Airplane artificial precipitation enhancement method and system based on multi-source observation data

Abstract

The application relates to the technical field of artificial rain enhancement, in particular to an airplane artificial rain enhancement method and system based on multi-source observation data, which comprises the following steps: based on star-based monitoring data, air-based monitoring data and ground-based monitoring data in a target area, a cross-source verification data set in the target area is constructed, and based on the cross-source verification data set, when the star-based monitoring data, the air-based monitoring data and the ground-based monitoring data do not exist abnormally, a multi-source observation data set is constructed; based on the cross-source verification data set, a global observation covariance matrix corresponding to the multi-source observation data set is constructed, and based on the global observation covariance matrix and a pre-set target function, a multi-source inversion data set is constructed; based on the inversion data set and the multi-source observation data set, a rain enhancement operation potential index of the target area is constructed to assist artificial rain enhancement. The method effectively improves the solution precision of cloud state parameters and the credibility of rain enhancement potential evaluation, and provides reliable data support and decision basis for artificial rain enhancement operation.

Description

Technical Field

[0001] This application relates to the field of artificial rain enhancement technology, and in particular to an aircraft-based artificial rain enhancement method and system based on multi-source observation data. Background Technology

[0002] At present, the development and utilization of aerial cloud water resources is an important means to alleviate regional water shortages and improve the ecological environment. Artificial rain enhancement, as the mainstream technology for the efficient development of aerial cloud water resources, has been widely used in various places.

[0003] Currently, artificial rain enhancement operations mainly rely on single or limited types of observational data, such as satellites, radar, and ground-based meteorological stations, for cloud system analysis and operational assessment. While this can largely complete basic tasks like cloud water resource assessment and initial selection of operational areas, the application of existing observational data is fragmented, failing to systematically integrate multi-source observational information from satellite, space, and ground sources. Furthermore, most artificial rain enhancement technologies only provide simple summaries and intuitive analysis of observational data, lacking consistency verification mechanisms for multi-source observational data. This makes it impossible to effectively identify abnormal observational information or objectively assess data reliability, further contributing to uncertainty in cloud physical characteristic analysis results and hindering the scientific rigor of artificial rain enhancement operational assessments. Summary of the Invention

[0004] (a) Technical problems to be solved In view of the above-mentioned shortcomings and deficiencies of the prior art, this application provides an aircraft-based artificial rain enhancement method and system based on multi-source observation data. It solves the technical problems in the prior art, such as the lack of consistency verification of multi-source observation data from satellites, airspace, and ground, which leads to the inability to effectively identify abnormal data and the low degree of multi-source data fusion and utilization. It also solves the technical problems in the prior art, such as the difficulty in distinguishing the reliability of data in order to quantify and evaluate data, which ultimately leads to inaccurate judgment of the potential of artificial rain enhancement operations.

[0005] (II) Technical Solution To achieve the above objectives, the main technical solutions adopted in this application include: In a first aspect, embodiments of this application provide an aircraft-based artificial rain enhancement method based on multi-source observation data, comprising: Based on satellite-based, space-based, and ground-based monitoring data within the target area, a cross-source verification dataset is constructed within the target area to determine whether there are observational anomalies in the satellite-based, space-based, and ground-based monitoring data; the cross-source verification dataset is data that quantifies the physical deviations between different observation sources; Once it is confirmed that there are no anomalies in the satellite-based, space-based, and ground-based monitoring data, a multi-source observation dataset is constructed based on these data; the multi-source observation dataset includes multi-dimensional monitoring vectors. Based on the cross-source verification dataset within the target area, a global observation covariance matrix corresponding to the multi-source observation dataset is constructed. Based on the global observation covariance matrix and a pre-set objective function, a multi-source inversion dataset is constructed. The multi-source inversion dataset includes the inversion data corresponding to each dimension monitoring vector in the multi-source observation dataset. Based on the multi-source inversion dataset and objective function, a multi-source inversion error dataset is constructed. Based on the multi-source inversion error dataset and the multi-source observation dataset, a rain enhancement operation potential index for the target area is constructed to assist in artificial rain enhancement. The multi-source inversion error dataset includes the inversion error corresponding to each dimension monitoring vector in the multi-source observation dataset.

[0006] Optionally, the satellite-based monitoring data includes cloud top height, cloud top temperature, cloud optical thickness, cloud water content, and cloud droplet effective radius; the space-based monitoring data includes cloud temperature, horizontal wind speed, liquid water content, and supercooled water content; the ground-based monitoring data includes atmospheric precipitable water, radar echo intensity baseline, cloud movement speed, ground precipitation, and cloud phase; the cross-source verification dataset includes cross-source verification data of various verification types. Based on satellite-based, space-based, and ground-based monitoring data within the target area, a cross-source verification dataset for the target area is constructed, including: Based on the radar echo intensity baseline, the cloud base height is determined, and based on the cloud top height, cloud base height, and pre-set geometric correction parameters, geometric verification data is determined; wherein, the cloud base height is the altitude corresponding to the radar echo intensity baseline -15dBz; the geometric verification data is the difference between the cloud top height, cloud base height, and geometric correction parameters; Based on cloud water content, liquid water content, supercooled water content, atmospheric precipitable water, and cloud phase, water content verification data is constructed. Particle optical verification data are constructed based on cloud optical thickness, cloud droplet effective radius, cloud phase, liquid water content, and cloud water content. Based on horizontal wind speed and cloud movement speed, dynamic wind field verification data is determined; wherein, dynamic wind field verification data is the vector magnitude of the difference between cloud movement speed and horizontal wind speed; Based on surface precipitation and a pre-set precipitation correction curve, atmospheric temperature decreasing correction data is determined. Based on this atmospheric temperature decreasing correction data, cloud temperature, cloud top temperature, cloud top height, and cloud base height, and a pre-set formula (Formula 1), thermal verification data is determined. The precipitation correction curve is the corresponding curve for surface precipitation and atmospheric temperature decreasing correction data. Formula 1 is: ; Among them, T top The cloud top temperature, T in The temperature inside the cloud, H top H represents the height of the cloud top.base k represents the cloud base height. std For pre-set standard atmospheric temperature decrease data, r G5 For thermal verification data, k pg Corrected data for decreasing atmospheric temperature; Geometric verification data, water content verification data, particle optical verification data, dynamic wind field verification data, and thermal verification data are all cross-source verification data of one type within the cross-source verification dataset.

[0007] Optionally, based on cloud water content, liquid water content, supercooled water content, and atmospheric precipitable water, water content verification data is constructed, including: Based on cloud phase state, a cloud phase state correction coefficient is determined, and based on cloud water content, liquid water content, supercooled water content, atmospheric precipitable water, and a pre-set formula (Formula 2), water content verification data is constructed; wherein, Formula 2 is: ; Where, r G2 For water content verification data, PWV is atmospheric precipitable water, LWC is liquid water content, SLW is supercooled water content, and k Φ ρ is the cloud phase correction coefficient. w For the pre-set density of liquid water, ρ air The cloud density is a pre-set value, and CWC is the cloud water content.

[0008] Optionally, particle optical verification data is constructed based on cloud optical thickness, cloud droplet effective radius, cloud phase, liquid water content, and cloud water content, including: Based on cloud phase state, cloud phase state correction coefficients are determined, and particle optical verification data are constructed based on cloud optical thickness, cloud droplet effective radius, liquid water content, cloud water content, and a pre-set formula (Formula 3). Formula 3 is: ; Where, r G3 For particle optical calibration data, LWC represents liquid water content, SLW represents supercooled water content, and k... Φ ρ is the cloud phase correction coefficient. w R is the pre-set density of liquid water. eff Let τ be the effective radius of the cloud droplet, and τ be the optical thickness of the cloud.

