inversion of multiple parameters such as cloud base height, cloud base temperature, cloud liquid water content, cloud optical thickness, effective radius of cloud particles, cloud particle spectrum distribution, and cloud droplet number concentration.
[0018] The multi-objective entropy-weighted adaptive hybrid optimization algorithm of this invention automatically coordinates and balances the two conflicting objectives of fitting error and physical consistency through dynamic weights of information entropy. It adaptively adjusts the intensity of global and local search through a quantum rotation gate driven by entropy difference, and maintains the diversity of solution set by combining Pareto non-dominated sorting. It effectively solves the ill-conditioned inversion problem of strong coupling between cloud optical thickness and cloud particle effective radius, and improves the convergence, stability and physical rationality of the inversion results.
[0019] This invention employs a 940nm eye-safe near-infrared band and a common-aperture optical structure. Through an optical path separation system, it achieves spatial isolation and efficient transmission of multi-band signals. Combined with modular lightweight design and low-power intelligent working mode scheduling, it significantly reduces equipment size, cost, and power consumption while ensuring multi-band collaborative detection performance. This enables the system to safely and flexibly adapt to the refined detection needs of scenarios with strict requirements for personnel safety, such as field mobile observation, urban environmental monitoring, and emergency support. It realizes the transformation of high-precision cloud parameter detection from fixed-site observation to mobile, portable, and rapid on-site detection. Attached Figure Description
[0020] Figure 1 This is a schematic diagram of the process of this invention. Detailed Implementation
[0021] To enable those skilled in the art to better understand the technical solution of the present invention, the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.
[0022] As shown in the figure, this embodiment discloses a portable cloud parameter remote sensing method with multi-channel active and passive coordination, including the following steps: Step 1: Transmit a near-infrared detection pulse into the upper cloud layer using a near-infrared transmitting antenna, and receive the echo formed by the reflection of the detection pulse into the cloud layer using a near-infrared receiving antenna. Determine the cloud base height based on the propagation time between the transmission time of the detection pulse and the reception time of the echo.
[0023] A near-infrared transmitting antenna emits 940nm wavelength near-infrared light upwards (towards the cloud base), and a near-infrared receiving antenna receives the 940nm echo. Assuming the propagation time from transmission to reception is t, the cloud base height is... , where c is the speed of light constant.
[0024] Near-infrared lasers in the 940nm band have advantages such as relative safety for the human eye, strong atmospheric penetration, and stable operation both day and night. Meanwhile, avalanche photodiodes have high quantum efficiency in this band, enabling high-sensitivity detection.
[0025] Step 2: Receive atmospheric downdraft radiation through the first and second thermal infrared channels of the thermal infrared receiving antenna, respectively, to obtain the first brightness temperature corresponding to the first thermal infrared channel and the second brightness temperature corresponding to the second thermal infrared channel. Determine the cloud base temperature based on the first and second brightness temperatures using the split window algorithm.
[0026] Thermal infrared split-window technology is based on two adjacent "atmospheric windows" (usually 10.5~11.5μm and 11.5~12.5μm) in the 8~14μm band. In this band, the atmosphere absorbs infrared radiation weakly, allowing infrared radiation from the ground to pass through the atmosphere and be received by the sensor, while also allowing infrared radiation from the clouds to be transmitted down from the cloud base to the ground detector.
[0027] The core idea of the split-window method is to utilize the difference in water vapor absorption between two adjacent channels to eliminate atmospheric influences. The 11μm and 12μm channels absorb atmospheric water vapor to different degrees—the latter being more sensitive to water vapor. By measuring the brightness temperature difference between the two channels, the atmospheric water vapor content can be quantitatively estimated, and the atmospheric radiation attenuation effect can be corrected accordingly. This principle is well-established in the field of land surface temperature retrieval, and the national standard "Land Surface Temperature Thermal Infrared Remote Sensing Retrieval Part 2: Split-Window Method" (GB / T43997.2-2024) provides systematic technical specifications for it.
[0028] Establish an empirical split-window algorithm to calculate the temperature formula. in, The coefficients were calculated using the MODTRAN radiative transfer model. Different combinations of parameters from different atmospheric models (such as mid-latitude summer, mid-latitude winter, and tropical) and different cloud base temperatures were input, and the optimal values were obtained through regression analysis. It is a dual-channel brightness temperature for thermal infrared.