[0009] Optionally, the cross-source verification dataset includes multiple verification types; Based on the cross-source verification dataset within the target region, a global observation covariance matrix corresponding to the multi-source observation dataset is constructed, including: Based on the pre-set verification type of each dimension monitoring vector in the multi-source observation dataset, the cross-source verification data corresponding to each dimension monitoring vector is selected from the cross-source verification dataset. Based on the multi-source observation dataset, the cross-source verification data corresponding to each dimension of the monitoring vector in the multi-source observation dataset, and the pre-set Formula 4, the observation standard deviation corresponding to each dimension of the monitoring vector in the multi-source observation dataset is constructed; Formula 4 is: ; Where i is the index of the monitoring vector in the multi-source observation dataset, Let $\mathbf{i}$ be the standard deviation of the observations corresponding to the monitoring vector at index $i$ in the multi-source observation dataset. The device nominal standard deviation r is pre-set for the monitoring vector of index i in the multi-source observation dataset. i This refers to the cross-source verification data corresponding to the monitoring vector in each dimension of the multi-source observation dataset. The theoretical standard deviation of the residuals is pre-set for the monitoring vector of index i in the multi-source observation dataset; Based on the observation standard deviations corresponding to all dimensions of the monitoring vectors in the multi-source observation dataset, a global observation covariance matrix corresponding to the multi-source observation dataset is constructed.

[0010] Optionally, based on the observation standard deviations corresponding to all dimensions of the monitoring vectors in the multi-source observation dataset, a global observation covariance matrix corresponding to the multi-source observation dataset is constructed, including: Based on the observation standard deviations corresponding to all dimensions of the monitoring vectors in the multi-source observation dataset, and the pre-set error correlation coefficients between every two dimensions of the monitoring vectors in the multi-source observation dataset, the covariance between every two dimensions of the monitoring vectors in the multi-source observation dataset is constructed. Based on the observation standard deviation corresponding to each dimension monitoring vector in the multi-source observation dataset, construct the observation variance corresponding to each dimension monitoring vector in the multi-source observation dataset; Based on the observation variance corresponding to each dimension monitoring vector in the multi-source observation dataset, and the covariance between every two dimensions monitoring vectors in the multi-source observation dataset, a global observation covariance matrix corresponding to the multi-source observation dataset is constructed.

[0011] Optionally, the objective function is: ; Where i is the index of the monitoring vector in the multi-source observation dataset, min J(X i Let X be the optimization objective of the monitoring vector at index i in the multi-source observation dataset. i X is the inversion data of the monitoring vector at index i in the multi-source observation dataset. i,obsLet be the monitoring vector of index i in the multi-source observation dataset, R be the global observation covariance matrix, and T be the transpose.

[0012] Optionally, based on the multi-source inversion dataset and the objective function, a multi-source inversion error dataset is constructed, including: Based on the multi-source observation dataset, the multi-source inversion dataset, and the objective function, the Hessian matrix corresponding to each inversion data in the multi-source inversion dataset is constructed. Based on the Hessian matrix corresponding to all inversion data in the multi-source inversion dataset, a multi-source inversion error dataset is constructed. The inversion error corresponding to any monitoring vector is the square root of the diagonal element in the corresponding Hessian matrix.

[0013] Optionally, based on multi-source inversion error datasets and multi-source observation datasets, a rain enhancement operation potential index for the target area is constructed, including: Based on a multi-source observation dataset, multiple job-sensitive derived variables are constructed; each job-sensitive derived variable is calculated from at least two different observation vectors. Using a multi-source inversion error dataset, determine the uncertainty of the derived quantity corresponding to the sensitive derived quantity for each operation; Based on each job-sensitive derivative quantity and the uncertainty of the derivative quantity corresponding to each job-sensitive derivative quantity, an error modulation confidence factor for each job-sensitive derivative quantity is constructed; Based on all operation-sensitive derivatives and the error modulation confidence factor of each operation-sensitive derivative, a rain enhancement operation potential index for the target area is constructed.

[0014] Secondly, embodiments of this application provide an aircraft-based artificial rain enhancement system based on multi-source observation data, including a memory, a processor, and a computer program stored in the memory. The processor executes the computer program to implement the aforementioned aircraft-based artificial rain enhancement method based on multi-source observation data.

[0015] (III) Beneficial Effects This application provides an aircraft-based artificial rain enhancement method based on multi-source observation data. It achieves anomaly identification of multi-source observation data by constructing a cross-source verification dataset, completes joint inversion of multi-source data by combining the global observation covariance matrix and objective function, and corrects the potential evaluation results based on the inversion error dataset. This effectively improves the accuracy of cloud state parameter solution and the credibility of rain enhancement potential assessment, providing reliable data support and decision-making basis for artificial rain enhancement operations. Attached Figure Description

[0016] Figure 1 A schematic flowchart of an aircraft-based artificial rain enhancement method based on multi-source observation data is provided for an embodiment of this application. Figure 2 A flowchart illustrating the construction of the global observation covariance matrix provided in this application embodiment; Figure 3 Flowchart for constructing the multi-source inversion error dataset provided in the embodiments of this application; Figure 4 A flowchart illustrating the construction of the rain enhancement operation potential index provided in this application embodiment. Detailed Implementation

[0017] To better explain and facilitate understanding of this application, the following detailed description of the application is provided in conjunction with the accompanying drawings and specific embodiments.

[0018] The development of cloud water resources is an important way to alleviate regional water shortages and improve the ecological environment, and artificial rain enhancement has been widely used as a mainstream technology. At present, artificial rain enhancement operations mainly rely on the analysis and judgment of single or limited observation data from satellites, radars, and ground meteorological stations, and multi-source observation data from satellites, airborne, and ground-based sources have not been systematically integrated. At the same time, existing schemes lack multi-source data consistency verification methods, making it difficult to identify abnormal observation information and quantitatively assess data reliability. This leads to biases in the analysis of cloud physical characteristics and significantly reduces the scientific rigor of artificial rain enhancement operations.

[0019] Therefore, this application provides an aircraft-based artificial rain enhancement method based on multi-source observation data. By constructing a cross-source verification dataset, anomaly identification of multi-source observation data is achieved. The global observation covariance matrix and objective function are combined to complete the joint inversion of multi-source data. The potential evaluation results are corrected based on the inversion error dataset, which effectively improves the accuracy of cloud state parameter solution and the credibility of rain enhancement potential assessment, providing reliable data support and decision-making basis for artificial rain enhancement operations.

[0020] To better understand the above technical solutions, exemplary embodiments of this application will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of this application are shown in the drawings, it should be understood that this application can be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this application can be understood more clearly and thoroughly, and that the scope of this application can be fully conveyed to those skilled in the art.