[0029] The thermal infrared receiving antenna receives atmospheric downdraft radiation at wavelengths of 11 μm and 12 μm. and , Substituting into the above equation, we get .
[0030] Step 3: Receive atmospheric microwave radiation through multiple K-band channels of the microwave receiving antenna to obtain the microwave brightness temperature of each K-band channel, and then invert the cloud liquid water content based on the microwave brightness temperature.
[0031] In addition to the contributions of oxygen and water vapor, the microwave radiation of the cloud-covered atmosphere is also absorbed by clouds. Therefore, the total atmospheric attenuation under cloud conditions can be expressed as: In the formula, the subscript Indicates the first Atmosphere This represents the total atmospheric attenuation profile. Indicates the first Layer oxygen decay profile, Indicates the first Layer water vapor attenuation profile, This indicates the decay of liquid water in the cloud; Assuming the atmospheric type at the equipment installation site is a mid-latitude summer atmosphere, then Known, and then , It is a constant; Will Convert to height Based on the brightness temperature of the K channel measured by the microwave antenna Atmospheric microwave remote sensing equations Using the Gauss-Newton iterative method, we can approximate the result to obtain... ; in, Tc represents the cloud liquid water attenuation coefficient, with units of ((dB / km) / (g / m3)), and Tc represents the cloud liquid water temperature (measured by a thermal infrared receiving antenna). The specific calculation process can be found in patent CN118225807B, where M represents the liquid water content of the cloud; then .
[0032] Step 4: Based on the cloud base height, cloud base temperature, and microwave brightness temperature, the cloud optical thickness and effective radius of cloud particles are inverted using a radiative transfer model and a multi-objective optimization algorithm. The multi-objective optimization algorithm uses the fitting error between the observed microwave brightness temperature and the simulated microwave brightness temperature, and the physical consistency error between each K-band channel as optimization objectives, and determines the cloud optical thickness and the effective radius of cloud particles based on the Pareto dominance principle.
[0033] In this embodiment, the multi-objective optimization algorithm is specifically a multi-objective entropy-weighted adaptive beetle whisker-quantum hybrid optimization algorithm. The specific inversion process of the multi-objective entropy-weighted adaptive beetle whisker-quantum hybrid optimization algorithm is as follows: Step 4.1: Construct a dual objective function for fitting error and physical consistency error, initialize the quantum longhorn beetle individual, and generate left and right beard probe solutions.
[0034] The objective function for the fitting error is: in, For the frequency of channel i, For cloud optical thickness, The effective radius of the cloud particle. Number of channels To observe brightness temperature, To calculate brightness temperature; The objective function for the physical consistency error is: .
[0035] Only a single quantum longhorn beetle is needed, encoding two variables: Each variable corresponds to a set of qubits: Initialize to a uniform superposition state: The initial solution was obtained from the observation: random unit direction vector Set the initial step size : right Substitute each value into the radiative transfer model and calculate the objective function value. In this embodiment, the radiative transfer model uses PAMTRA2.
[0036] Step 4.2: Calculate the information entropy for the target values of the fitting error and physical consistency error of the current iteration, calculate the dynamic weights of the fitting error and physical consistency error based on the information entropy, and obtain the comprehensive fitness based on the dynamic weights.
[0037] Calculate the entropy for all target values in the current iteration: ; Explanation of the formula for calculating information entropy: In bi-objective optimization (inverting cloud parameters), for the i-th objective function, the information entropy is calculated as: ; Where the probability components are: Let be the function value of the i-th target at the j-th solution; i=1 represents the corresponding target 1 - the error in fitting the observed and simulated reflectance; i=2 represents the corresponding target 2 - the constraint of spectral smoothing / physical consistency; j represents the solution currently involved in the calculation: the current position of the longhorn beetle xnow, the left whisker xleft, and the right whisker xright, a total of 3 points, so j=1,2,3.
[0038] Dynamic weights: Overall adaptability: .
[0039] Step 4.3: Compare the adaptability of the left and right whiskers to determine the direction of movement of the longhorn beetle whiskers.
[0040] Compare left and right sides for adaptability: if → Move left; conversely → Move right; close → Initiate quantum rotation perturbation Step 4.4: Update the candidate solution using a quantum rotation gate. The quantum rotation angle corresponding to the j-th parameter to be inverted is: in, The base angle, with a value of 0.01π to 0.1π (approximately 1.8° to 18°), represents the fundamental angular amplitude of quantum rotation and controls the overall rotation intensity. The average entropy of all targets is used as a balancing benchmark to determine whether the current target is "above average disorder" or "below average disorder". For the direction symbol item, and E represents the objective function values at the current solution and the historical best solution, respectively. j Let E1 be the entropy of the j-th objective, E2 be the entropy of the fitting error, and E3 be the entropy of the physical consistency error. The larger the entropy, the more chaotic and unstable the objective function is.
[0041] The qubits corresponding to each parameter to be inverted are rotated and updated according to the quantum rotation angle, and the updated qubits are observed to obtain a new generation of candidate solutions. Boundary constraints are imposed on the new generation of candidate solutions, wherein the cloud optical thickness ranges from 2 to 50, and the effective radius of the cloud particles ranges from 4 μm to 30 μm; when the updated candidate solution crosses the boundary, the out-of-bounds component is reflected back into the feasible region using a mirror bounce method.
[0042] Step 4.5: Merge the contemporary solution with the historical Pareto elite library, perform fast non-dominated sorting and crowding calculation, and retain the optimal solution set with uniform front distribution; when the crowding is less than a preset lower limit, increase the quantum perturbation step size; when the crowding is greater than a preset upper limit, perform fine local search; terminate the iteration when the maximum number of iterations is reached or the change in the objective function is less than the convergence threshold, and select the solution with the smallest comprehensive error from the Pareto front as the final inversion result of the cloud optical thickness and the cloud particle effective radius.
[0043] Step 5: Based on the cloud liquid water content, cloud optical thickness, and cloud particle effective radius, solve the spectral parameters of the cloud particle spectrum distribution using the particle spectrum model, and integrate the cloud particle spectrum distribution over the entire particle size range to obtain the cloud droplet number concentration.
[0044] The physical meaning of cloud particle spectral distribution: the number of cloud droplets per unit volume and unit diameter interval. Based on the cloud microphysical integral relationship and the Gamma log-normal spectrum assumption, the cloud particle spectral distribution (N0, μ, λ) can be solved by the three parameters LWC, COT, and re in a closed loop. The distribution of cloud particle spectra is commonly described by the Gamma distribution: ) in, :diameter Cloud particle spectrum distribution ( ); : Intercept parameter ( ); : Spectral parameters (dimensionless); : Slope parameter ( ); The factorial function (gamma function) is a generalization of the factorial to real and complex numbers, derived from the integral. Define, satisfy It has wide applications in probability, physics, and number theory.
[0045] Integral relationships between the three main observation parameters and the spectral distribution: (1) Liquid Water Content (LWC) Mass of liquid water per unit volume: Integrating Gamma yields: ① (2) Cloud Optical Thickness (COT) Vertical extinction integral (τ): Visible / Near Infrared: Q ext ≈2 (Geometric optics + Rayleigh approximation, small particles) Simplified (vertical thickness Δz normalized): ② (3) Effective radius r of cloud droplets e (EffectiveRadius) Area-weighted average radius (core observation of remote sensing): a radiometric equivalent scale representing the average particle size of cloud droplets. It is defined as the ratio of the third moment to the second moment of the particle size distribution. Substitute into the Gamma integral: ③ By combining equations ①②③, we can utilize... , simplify: Simultaneous solution Thus, to obtain The physical meaning of cloud droplet number concentration: the total number of cloud droplets of all diameters per unit volume. Integrating the particle spectral distribution over the entire particle size gives the cloud droplet number concentration. Gamma spectrum represented by diameter: The number concentration is then: .
[0046] When the above steps are executed in synergy, a significant synergistic enhancement effect can be produced. The cloud base height locked by active laser detection provides a definite lower boundary for the layered integration of the microwave remote sensing equation, eliminating the integration path error caused by the uncertainty of the cloud base height; the cloud base temperature measured by the thermal infrared split window corrects the cloud liquid water attenuation coefficient in real time, making the cloud liquid water content retrieved by microwave adaptive to changes in cloud temperature; the multi-channel microwave brightness temperature simultaneously constrains the cloud optical thickness and the effective radius of cloud particles under the entropy weight adaptive multi-objective framework, overcoming the ill-conditioned nature of strong coupling between the two parameters; the three inversion quantities are then solved by the spectral model closure, and the inversion process of each parameter is mutually constrained and mutually verified. Any abnormal deviation of any parameter will be exposed in the simultaneous solution, thus effectively avoiding the systematic deviation caused by single-point measurement and constructing a complete cloud parameter system covering geometric structure, thermal characteristics and microphysical properties.