[0021] This application provides an aircraft-based artificial rain enhancement method based on multi-source observation data, such as... Figure 1 As shown, it includes: S1. Based on satellite-based monitoring data, space-based monitoring data, and ground-based monitoring data within the target area, construct a cross-source verification dataset within the target area to determine whether there are observational anomalies in the satellite-based monitoring data, space-based monitoring data, and ground-based monitoring data; the cross-source verification dataset is data that quantifies the physical deviations between different observation sources; S2. Once it is determined that there are no anomalies in the satellite-based monitoring data, space-based monitoring data, and ground-based monitoring data, a multi-source observation dataset is constructed based on the satellite-based monitoring data, space-based monitoring data, and ground-based monitoring data; wherein, the multi-source observation dataset includes multi-dimensional monitoring vectors; S3. Based on the cross-source verification dataset within the target area, construct the global observation covariance matrix corresponding to the multi-source observation dataset, and construct the multi-source inversion dataset based on the global observation covariance matrix and the pre-set objective function; wherein, the multi-source inversion dataset includes the inversion data corresponding to each dimension monitoring vector in the multi-source observation dataset; S4. Based on the multi-source inversion dataset and objective function, construct a multi-source inversion error dataset, and based on the multi-source inversion error dataset and multi-source observation dataset, construct a rain enhancement operation potential index for the target area to assist in artificial rain enhancement; wherein, the multi-source inversion error dataset includes the inversion error corresponding to each dimension monitoring vector in the multi-source observation dataset.

[0022] This application provides an aircraft-based artificial rain enhancement method based on multi-source observation data. It accurately identifies observational anomalies in satellite-based, airborne, and ground-based monitoring data to ensure data quality from the source. Then, it integrates effective observational data to form a multi-dimensional monitoring vector. Using the global observation covariance matrix and an objective function, it performs joint inversion to obtain high-precision inversion data. Simultaneously, it combines the inversion results with the objective function to quantify the inversion errors of each dimension of data, constructing an inversion error dataset. Finally, it calculates the rain enhancement operation potential index by comprehensively considering the observational data and error data. This method achieves integrated processing of multi-source data quality verification, high-precision inversion, error quantification, and potential assessment, effectively reducing interference from abnormal data, observational errors, and inversion errors, and improving the accuracy of cloud parameter solutions and the reliability of rain enhancement potential assessment results.

[0023] Furthermore, in the embodiments of this application, after any range is gridded, one of the target grids is the target region.

[0024] Furthermore, in the embodiments of this application, the satellite-based monitoring devices include, but are not limited to, the Fengyun series (FY-2, FY-3, FY-4) satellites, Cloudsat satellites, and MODIS satellites; the air-based monitoring devices include, but are not limited to, the King Air 350 aircraft carrying a cloud physics integrated observation platform; and the ground-based monitoring devices include, but are not limited to, X-band weather radar, Ka-band cloud radar, micro-rain radar, ground-based microwave radiometer, ceilometer, ground-based GNSS, laser raindrop spectrometer, radiosonde, and ground meteorological station.

[0025] Optionally, in a specific embodiment, the satellite-based monitoring data includes cloud top height, cloud top temperature, cloud optical thickness, cloud water content, and cloud droplet effective radius; the space-based monitoring data includes cloud temperature, horizontal wind speed, liquid water content, and supercooled water content; the ground-based monitoring data includes atmospheric precipitable water, radar echo intensity baseline, cloud movement speed, ground precipitation, and cloud phase; the cross-source verification dataset includes cross-source verification data of various verification types. Based on satellite-based, space-based, and ground-based monitoring data within the target area, a cross-source verification dataset for the target area is constructed, including: Based on the radar echo intensity baseline, the cloud base height is determined, and based on the cloud top height, cloud base height, and pre-set geometric correction parameters, geometric verification data is determined; wherein, the cloud base height is the altitude corresponding to the radar echo intensity baseline -15dBz; the geometric verification data is the difference between the cloud top height, cloud base height, and geometric correction parameters; Based on cloud water content, liquid water content, supercooled water content, atmospheric precipitable water, and cloud phase, water content verification data is constructed. Particle optical verification data are constructed based on cloud optical thickness, cloud droplet effective radius, cloud phase, liquid water content, and cloud water content. Based on horizontal wind speed and cloud movement speed, dynamic wind field verification data is determined; wherein, dynamic wind field verification data is the vector magnitude of the difference between cloud movement speed and horizontal wind speed; Based on surface precipitation and a pre-set precipitation correction curve, atmospheric temperature decreasing correction data is determined. Based on this atmospheric temperature decreasing correction data, cloud temperature, cloud top temperature, cloud top height, and cloud base height, and a pre-set formula (Formula 1), thermal verification data is determined. The precipitation correction curve is the corresponding curve for surface precipitation and atmospheric temperature decreasing correction data. Formula 1 is: ; Among them, T top The cloud top temperature, T in The temperature inside the cloud, H top H represents the height of the cloud top. base The cloud base height, k std For pre-set standard atmospheric temperature decrease data, r G5 For thermal verification data, k pg Corrected data for decreasing atmospheric temperature; Geometric verification data, water content verification data, particle optical verification data, dynamic wind field verification data, and thermal verification data are all cross-source verification data of one type within the cross-source verification dataset.

[0026] Furthermore, based on cloud water content, liquid water content, supercooled water content, and atmospheric precipitable water, water content verification data is constructed, including: Based on cloud phase state, a cloud phase state correction coefficient is determined, and based on cloud water content, liquid water content, supercooled water content, atmospheric precipitable water, and a pre-set formula (Formula 2), water content verification data is constructed; wherein, Formula 2 is: ; Where, r G2 For water content verification data, PWV is atmospheric precipitable water, LWC is liquid water content, SLW is supercooled water content, and k Φ ρ is the cloud phase correction coefficient. w For the pre-set density of liquid water, ρ air The cloud density is a pre-set value, and CWC is the cloud water content.

[0027] Furthermore, based on cloud optical thickness, cloud droplet effective radius, cloud phase, liquid water content, and cloud water content, particle optical verification data are constructed, including: Based on cloud phase state, cloud phase state correction coefficients are determined, and particle optical verification data are constructed based on cloud optical thickness, cloud droplet effective radius, liquid water content, cloud water content, and a pre-set formula (Formula 3). Formula 3 is: ; Where, r G3 For particle optical calibration data, LWC represents liquid water content, SLW represents supercooled water content, and k... Φ ρ is the cloud phase correction coefficient. w R is the pre-set density of liquid water. eff Let τ be the effective radius of the cloud droplet, and τ be the optical thickness of the cloud.

[0028] Specifically, this application embodiment takes a typical artificial rain enhancement operation in a stratiform cloud as an example, and selects grid point P for complete numerical calculation, which specifically includes: The satellite-based monitoring data obtained by the satellite-based monitoring device includes a cloud top altitude of 7820m, a cloud top temperature of -22.4℃, a cloud optical thickness of 24.6m, and a cloud water content of 0.42g / m². 3 The effective radius of the cloud droplet is 18.2 μm.

[0029] The airborne monitoring data obtained by the airborne monitoring device includes cloud temperature -8.6℃, horizontal wind speed (12.6m / s, 4.8m / s), and liquid water content of 0.31g / m³. 3 and supercooled water content 0.30g / m 3 .

[0030] The ground monitoring data obtained by the ground monitoring device include atmospheric precipitable water of 18.2 mm, radar echo intensity baseline, cloud movement speed (13.4 m / s, 5.2 m / s), surface precipitation of 0.46 mm / h, and cloud phase M mixed phase.

[0031] Since satellites focus on overall macroscopic remote sensing of clouds, aircraft detect local parameters within clouds, and radar and ground equipment focus on cloud base and near-surface features, single-dimensional verification cannot comprehensively identify data problems. Therefore, this embodiment sets up five types of cross-source verification: geometry, water content, particle optics, dynamic wind field, and thermodynamics. This covers five core dimensions of cloud physics: cloud geometry, water vapor conservation, cloud microphysical optical properties, dynamic motion characteristics, and thermal stratification. This serves as the prerequisite for conducting multi-source data quality control, error correction, joint inversion, and rain enhancement potential assessment.