[0047] Based on the above method, this embodiment also discloses a multi-channel active-passive coordinated portable cloud parameter remote sensing system, including a near-infrared detection module, a thermal infrared detection module, a microwave radiation detection module, an optical path separation system, an inversion processing module, a power management module, an optical window, a storage unit, an input / output interface, and an auxiliary environmental sensor group integrated within a protective housing. The optical window is located on the detection surface of the protective housing, protecting the sensitive components inside the system from environmental influences such as dehumidification and waterproofing, while allowing target light and electromagnetic radiation to pass through efficiently and with high fidelity. In low-temperature environments, the optical window 201 works in conjunction with an automatic heater to prevent icing and condensation from affecting transmittance.
[0048] The near-infrared detection module includes a near-infrared transmitting antenna and a near-infrared receiving antenna. It transmits near-infrared detection pulses into the upper clouds and receives the echoes reflected by the clouds, determining the cloud base height based on the propagation time. The near-infrared transmitting antenna uses a pulsed semiconductor laser with a center wavelength of 940nm±3nm, an adjustable pulse width of 1~10ns, a single pulse energy of not less than 10μJ, and a repetition frequency of 1~10kHz. After beam expansion and collimation, the laser divergence angle is controlled within 0.5mrad to ensure energy density for long-distance detection. The 940nm band offers advantages such as relative safety for the human eye, strong atmospheric penetration, and stable operation day and night. The near-infrared receiving antenna uses an avalanche photodiode (APD) in conjunction with a transimpedance amplifier (TAA) to detect weak echo signals. The APD has high quantum efficiency, high gain, low noise, and fast response speed at a wavelength of 940nm. The echo enters the detector after being filtered by a narrowband filter with a center wavelength of 940nm and a bandwidth of no more than 3nm, effectively suppressing solar background light interference and making it suitable for continuous day and night operation.
[0049] The thermal infrared detection module includes a thermal infrared receiving antenna with a first thermal infrared channel and a second thermal infrared channel, used to receive downward atmospheric radiation and output the brightness temperature corresponding to each of the two thermal infrared channels. The center wavelength of the first thermal infrared channel is approximately 11 μm, located in the strong channel of the atmospheric window; the center wavelength of the second thermal infrared channel is approximately 12 μm, a water vapor absorption sensitive channel. The two channels share the same optical system, and synchronous detection of the two channels is achieved through beam splitters or filter wheels, ensuring geometric consistency of observation between the two channels; the core detector can be a cooled type or a high-sensitivity uncooled infrared focal plane array.
[0050] The microwave radiation detection module includes a microwave receiving antenna with multiple K-band channels for passively receiving thermal radiation signals from cloud droplets in the K-band. The center frequencies of the multiple K-band channels are 22.235 GHz, 23.8 GHz, 26.24 GHz, 30.0 GHz, and 31.4 GHz. The K-band is located at the edge of the water vapor absorption band, making it sensitive to liquid water but moderately affected by water vapor interference. The wavelength (approximately 1 cm) and cloud droplet size (1–100 μm) of this band satisfy the Rayleigh scattering condition, exhibiting stable scattering and absorption patterns that are easy to model. It also possesses strong penetration and can operate in all weather conditions. The microwave radiation detection module further includes an antenna feed unit, an RF receiving unit, and a precision calibration unit: the antenna feed unit captures atmospheric K-band microwave radiation and transmits it to the receiver; the RF receiving unit employs a superheterodyne or direct detection architecture to amplify, frequency-convert, and detect the weak radiation signal before converting it into a digital signal; and the precision calibration unit eliminates system drift through absolute calibration to obtain the true atmospheric brightness temperature.
[0051] The optical path separation system is used to distinguish, guide, and isolate the detection optical paths in the near-infrared and thermal infrared bands according to wavelength and propagation direction, ensuring that signals from different bands are transmitted to their respective receiving antennas without interference. The optical path separation system employs a coaxial optical structure: the transmitting and receiving optical paths share a main lens; the receiving reflector is tilted at 45° and has a transmission area and a reflection area; the transmitted light exits through the transmission area, and the echo signal enters the receiving optical path through the reflection area; the near-infrared detection module, the thermal infrared detection module, and the microwave radiation detection module are integrated into a common aperture optical structure. This coaxial design effectively eliminates the blind zone problem of non-coaxial structures in close-range detection, improving measurement accuracy. In an optional embodiment, the receiving telescope can adopt a Cassegrain structure to further compress the axial dimension; this application does not limit this.