[0032] The cross-source verification data for five verification types is calculated using the aforementioned space-based, satellite-based, and ground-based monitoring data, including the following process: First, the cloud base location is identified based on the spatial distribution characteristics of the radar echo intensity baseline. The bottom of the cloud is a transitional region where radar echoes transition from non-existent to present. In this embodiment, a fixed echo intensity threshold is selected as the basis for cloud base determination, and the altitude corresponding to this threshold is extracted as the cloud base height of the target area. In this embodiment, the radar echo intensity baseline is the baseline formed by the radar echo intensities corresponding to different altitudes obtained from the spatial scanning data of the ground-based radar in the target area. Specifically, in this embodiment, a radar echo intensity of -15 dBz is used as the critical value for cloud base identification. This value represents a typical weak echo range for the cloud base, which can effectively distinguish between clear sky areas and cloud areas, avoiding interference from near-ground clutter and aerosol echoes.

[0033] The cloud base height H is obtained from the radar echo intensity at -15dBz. base =3320m, according to the cloud top height H top Cloud base height H base and the pre-set geometric verification tolerance ξ geom Determine the geometric verification data r G1 The calculation formula is: r G1 =H top -H base -ξ geom The geometric calibration tolerance is a pre-set fixed deviation allowable value used to characterize the normal systematic deviation threshold between two types of observation data: satellite-based cloud top altitude and radar-inverted cloud bottom altitude, caused by differences in the detection platform, observation principle, and spatial sampling. In this embodiment, it is set to 300m. Substituting the data from this embodiment, we get 7820-4200-300=3320m.

[0034] The ensemble verification residual (geometric verification data) of 3320m indicates that the cloud top height has a significant thickness structure relative to the radar-detected cloud base, which is consistent with the typical geometric characteristics of layered clouds.

[0035] Secondly, based on cloud phase state, a cloud phase state correction coefficient is determined, and based on cloud water content, liquid water content, supercooled water content, atmospheric precipitable water, and a pre-set formula two, water content verification data is constructed; wherein, formula two is: ; Where, r G2 For water content verification data, PWV is atmospheric precipitable water, LWC is liquid water content, SLW is supercooled water content, and k Φ ρ is the cloud phase correction coefficient. w For the pre-set density of liquid water, ρ air The cloud density is a pre-set value, and CWC is the cloud water content.

[0036] The cloud phase correction coefficient is a pre-set correction parameter based on the physical phase of the cloud. It is used to convert cloud water content (CWC), unify the condensate equivalent of clouds in different phases, and adapt to water vapor mass conservation calculations and cloud optical theory calculations. Cloud phases are jointly identified by ground-based radar and satellite-based systems, and are mainly divided into three categories: pure liquid clouds, mixed-phase clouds, and pure ice-phase clouds. Different phases correspond to different correction coefficients. Generally, the cloud phase correction coefficient is 1.0 for pure liquid clouds, 0.7 for mixed-phase clouds, and 0.4 for pure ice-phase clouds.

[0037] In this embodiment, the cloud thickness is 7820-4200=3620m. Therefore, assuming a uniform distribution of water content within the cloud, the result is: =(0.31+0.30+0.7×0.42)×3620=0.902×3620=3265.24g / m 2 .

[0038] but, =3265.24 / (1000×0.78)=4.186mm.

[0039] r G2 =18.2-4.186=14.014mm.

[0040] The residual for water content verification (water content verification data) is 14.014 mm, indicating that there is a systematic deviation between the satellite-based precipitable water observation and the space-based water content integration, which may be related to the fact that the cloud body did not completely fill the observation volume.

[0041] Furthermore, as mentioned above, cloud phase correction coefficients are determined based on cloud phase state.

[0042] Based on cloud optical thickness, cloud droplet effective radius, liquid water content, cloud water content, and a pre-set formula (Formula 3), particle optical verification data is constructed; formula 3 is: ; Where, rG3 For particle optical calibration data, LWC represents liquid water content, SLW represents supercooled water content, and k... Φ ρ is the cloud phase correction coefficient. w R is the pre-set density of liquid water. eff Let τ be the effective radius of the cloud droplet, and τ be the optical thickness of the cloud.

[0043] In this embodiment, the numerator is 3×(0.31+0.7×0.42)=1.812g / m 3 The denominator is 2 × 1000 × 18.2 × 10⁻⁶ = 0.0364 kg / m³ 3 The integral result is: 49.78×3620×10-3=0.1802, and the particle optics verification residual (particle optics verification data) is 24.6-0.1802=24.420.

[0044] In this embodiment, the particle optical verification residual is 24.42. The optical thickness retrieved by MODIS is much larger than the theoretical integral value based on the assumptions of airborne water content and particle spectrum, reflecting the difference in spatial representativeness between satellite and aircraft observations.

[0045] Furthermore, dynamic wind field verification data is determined based on horizontal wind speed and cloud movement speed; wherein, the dynamic wind field verification data is the vector magnitude of the difference between cloud movement speed and horizontal wind speed, i.e.: In this embodiment, the dynamic wind field verification data is the magnitude of the vector difference between cloud movement speed and horizontal wind speed, calculated using the formula: r G4 =||U cld -U h In this embodiment, the wind field verification residual (wind field verification data) is 0.894 m / s.

[0046] In this embodiment, the wind field verification residual is 0.894 m / s, and the cloud movement speed retrieved by TREC has good consistency with the wind field measured by the aircraft.

[0047] Finally, based on surface precipitation and a pre-set precipitation correction curve, atmospheric temperature decrease correction data is determined. Based on the atmospheric temperature decrease correction data, cloud temperature, cloud top temperature, cloud top height, and cloud base height, and a pre-set formula (Formula 1), thermal verification data is determined. The precipitation correction curve is the corresponding curve for surface precipitation and atmospheric temperature decrease correction data. Formula 1 is: ; Among them, T top The cloud top temperature, T in The temperature inside the cloud, H top H represents the height of the cloud top. base The cloud base height, k stdFor pre-set standard atmospheric temperature decrease data, r G5 For thermal verification data, k pg This is a correction for decreasing atmospheric temperature.

[0048] That is, in this embodiment, the temperature difference is -22.4 - (-8.6) = -13.8℃, the height difference is 3620m = 3.62km, and the overall temperature lapse rate is 6.5 + 0.8 = 7.3℃ / km.

[0049] Therefore, the thermal verification residual (thermal verification data) is -0.522 (dimensionless), and the cloud temperature stratification conforms to the typical characteristics of a stratiform cloud.

[0050] In summary, the results of the five types of cross-source verification are summarized as follows: The geometric verification data is 3320m, with a judgment threshold of ±500m, and the judgment result is that it exceeds the threshold; the moisture content verification data is 14.014mm, with a judgment threshold of ±5mm, and the judgment result is that it exceeds the threshold; the particle optical verification data is 24.420, with a judgment threshold of ±10, and the judgment result is that it exceeds the threshold; the dynamic wind field verification data is 0.894m / s, with a judgment threshold of ±2m / s, and the judgment result is that it passes; the thermal verification data is -0.522, with a judgment threshold of ±1, and the judgment result is that it passes.

[0051] The residuals exceeding the threshold terms reflect systematic differences in spatial representativeness, observation principles, and inversion algorithms among satellite-based, space-based, and ground-based observations, which are normal phenomena in multi-source data fusion scenarios. These differences are the physical basis for optimal fusion through covariance matrices and objective functions, rather than data anomalies. In other words, the multi-source observation data at grid point P does not exhibit anomalies caused by instrument malfunctions or significant gross errors, and all five types of verification residuals are within interpretable limits.

[0052] This embodiment uses five types of cross-source verification data to conduct comprehensive consistency verification of multi-source observation data from space-based, air-based, and ground-based sources, based on five core dimensions: cloud geometry, water vapor conservation, cloud microphysical and optical properties, dynamic motion characteristics, and thermo-stratification. This effectively distinguishes between instrument gross errors, observational anomalies, and normal systematic deviations caused by different detection platforms due to their observation principles, spatial sampling, and inversion algorithms, avoiding misjudgments or direct rejection of valid observation data. At the same time, various verification residuals can serve as important bases for error correction, providing a realistic bias foundation for subsequent global observation covariance matrix construction and multi-source data joint inversion.

[0053] Optionally, in a specific embodiment, a global observation covariance matrix corresponding to the multi-source observation dataset is constructed based on the cross-source verification dataset within the target region, such as... Figure 2 As shown, it includes: S31. Based on the pre-set verification type of each dimension monitoring vector in the multi-source observation dataset, filter out the cross-source verification data corresponding to each dimension monitoring vector from the cross-source verification dataset. S32. Based on the multi-source observation dataset, the cross-source verification data corresponding to each dimension monitoring vector in the multi-source observation dataset, and the pre-set Formula 4, construct the observation standard deviation corresponding to each dimension monitoring vector in the multi-source observation dataset; Formula 4 is: ; Where i is the index of the monitoring vector in the multi-source observation dataset, Let $\mathbf{i}$ be the standard deviation of the observations corresponding to the monitoring vector at index $i$ in the multi-source observation dataset. The device nominal standard deviation r is pre-set for the monitoring vector of index i in the multi-source observation dataset. i This refers to the cross-source verification data corresponding to the monitoring vector in each dimension of the multi-source observation dataset. The theoretical standard deviation of the residuals is pre-set for the monitoring vector of index i in the multi-source observation dataset; S33. Based on the observation standard deviations of all dimensions of the monitoring vectors in the multi-source observation dataset, construct the global observation covariance matrix corresponding to the multi-source observation dataset.

[0054] Furthermore, based on the observation standard deviations corresponding to all dimensions of the monitoring vectors in the multi-source observation dataset, a global observation covariance matrix corresponding to the multi-source observation dataset is constructed, including: Based on the observation standard deviations corresponding to all dimensions of the monitoring vectors in the multi-source observation dataset, and the pre-set error correlation coefficients between every two dimensions of the monitoring vectors in the multi-source observation dataset, the covariance between every two dimensions of the monitoring vectors in the multi-source observation dataset is constructed. Based on the observation standard deviation corresponding to each dimension monitoring vector in the multi-source observation dataset, construct the observation variance corresponding to each dimension monitoring vector in the multi-source observation dataset; Based on the observation variance corresponding to each dimension monitoring vector in the multi-source observation dataset, and the covariance between every two dimensions monitoring vectors in the multi-source observation dataset, a global observation covariance matrix corresponding to the multi-source observation dataset is constructed.

[0055] Specifically, the multi-source observation dataset in this embodiment contains 14-dimensional monitoring vectors, namely, cloud top height, cloud top temperature, cloud optical thickness, cloud water content, cloud droplet effective radius, cloud temperature, horizontal wind speed, liquid water content, supercooled water content, atmospheric precipitable water, radar echo intensity baseline, cloud movement speed, surface precipitation, and cloud phase. Based on the attributes of each dimension of the monitoring vector, matching verification residuals r are selected from the cross-source verification dataset. iFor example, cloud top height and radar echo intensity are matched with geometric verification residuals; various water contents and atmospheric precipitable water are matched with water content verification residuals; temperature variables are matched with thermal verification residuals; wind speed variables are matched with dynamic wind field verification residuals; and cloud optical thickness and cloud droplet effective radius are matched with particle optical verification residuals, thus completing the pairing of residuals across all dimensions.

[0056] Substituting the nominal standard deviation of the equipment, the calibration residual, and the theoretical standard deviation of the residual into Formula 4, the observation standard deviation of each monitoring vector is calculated dimension by dimension. Among them, the nominal standard deviation of the equipment, also known as the instrument's inherent standard deviation, is the inherent measurement error specified in the factory calibration and industry standards for various meteorological observation equipment, including satellite-based, space-based, and ground-based equipment. It is the standard deviation obtained by repeatedly measuring the same physical quantity under standard laboratory environment and ideal detection conditions. The theoretical standard deviation of the residual is a preset theoretical reference value. For a certain type of cross-source calibration (geometric, water content, thermal, etc.), the standard residual dispersion is calculated based on a large number of historical observation samples and theoretical models.

[0057] By performing calculations for all 14 dimensions as described above, a complete sequence of observed standard deviations is obtained. The observed variance is the square of the observed standard deviation. The observed variances for all dimensions are calculated sequentially and used as the diagonal elements of the global observed covariance matrix.

[0058] Based on meteorological physical laws, the error correlation coefficients for different variable groups were pre-set: wind field variable ρ=0.95, cloud top temperature and cloud interior temperature ρ=0.85, liquid water content and supercooled water content ρ=0.70, and the correlation coefficients for variables with no obvious physical correlation were set to 0.

[0059] Taking the x-components of horizontal wind speed and cloud movement speed as an example: the covariance between the two is 0.95× × ;in, This represents the observed standard deviation for horizontal wind speed. This represents the observation standard deviation corresponding to the cloud movement speed.

[0060] Similarly, calculate the covariance between all pairs of dimensions and use it as the off-diagonal element of the covariance matrix.

[0061] The diagonal positions of the matrix are filled with the observation variances of each dimension, and the off-diagonal positions are filled with the covariances between each pair of variables, finally yielding a 14th-order global observation covariance matrix R. This matrix fully contains the observation errors of each variable and the error correlations between variables.

[0062] This embodiment provides a reliable error constraint basis for subsequent objective function solving and multi-source data joint inversion, effectively improving the accuracy of multi-source data fusion and ensuring the scientificity and accuracy of subsequent inversion results and rain enhancement operation potential index.

[0063] Optionally, in a specific embodiment, the objective function is: ; Where i is the index of the monitoring vector in the multi-source observation dataset, min J(X i Let X be the optimization objective of the monitoring vector at index i in the multi-source observation dataset. i X is the inversion data of the monitoring vector at index i in the multi-source observation dataset. i,obs Let be the monitoring vector of index i in the multi-source observation dataset, R be the global observation covariance matrix, and T be the transpose.

[0064] Specifically, in this embodiment, the objective function adopts the least binary objective function. The physical meaning of this objective function is: to characterize the error level and correlation of each dimension of data with the global observation covariance matrix, minimize the weighted residual between the observed value and the inverted value, and solve for the optimal inversion result that fits the true state of the cloud.

[0065] Perform matrix inversion on the constructed 14th-order global observation covariance matrix R to obtain the inverse matrix R. -1 This inverse matrix is ​​the core weight matrix for solving the objective function. The larger the observation error, the lower the constraint weight in the objective function, thus achieving adaptive weighting of multi-source data.

[0066] The Newton-Gaussian iterative algorithm is used to find the minimum value of the objective function, and optimization calculations are performed on each of the 14-dimensional monitoring vectors: Calculate the gradient of the objective function and the Hessian matrix, combine them with the inverse matrix, adjust the iteration step size, and continuously update the variable values ​​until the objective function converges. The iteration terminates, and the value obtained after convergence is the optimal inversion data corresponding to that dimension.