[0052] The auxiliary environmental sensor group integrates sensors for ground meteorological parameters such as temperature, humidity, air pressure, and wind force to obtain real-time ground environmental conditions and provide input for atmospheric transmission correction and atmospheric model selection. At the same time, it is equipped with a temperature and humidity control unit to monitor the internal temperature and humidity of the instrument, ensuring the stable operation of the equipment in complex field environments.
[0053] The inversion processing module is the core of the system's data processing. It adopts an embedded lightweight design and can be optionally equipped with a GPU or neural network acceleration unit. Integrating a collaborative fusion algorithm, it can perform real-time data acquisition, time-series synchronization, radiometric calibration, brightness temperature calculation, split-window inversion, cloud liquid water content inversion, multi-objective optimization inversion, and spectral parameter solving without an external computer. The inversion processing module performs time-series synchronization processing on active and passive detection signals, aligning the timestamps of data from each channel to ensure data correspondence and matching within the same detection area at the same time. The storage unit supports local data storage with a capacity of no less than 16GB. The input / output interface is a human-machine interface used to modify system configurations and output cloud remote sensing information and visualization results. The system also supports 4G / 5G, Ethernet, WiFi, Bluetooth, and other communication methods to adapt to cloud data synchronization and remote management requirements.
[0054] The power management unit manages the system's power supply, including power display, charging reminders, and overload voltage protection. The system is powered by a built-in high-energy-density lithium battery, providing at least 6 hours of battery life, and supports vehicle-mounted power and solar charging. The entire unit features a lightweight carbon fiber casing, weighing less than 5kg, and is approximately the size of a laptop. It can be handheld, carried in a backpack, mounted in a vehicle, or set up on a tripod. It offers convenient operation capabilities including one-button power-on, automatic calibration, automatic data acquisition, and automatic cloud recognition.
[0055] In this embodiment, the system further includes a radiometric calibration module and a mode switching module. The radiometric calibration module includes a first blackbody radiation source and a second blackbody radiation source respectively disposed on both sides of the optical system, with a temperature difference of not less than 20°C between the two blackbodies. Before each observation, the thermal infrared detection module sequentially collects the infrared radiation values of the two blackbodies and obtains the real-time calibration coefficients by solving the calibration model. This can effectively compensate for errors introduced by lens thermal radiation and optical component aging, reduce the frequency of manual calibration, and ensure the stability and reliability of long-term field observation data. In another implementation, an on-chip blackbody can be built in and automatic calibration can be achieved using the cold sky as a reference; this application does not limit this approach.
[0056] The mode switching module is used to automatically switch the working mode according to the ambient brightness and the probability of cloud presence: when it is determined to be a cloudless state, the near-infrared detection module is turned off, and only the thermal infrared detection module and the microwave radiation detection module are kept working; when it is determined that there are clouds based on passive detection data, the near-infrared detection module is activated to perform active detection.
[0057] In one possible implementation, the computational steps of the above-mentioned inversion algorithm may not be deployed on the device side. Instead, after the device side completes data acquisition and calibration, the original brightness temperature data is uploaded to a remote server via a communication unit to perform cloud liquid water content inversion, multi-objective optimization inversion, and spectral parameter solving. The results are then sent back to the device side for display, thereby further reducing the computing power and power consumption requirements of the device side. This application does not limit this aspect.
[0058] In another implementation, the laser of the near-infrared detection module can be replaced with a pulsed semiconductor laser operating in other near-infrared human eye-safe bands. The center wavelength of the channel of the thermal infrared detection module can be adjusted according to the actual atmospheric window and target characteristics. The number of channels of the microwave radiation detection module can be increased or decreased while meeting the information requirements of multi-channel joint inversion. The corresponding split window coefficient and cloud liquid water attenuation coefficient can be recalibrated according to the adjusted channels. This application embodiment does not limit this.
[0059] The above embodiments are merely exemplary embodiments of the present invention and are not intended to limit the present invention. Those skilled in the art can make various modifications or equivalent substitutions to the present invention within its scope and spirit, and such modifications or equivalent substitutions should also be considered to fall within the scope of protection of the present invention.