[0067] The inversion data obtained by solving all 14 dimensions are integrated in the original dimensional order to form a multi-source inversion dataset.

[0068] This embodiment utilizes the global observation covariance matrix to quantify the error magnitude and variable correlation of multi-source observation data, and conducts joint inversion through a preset weighted least squares objective function. Compared to directly using raw observation data, this method can fully compensate for the systematic biases and random errors introduced by different detection platforms (satellite, space, and ground), weaken the interference of low-precision observation data on the results, and obtain inversion data that more closely reflects the true macro- and micro-states of clouds. The completed multi-source inversion dataset can be directly used for subsequent inversion error calculation, solving for operational sensitivity derivatives, and constructing a rain enhancement operation potential index, providing high-precision basic data for the assessment of artificial rain enhancement operations.

[0069] Optionally, in a specific embodiment, a multi-source inversion error dataset is constructed based on the multi-source inversion dataset and the objective function, such as... Figure 3As shown, it includes: S41. Based on the multi-source observation dataset, the multi-source inversion dataset, and the objective function, construct the Hessian matrix corresponding to each inversion data in the multi-source inversion dataset; S42. Based on the Hessian matrix corresponding to all inversion data in the multi-source inversion dataset, construct a multi-source inversion error dataset. The inversion error corresponding to any monitoring vector is the square root of the diagonal element in the corresponding Hessian matrix.

[0070] Specifically, in this embodiment, the Hessian matrix is ​​a matrix composed of the second-order partial derivatives of the objective function with respect to the inverse variable. In the univariate case, the Hessian matrix degenerates into a second-order derivative scalar, expressed as: ; Among them, H J,i The objective function J is related to the inversion variable X. i The Hessian matrix, the diagonal elements of the matrix (H J,i ) ii The second derivative corresponding to the i-th dimension variable is used to characterize the curvature of the objective function at the optimal inversion solution and is the core matrix for solving the inversion error.

[0071] Once the diagonal elements of the matrix for each inversion vector are obtained, the inversion error corresponding to the monitoring vector is calculated. The inversion error is the square root of the diagonal element of the Hessian matrix in that dimension.

[0072] The inversion errors corresponding to the 14-dimensional monitoring vectors are organized and summarized according to the same dimensional order as the multi-source observation dataset and the multi-source inversion dataset, ultimately forming a multi-source inversion error dataset. This dataset fully includes the inversion error σ corresponding to each dimension of the monitoring vector in the multi-source observation dataset. retrieval,i It can be directly used to calculate subsequent derivative uncertainties, error modulation confidence factors, and rain enhancement operation potential indices.

[0073] This embodiment obtains the Hessian matrix by solving the second-order partial derivatives of the objective function, and calculates the inversion error in each dimension using the diagonal elements of the Hessian matrix, quantifying the error level introduced during the joint inversion of multi-source data. The magnitude of the inversion error is negatively correlated with the observation standard deviation: the higher the observation accuracy and the smaller the observation standard deviation, the larger the inversion error, and the higher the weight of this variable in the joint inversion. This method achieves standardized and quantitative extraction of inversion errors. The constructed multi-source inversion error dataset is the core input for subsequent calculations of the uncertainty of sensitive derivatives and the confidence factor of error modulation, effectively quantifying the data reliability of the inversion process and further improving the calculation accuracy and reliability of the rain enhancement operation potential index.

[0074] Optionally, in a specific embodiment, a rain enhancement operation potential index for the target area is constructed based on a multi-source inversion error dataset and a multi-source observation dataset, such as... Figure 4As shown, it includes: S43. Based on a multi-source observation dataset, construct multiple job-sensitive derived quantities; each job-sensitive derived quantity is calculated from at least two different observation vectors; S44. Multi-source inversion error dataset, determine the uncertainty of the derived quantity corresponding to the sensitive derived quantity of each operation; S45. Based on the sensitive derivative quantity of each operation and the uncertainty of the derivative quantity corresponding to each sensitive derivative quantity, construct the error modulation confidence factor of each sensitive derivative quantity of the operation; S46. Based on all operation-sensitive derivatives and the error modulation confidence factor of each operation-sensitive derivative, construct the rain enhancement operation potential index for the target area.

[0075] Specifically, in this embodiment, based on the core physical conditions of artificial rain enhancement using cold stratospheric clouds, five types of operationally sensitive derived quantities are constructed. Each type integrates multi-source observation vectors. The definitions, participating variables, and calculation formulas of each derived quantity are as follows: The Vertical Accumulated Supercooled Column (VISLW) (a core indicator of supercooled catalysis) involves vectors such as satellite-based cloud top height, ground-based cloud bottom height, airborne supercooled water content, and cloud temperature. Its calculation formula is as follows: ; Where I(·) is the interval indicator function, only retaining the integral results of the effective subcooled water temperature zone.

[0076] The warm cloud collision-merger potential (WCCP) (warm cloud catalytic index) involves vectors such as airborne liquid water content, effective radius of star-based cloud droplets, and cloud temperature. Its calculation formula is as follows: WCCP=LWC·[1-(R eff / R0)]·I(T in >0℃); Where R0 is a pre-set reference threshold for the cloud base radius.

[0077] The Cloud Dynamic Support Index (CDSI) incorporates vectors including satellite-based cloud top height, ground-based cloud base height, and airborne horizontal wind speed. Its calculation formula is as follows: CDSI=||U h ||·[(H top -H base ) / H ref ]; Among them, H ref This is the normalized baseline constant for cloud thickness.

[0078] The water abundance index CW AI, whose participating vectors include ground-based atmospheric precipitable water, airborne liquid water / supercooled water, and spaceborne cloud water content, is calculated using the following formula: CW AI=α[(LWC+SLW) / LWC ref]+β(CWC / CWC ref )+γ(PWV / PWV ref ); Among them, LWC ref CWC serves as a reference threshold for liquid water content. ref PWV is the reference threshold for cloud water content. ref α is the reference threshold for atmospheric precipitable water, and β and γ are pre-set weighting coefficients.

[0079] The cloud-wind consistency ΔU (negative penalty index) involves horizontal wind speed and cloud movement speed as vectors, and its calculation formula is as follows: △U=||Ucld-Uh||.

[0080] In this embodiment, by substituting the multi-source observation values ​​of grid point P, the numerical solutions for five types of derived quantities are completed, and the original set of derived quantities is obtained.

[0081] The total uncertainty σ of the derived quantity is calculated using the full covariance error propagation formula. D : Suppose the error propagation formula for a certain derived quantity D is: ; Among them, v k Let σ be the k-th dimension monitoring vector. retrieval,k Let v be the inversion error corresponding to the k-th dimension monitoring vector. m Let Cov(v) be the m-th dimension monitoring vector. k ,v m ) for v k With v m The error covariance between them.

[0082] That is, the calculation process is as follows: for each type of job-sensitive derived quantity, calculate the partial derivatives of all its input observation vectors, read the inversion error of the corresponding dimension of the multi-source inversion error dataset, combine it with the variable covariance, substitute it into the error propagation formula, and take the square root to obtain the derived quantity uncertainty. The uncertainty of the five types of derived quantities is completed in sequence.

[0083] The normalized confidence factor calculation formula is used to quantify the data reliability of single-class derivatives: Q D =1 / [1+(σ D / D)]; D represents the calculated value of the sensitive derivative quantity in the current operation, Q D is the error modulation confidence factor, with a value range of (0,1).

[0084] The five types of original derived quantities are globally normalized and uniformly mapped to the interval [0,1].

[0085] Based on the above-mentioned original derived quantities multiplied by the data credibility and the pre-set index weight, the potential index corresponding to each type of original derived quantity is obtained. The potential indices corresponding to all original derived quantities are added together to obtain the final rain enhancement operation potential index of the target area, that is, the final rain enhancement operation potential index of grid point P.

[0086] In this embodiment, grid point P is a mixed-phase layered cold cloud with abundant supercooled water resources, stable cloud dynamics, and small cloud-wind deviation. The calculated confidence factors of the vertical cumulative supercooled water column, cloud dynamic support index, and water content abundance index are all close to 1. The confidence factor of cloud-wind consistency deviation is high and the penalty term is small. The final synthesized rain enhancement operation potential index is in the high potential range, and it is determined that the area is suitable for carrying out cold cloud aircraft artificial rain enhancement catalysis operations.

[0087] This embodiment achieves objective quantitative classification of cloud conditions, which can accurately distinguish the operational areas suitable for cold and warm cloud catalysis, and also intuitively reflect the data reliability of the evaluation results. It provides a quantitative decision-making basis that combines physical rigor and error reliability for the selection of artificial rain enhancement operation areas, the formulation of catalysis schemes, and the planning of aircraft operation routes.

[0088] Furthermore, after obtaining the final rain enhancement operation potential index for all grid points within the entire range, the entire spatial grid is traversed, and the grid potential index is divided into four levels: high potential, medium potential, low potential, and no potential, based on historical artificial rain enhancement operation thresholds. Simultaneously, the cloud phase and temperature stratification data of each grid are combined to label the appropriate catalytic type. High-potential mixed phase / cold cloud grids are marked as cold cloud aircraft catalytic zones, high-potential warm cloud grids are marked as warm cloud hygroscopic particle operation zones, and low-potential grids are marked with limiting factors such as insufficient water vapor, cloud dispersion, or poor dynamic conditions, forming a global hierarchical labeling layer.

[0089] The error modulation confidence factor corresponding to the five types of derived quantities for each grid is retrieved, and the overall confidence level of the grid is calculated. The entire grid is then stratified: grids with high overall confidence levels are designated as core reliable operation identification areas and are directly used for operation planning; grids with low confidence levels and large inversion errors are marked as suspicious grids and are not given priority for operation. Error source tracing records are also retained to facilitate subsequent verification of observation data and inversion results; invalid grids with confidence levels close to 0 and obvious biases in observation data are removed to avoid distorted results interfering with business decisions.

[0090] An eight-neighbor connectivity algorithm is used to spatially cluster potential grids of the same level, eliminating isolated and scattered small-area high-potential grids, extracting continuous high and medium-potential cloud regions, and outputting the range of closed polygonal target areas. Simultaneously, the average potential index, average comprehensive confidence, cloud vertical thickness, and duration of continuous prediction are statistically analyzed for each contiguous target area, distinguishing between short-term scattered cloud clusters and stable layered cloud regions that can be operated for a long time, providing spatial boundaries for aircraft to plan continuous flight paths.

[0091] By overlaying operational constraint layers such as regional terrain elevation, no-fly airspace, airport take-off and landing range, and radar detection blind spots, contiguous potential target areas are cropped and corrected: grids within high mountain obstructions and no-fly airspaces are removed, large target areas exceeding the one-way flight range of aircraft are broken down, and segmented operational sub-regions are divided; at the same time, the movement and dissipation trends of target areas in the next 1-3 hours are extrapolated based on the time series of grid cloud movement speed, and the effective operational window period of each potential target area is predicted.

[0092] Output grid potential index distribution map, graded target area vector boundary, grid confidence layer, catalysis type partition map, as well as the area ratio of each level of potential grid in the whole domain, the comprehensive potential score of each contiguous target area, the effective operation time, and the data reliability statistics table. Overlay the map base map on the meteorological command platform to render the potential index heat map, highlight the high confidence contiguous operation area, and mark the recommended catalysis method and operation priority level of the target area.

[0093] Based on the priority level of contiguous target areas, flight range, and window timing, multiple sets of aircraft operation route planning schemes are automatically generated, including takeoff time, target area coordinates, catalyst dissemination duration, and return point; simultaneously, grid potential classification, target area boundaries, recommended catalyst types, and data confidence notes are pushed to ground command and airborne operation terminals for on-site reference by operators.

[0094] In addition, this application provides an aircraft-based artificial rain enhancement system based on multi-source observation data, including a memory, a processor, and a computer program stored in the memory. The processor executes the computer program to implement the above-mentioned aircraft-based artificial rain enhancement method based on multi-source observation data.

[0095] In the description of this application, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0096] In this application, unless otherwise expressly specified and limited, the terms "installation," "connection," "linking," and "fixing," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. Those skilled in the art can understand the specific meaning of the above terms in this application according to the specific circumstances.

[0097] In this application, unless otherwise expressly specified and limited, "above" or "below" the second feature can mean that the first and second features are in direct contact, or that they are in indirect contact through an intermediate medium. Furthermore, "above," "on top of," and "over" the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.

[0098] In the description of this specification, the terms "one embodiment," "some embodiments," "embodiment," "example," "specific example," or "some examples," etc., refer to specific features, structures, materials, or characteristics described in connection with that embodiment or example, which are included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0099] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make modifications, alterations, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for artificial rain enhancement based on multi-source observation data from aircraft, characterized in that, include: Based on satellite-based, space-based, and ground-based monitoring data within the target area, a cross-source verification dataset is constructed within the target area to determine whether there are observational anomalies in the satellite-based, space-based, and ground-based monitoring data; the cross-source verification dataset is data that quantifies the physical deviations between different observation sources; Once it is confirmed that there are no anomalies in the satellite-based, space-based, and ground-based monitoring data, a multi-source observation dataset is constructed based on these data; the multi-source observation dataset includes multi-dimensional monitoring vectors. Based on the cross-source verification dataset within the target area, a global observation covariance matrix corresponding to the multi-source observation dataset is constructed. Based on the global observation covariance matrix and a pre-set objective function, a multi-source inversion dataset is constructed. The multi-source inversion dataset includes the inversion data corresponding to each dimension monitoring vector in the multi-source observation dataset. Based on the multi-source inversion dataset and objective function, a multi-source inversion error dataset is constructed. Based on the multi-source inversion error dataset and the multi-source observation dataset, a rain enhancement operation potential index for the target area is constructed to assist in artificial rain enhancement. The multi-source inversion error dataset includes the inversion error corresponding to each dimension monitoring vector in the multi-source observation dataset.