Claims
1. A multi-channel active-passive cooperative portable cloud parameter remote sensing method, characterized in that, Includes the following steps: Step 1: Transmit a near-infrared detection pulse into the upper cloud layer through a near-infrared transmitting antenna, and receive the echo formed by the reflection of the detection pulse through the cloud layer through a near-infrared receiving antenna. Determine the cloud base height based on the propagation time between the transmission time of the detection pulse and the reception time of the echo. Step 2: Receive atmospheric down-current radiation through the first and second thermal infrared channels of the thermal infrared receiving antenna to obtain the first brightness temperature corresponding to the first thermal infrared channel and the second brightness temperature corresponding to the second thermal infrared channel. Determine the cloud base temperature based on the first and second brightness temperatures using the split window algorithm. Step 3: Receive atmospheric microwave radiation through multiple K-band channels of the microwave receiving antenna to obtain the microwave brightness temperature of each K-band channel, and invert the cloud liquid water content based on the microwave brightness temperature. Step 4: Based on the cloud base height, cloud base temperature, and microwave brightness temperature, the cloud optical thickness and cloud particle effective radius are retrieved using a radiative transfer model and a multi-objective optimization algorithm. The multi-objective optimization algorithm uses the fitting error between the observed microwave brightness temperature and the simulated microwave brightness temperature, and the physical consistency error between each K-band channel as optimization objectives, and determines the cloud optical thickness and the cloud particle effective radius based on the Pareto dominance principle. Step 5: Based on the cloud liquid water content, cloud optical thickness, and cloud particle effective radius, solve the spectral parameters of the cloud particle spectrum distribution using the particle spectrum model, and integrate the cloud particle spectrum distribution over the entire particle size range to obtain the cloud droplet number concentration.
2. The portable cloud parameter remote sensing method with multi-channel active-passive coordination according to claim 1, characterized in that, Step 2 specifically includes: According to the formula Calculate cloud base temperature, where, The coefficients were calculated using the MODTRAN radiative transfer model. Different combinations of parameters from different atmospheric models and cloud base temperatures were input, and the optimal values were obtained through regression analysis. First brightness temperature The second brightness temperature; The center wavelength of the first thermal infrared channel is 11 μm, and the center wavelength of the second thermal infrared channel is 12 μm.
3. The portable cloud parameter remote sensing method with multi-channel active-passive coordination according to claim 1, characterized in that, Step 3 specifically involves: based on the microwave brightness temperature and atmospheric microwave remote sensing equations for each K-band channel, using the Gauss-Newton iterative method to solve for the cloud water attenuation component in the total cloud and atmospheric attenuation; wherein the total cloud and atmospheric attenuation is the sum of oxygen attenuation, water vapor attenuation, and cloud water attenuation, the cloud water attenuation component is determined based on the cloud water attenuation coefficient and the cloud water path, and the cloud water attenuation coefficient is determined by the cloud base temperature.
4. The portable cloud parameter remote sensing method with multi-channel active-passive coordination according to claim 1, characterized in that, The multi-objective optimization algorithm is specifically a multi-objective entropy-weighted adaptive beetle whisker-quantum hybrid optimization algorithm.
5. A portable cloud parameter remote sensing method with multi-channel active-passive coordination according to claim 4, characterized in that, The specific inversion process of the multi-objective entropy-weighted adaptive beetle whisker-quantum hybrid optimization algorithm is as follows: Step 4.1: Construct a dual objective function for fitting error and physical consistency error, initialize the quantum longhorn beetle individual, and generate left and right probe solutions; Step 4.2: Calculate the information entropy for the target values of the fitting error and physical consistency error of the current iteration, calculate the dynamic weights of the fitting error and physical consistency error based on the information entropy, and obtain the comprehensive fitness based on the dynamic weights. Step 4.3: Compare the adaptability of the left and right whiskers to determine the direction of movement of the longhorn beetle whiskers; Step 4.4: Update the candidate solution using a quantum rotation gate. The quantum rotation angle corresponding to the j-th parameter to be inverted is: in, The fundamental angle represents the basic angular amplitude of quantum rotation and controls the overall rotation intensity; E j Let the information entropy of the j-th target be , The average entropy of all targets, For the direction symbol item, and These are the objective function values at the current solution and the historical best solution, respectively; the qubits corresponding to each parameter to be inverted are rotated and updated according to the quantum rotation angle, and the updated qubits are observed to obtain a new generation of candidate solutions; Step 4.5: Merge the contemporary solution with the historical Pareto elite library, perform fast non-dominated sorting and crowding calculation, and retain the optimal solution set with uniform front distribution; when the crowding is less than a preset lower limit, increase the quantum perturbation step size; when the crowding is greater than a preset upper limit, perform fine local search; terminate the iteration when the maximum number of iterations is reached or the change in the objective function is less than the convergence threshold, and select the solution with the smallest comprehensive error from the Pareto front as the final inversion result of the cloud optical thickness and the cloud particle effective radius.