2. The aircraft-based artificial rain enhancement method based on multi-source observation data according to claim 1, characterized in that, The satellite-based monitoring data includes cloud top height, cloud top temperature, cloud optical thickness, cloud water content, and cloud droplet effective radius; the space-based monitoring data includes cloud temperature, horizontal wind speed, liquid water content, and supercooled water content; the ground-based monitoring data includes atmospheric precipitable water, radar echo intensity baseline, cloud movement speed, ground precipitation, and cloud phase; the cross-source verification dataset includes cross-source verification data of various verification types. Based on satellite-based, space-based, and ground-based monitoring data within the target area, a cross-source verification dataset for the target area is constructed, including: Based on the radar echo intensity baseline, the cloud base height is determined, and based on the cloud top height, cloud base height, and pre-set geometric correction parameters, geometric verification data is determined; wherein, the cloud base height is the altitude corresponding to the radar echo intensity baseline -15dBz; the geometric verification data is the difference between the cloud top height, cloud base height, and geometric correction parameters; Based on cloud water content, liquid water content, supercooled water content, atmospheric precipitable water, and cloud phase, water content verification data is constructed. Particle optical verification data are constructed based on cloud optical thickness, cloud droplet effective radius, cloud phase, liquid water content, and cloud water content. Based on horizontal wind speed and cloud movement speed, dynamic wind field verification data is determined; whereby, dynamic wind field verification data is the vector magnitude of the difference between cloud movement speed and horizontal wind speed. Based on surface precipitation and a pre-set precipitation correction curve, atmospheric temperature decreasing correction data is determined. Based on this atmospheric temperature decreasing correction data, cloud temperature, cloud top temperature, cloud top height, and cloud base height, and a pre-set formula (Formula 1), thermal verification data is determined. The precipitation correction curve is the corresponding curve for surface precipitation and atmospheric temperature decreasing correction data. Formula 1 is: ； Among them, T top The cloud top temperature, T in The temperature inside the cloud, H top H represents the height of the cloud top. base k represents the cloud base height. std For pre-set standard atmospheric temperature decrease data, r G5 For thermal verification data, k pg Corrected data for decreasing atmospheric temperature; Geometric verification data, water content verification data, particle optical verification data, dynamic wind field verification data, and thermal verification data are all cross-source verification data of one type within the cross-source verification dataset.

3. The aircraft-based artificial rain enhancement method based on multi-source observation data according to claim 2, characterized in that, Based on cloud water content, liquid water content, supercooled water content, and atmospheric precipitable water, water content verification data is constructed, including: Based on cloud phase state, a cloud phase state correction coefficient is determined, and based on cloud water content, liquid water content, supercooled water content, atmospheric precipitable water, and a pre-set formula (Formula 2), water content verification data is constructed; wherein, Formula 2 is: ； Where, r G2 For water content verification data, PWV is atmospheric precipitable water, LWC is liquid water content, SLW is supercooled water content, and k is... Φ ρ is the cloud phase correction coefficient. w For the pre-set density of liquid water, ρ air The cloud density is a pre-set value, and CWC is the cloud water content.

4. The aircraft-based artificial rain enhancement method based on multi-source observation data according to claim 2, characterized in that, Based on cloud optical thickness, cloud droplet effective radius, cloud phase, liquid water content, and cloud water content, particle optical verification data is constructed, including: Based on cloud phase state, cloud phase state correction coefficients are determined, and particle optical verification data are constructed based on cloud optical thickness, cloud droplet effective radius, liquid water content, cloud water content, and a pre-set formula (Formula 3). Formula 3 is: ； Where, r G3 For particle optical calibration data, LWC represents liquid water content, SLW represents supercooled water content, and k... Φ ρ is the cloud phase correction coefficient. w R is the pre-set density of liquid water. eff Let τ be the effective radius of the cloud droplet, and τ be the optical thickness of the cloud.

5. The aircraft-based artificial rain enhancement method based on multi-source observation data according to claim 1, characterized in that, The cross-source verification dataset includes various verification types; Based on the cross-source verification dataset within the target region, a global observation covariance matrix corresponding to the multi-source observation dataset is constructed, including: Based on the pre-set verification type of each dimension monitoring vector in the multi-source observation dataset, the cross-source verification data corresponding to each dimension monitoring vector is selected from the cross-source verification dataset. Based on the multi-source observation dataset, the cross-source verification data corresponding to each dimension of the monitoring vector in the multi-source observation dataset, and the pre-set Formula 4, the observation standard deviation corresponding to each dimension of the monitoring vector in the multi-source observation dataset is constructed; Formula 4 is: ； Where i is the index of the monitoring vector in the multi-source observation dataset, Let $\mathbf{i}$ be the standard deviation of the observations corresponding to the monitoring vector at index $i$ in the multi-source observation dataset. The device nominal standard deviation r is pre-set for the monitoring vector of index i in the multi-source observation dataset. i This refers to the cross-source verification data corresponding to the monitoring vector in each dimension of the multi-source observation dataset. The theoretical standard deviation of the residuals is pre-set for the monitoring vector of index i in the multi-source observation dataset; Based on the observation standard deviations corresponding to all dimensions of the monitoring vectors in the multi-source observation dataset, a global observation covariance matrix corresponding to the multi-source observation dataset is constructed.

6. The aircraft-based artificial rain enhancement method based on multi-source observation data according to claim 5, characterized in that, Based on the observation standard deviations corresponding to all dimensions of the monitoring vectors in the multi-source observation dataset, a global observation covariance matrix corresponding to the multi-source observation dataset is constructed, including: Based on the observation standard deviations corresponding to all dimensions of the monitoring vectors in the multi-source observation dataset, and the pre-set error correlation coefficients between every two dimensions of the monitoring vectors in the multi-source observation dataset, the covariance between every two dimensions of the monitoring vectors in the multi-source observation dataset is constructed. Based on the observation standard deviation corresponding to each dimension monitoring vector in the multi-source observation dataset, construct the observation variance corresponding to each dimension monitoring vector in the multi-source observation dataset; Based on the observation variance corresponding to each dimension monitoring vector in the multi-source observation dataset, and the covariance between every two dimensions monitoring vectors in the multi-source observation dataset, a global observation covariance matrix corresponding to the multi-source observation dataset is constructed.

7. The aircraft-based artificial rain enhancement method based on multi-source observation data according to claim 1, characterized in that, The objective function is: ； Where i is the index of the monitoring vector in the multi-source observation dataset, min J(X i Let X be the optimization objective of the monitoring vector at index i in the multi-source observation dataset. i X is the inversion data of the monitoring vector at index i in the multi-source observation dataset. i,obs Let be the monitoring vector of index i in the multi-source observation dataset, R be the global observation covariance matrix, and T be the transpose.

8. The aircraft-based artificial rain enhancement method based on multi-source observation data according to claim 1, characterized in that, Based on the multi-source inversion dataset and objective function, a multi-source inversion error dataset is constructed, including: Based on the multi-source observation dataset, the multi-source inversion dataset, and the objective function, the Hessian matrix corresponding to each inversion data in the multi-source inversion dataset is constructed. Based on the Hessian matrix corresponding to all inversion data in the multi-source inversion dataset, a multi-source inversion error dataset is constructed. The inversion error corresponding to any monitoring vector is the square root of the diagonal element in the corresponding Hessian matrix.

9. The aircraft-based artificial rain enhancement method based on multi-source observation data according to claim 1, characterized in that, Based on multi-source inversion error datasets and multi-source observation datasets, a rain enhancement operation potential index for the target area is constructed, including: Based on a multi-source observation dataset, multiple job-sensitive derived variables are constructed; each job-sensitive derived variable is calculated from at least two different observation vectors. Using a multi-source inversion error dataset, determine the uncertainty of the derived quantity corresponding to the sensitive derived quantity for each operation; Based on each job-sensitive derivative quantity and the uncertainty of the derivative quantity corresponding to each job-sensitive derivative quantity, an error modulation confidence factor for each job-sensitive derivative quantity is constructed; Based on all operation-sensitive derivatives and the error modulation confidence factor of each operation-sensitive derivative, a rain enhancement operation potential index for the target area is constructed.

10. An aircraft-based artificial rain enhancement system based on multi-source observation data, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the aircraft-based artificial rain enhancement method based on multi-source observation data as described in any one of claims 1-9.

Images