6. The portable cloud parameter remote sensing method with multi-channel active-passive coordination according to claim 5, characterized in that, The objective function for the fitting error is: in, For the frequency of channel i, For cloud optical thickness, The effective radius of the cloud particle. Number of channels To observe brightness temperature, To calculate brightness temperature; The objective function for the physical consistency error is: 。 7. A portable cloud parameter remote sensing method with multi-channel active-passive coordination according to claim 5, characterized in that, In step 4.4, boundary constraints are applied to the new generation of candidate solutions. The value of the cloud optical thickness ranges from 2 to 50, and the value of the effective radius of the cloud particles ranges from 4 μm to 30 μm. When the updated candidate solution crosses the boundary, the out-of-bounds component is reflected back into the feasible region using a mirror bounce method.
8. The portable cloud parameter remote sensing method with multi-channel active-passive coordination according to claim 1, characterized in that, Cloud droplet number concentration in step 5 for: in, For the intercept parameter, For gamma function, For spectral parameters, This is the slope parameter.
9. A portable cloud parameter remote sensing system with multi-channel active-passive coordination, characterized in that, A portable cloud parameter remote sensing method for implementing a multi-channel active-passive coordinated method as described in any one of claims 1-8 includes a near-infrared detection module, a thermal infrared detection module, a microwave radiation detection module, an optical path separation system, an inversion processing module, and a power management module integrated within a protective housing. The near-infrared detection module includes a near-infrared transmitting antenna and a near-infrared receiving antenna, used to transmit near-infrared detection pulses to the upper cloud layer and receive echoes formed by reflections from the cloud layer, and determine the cloud base height based on the propagation time. The thermal infrared detection module includes a thermal infrared receiving antenna with a first thermal infrared channel and a second thermal infrared channel, used to receive downward atmospheric radiation and output the brightness temperature corresponding to each of the two thermal infrared channels. The microwave radiation detection module includes a microwave receiving antenna with multiple K-band channels, used to receive atmospheric microwave radiation and output the microwave brightness temperature of each K-band channel. The optical path separation system is used to distinguish, guide and isolate the detection optical paths of the near-infrared band and the thermal infrared band according to wavelength and propagation direction, so that signals of different bands are transmitted to the corresponding receiving antennas without interference. The inversion processing module is used to determine the cloud base temperature based on the brightness temperature of the two thermal infrared channels using a split window algorithm, invert the cloud liquid water content based on the microwave brightness temperature, and invert the cloud optical thickness and cloud particle effective radius based on the cloud base height, cloud base temperature, microwave brightness temperature, radiative transfer model and multi-objective optimization algorithm. Then, based on the particle spectrum model, it solves the cloud particle spectrum distribution parameters and cloud droplet number concentration. The power management unit is used to manage the power supply to the system; The optical path separation system adopts a coaxial optical structure: the transmitting optical path and the receiving optical path share the same main lens, the receiving reflector is placed at an angle of 45° and is provided with a transmission area and a reflection area, the transmitted light is emitted through the transmission area, and the echo signal enters the receiving optical path through the reflection area. The near-infrared detection module, the thermal infrared detection module, and the microwave radiation detection module are integrated into a common aperture optical structure.
10. A multi-channel active-passive coordinated portable cloud parameter remote sensing system according to claim 9, characterized in that, It also includes a radiation calibration module and a mode switching module; the radiation calibration module includes a first blackbody radiation source and a second blackbody radiation source respectively set on both sides of the optical system, and the temperature difference between the two blackbodies is not less than 20°C; before each observation, the thermal infrared detection module sequentially collects the infrared radiation values of the two blackbodies, and obtains the real-time calibration coefficient by solving the calibration model. The mode switching module is used to automatically switch the working mode according to the ambient brightness and the probability of cloud presence: when it is determined to be a cloudless state, the near-infrared detection module is turned off, and only the thermal infrared detection module and the microwave radiation detection module are kept working; when it is determined that there are clouds based on passive detection data, the near-infrared detection module is activated to perform active detection.
