Diagnostic method and system for marine rainstorm based on multi-source remote sensing south china sea summer monsoon index

Abstract

The application provides a marine rainstorm diagnosis method and system based on a multi-source remote sensing South China Sea summer monsoon index, relates to the meteorological monitoring technical field, and comprises the following steps: acquiring multi-source remote sensing data such as a South China Sea wind field, a temperature field and a humidity field, calculating a monsoon comprehensive index field, and performing time-space gradient analysis to identify an abnormal region. Based on the spatial position and evolution trend of the abnormal region, the intensity of kinetic energy input and thermal energy output is calculated respectively, and whether the energy budget is in a positive feedback enhancement state is determined by comparison. Finally, the rainstorm diagnosis result is generated by coupling the evolution path of the index field and the energy state, and more accurate and timely diagnosis and early warning of the South China Sea marine rainstorm are realized.

Description

Technical Field

[0001] This invention relates to the field of meteorological monitoring technology, and in particular to a method and system for diagnosing marine rainstorms based on the multi-source remote sensing South China Sea summer monsoon index. Background Technology

[0002] In the field of marine meteorological monitoring, particularly for the diagnosis and early warning of maritime torrential rains triggered by the South China Sea summer monsoon, existing technologies typically rely on remote sensing data from single or limited sources. The conventional approach involves independently analyzing the spatiotemporal variation characteristics of individual meteorological elements such as wind fields, temperature fields, or humidity fields. For example, analyzing vertical wind shear or low-level vorticity can identify potential convection triggering conditions, or monitoring sea surface temperature anomalies or atmospheric precipitable water can assess torrential rain potential. These methods often focus on describing individual physical processes and determining thresholds for isolated indicators, attempting to comprehensively assess the likelihood of torrential rain by linearly superimposing the results of multiple independent indicators. Such methods form the technological foundation widely used in current operational forecasting and scientific research analysis.

[0003] Due to the lack of effective characterization of nonlinear coupling effects among multiple factors, diagnostic methods based on single factors or simple linear combinations struggle to accurately capture the complex feedback processes of energy and material exchange within monsoon systems. This leads to inaccurate identification of key dynamic and thermodynamic coupling mechanisms that trigger and sustain rainstorms, limiting forecast accuracy and timeliness. Existing methods typically separate the analysis of dynamic and thermodynamic processes in time and space, failing to achieve integrated dynamic diagnosis from the perspective of system energy balance. They cannot effectively determine the net energy balance and its positive feedback enhancement effect on the development of rainstorm systems. Consequently, the diagnosis of rainstorm intensity, location, and evolution trend involves significant uncertainty, making it difficult to meet the needs of high-precision early warning. Summary of the Invention

[0004] This invention provides a method and system for diagnosing marine rainstorms based on the South China Sea summer monsoon index using multi-source remote sensing, which can solve the problems in the prior art.

[0005] A first aspect of this invention provides a method for diagnosing marine heavy rainfall based on a multi-source remote sensing South China Sea summer monsoon index, comprising: Acquire multi-source remote sensing observation data of the South China Sea, including wind field data, temperature field data, and humidity field data; Vertical layered integration is performed on the wind field data, and multi-element coupling is performed by combining the temperature field data and the humidity field data to obtain the monsoon comprehensive index field distribution and perform spatiotemporal gradient analysis. After identifying abnormal change areas that meet the preset gradient threshold conditions, the corresponding spatial location information and evolution trend information are extracted. Based on the spatial location information, the spatial integration range of energy calculation is determined, and the horizontal kinetic energy flux convergence and vertical kinetic energy transport are calculated based on the wind field data. Based on the evolution trend information, the time series direction of energy budget analysis is determined, and the latent heat release rate and sensible heat accumulation rate are calculated based on the temperature field data and humidity field data. By comparing the combined thermal driving intensity of the latent heat energy release rate and the sensible heat energy accumulation rate with the combined power input intensity of the horizontal kinetic energy flux convergence and the vertical kinetic energy transport, the net energy conversion efficiency is calculated. When the net energy conversion efficiency is greater than zero, the energy conversion is determined to be in a positive feedback enhancement state. The spatiotemporal evolution path of the monsoon composite index field distribution along the time series direction within the spatial integration range is coupled with the positive feedback enhancement state of energy balance to generate rainstorm diagnosis results for the South China Sea.

[0006] Vertical stratified integration is performed on the wind field data, and multi-factor coupling is performed by combining the temperature field data and the humidity field data to obtain the monsoon comprehensive index field distribution and perform spatiotemporal gradient analysis, including: The wind field data is vertically divided according to pressure levels. The horizontal divergence field and vorticity field of the wind field vector in each pressure level are calculated and integrated in the vertical direction to obtain the vertical integral of the wind field characterizing the vertical circulation structure. The temperature field data is extracted vertically according to the pressure level, and the equivalent potential temperature field of each pressure level is calculated by combining the humidity field data. Based on the distribution characteristics of the equivalent potential temperature field in the vertical direction, thermodynamic structural parameters characterizing the thermal stability of the atmosphere are constructed. Based on the ratio of the spatial distribution variance of the vertical integral of the wind field to the spatial distribution variance of the thermal structure parameters, the weighting coefficients are dynamically allocated, and the vertical integral of the wind field and the thermal structure parameters are weighted and coupled to obtain the monsoon comprehensive index field distribution that characterizes the three-dimensional structural features of the monsoon system. The temporal derivative field is calculated along the time dimension and the spatial gradient field is calculated along the spatial dimension for the distribution of the monsoon composite index field. By extracting the joint distribution pattern of the temporal derivative field and the spatial gradient field and identifying the regions in the joint distribution pattern where the values ​​of the temporal derivative field and the spatial gradient field increase synchronously, the abnormal change regions that meet the preset gradient threshold conditions are obtained.

[0007] After identifying abnormal change regions that meet the preset gradient threshold conditions, the corresponding spatial location information and evolution trend information are extracted, including: Extract the geographic coordinate boundary range and center coordinates of the abnormal change area as the spatial location information; A multi-moment snapshot sequence is constructed along the time dimension for the distribution of the monsoon composite index field within the abnormal change region. The current moment snapshot in the multi-moment snapshot sequence is used as the reference snapshot and the next moment snapshot is used as the target snapshot. A sliding window is set within the spatial range of the reference snapshot and sliding matching is performed within the spatial range of the target snapshot. For each sliding position, the spatial correlation coefficient of the monsoon composite index field values ​​of the reference snapshot and the target snapshot within the coverage area of ​​the sliding window is calculated. All spatial correlation coefficients corresponding to the sliding positions are distributed and mapped according to the spatial coordinates of the sliding positions to form a spatial correlation coefficient field. The maximum position trajectory of the spatial correlation coefficient field is extracted, and the movement direction vector and movement speed scalar of the maximum position trajectory are combined as the evolution trend information.

[0008] Based on the spatial location information, the spatial integration range for energy calculation is determined, and the horizontal kinetic energy flux convergence and vertical kinetic energy transport are calculated based on the wind field data, including: The horizontal integration boundary is determined based on the geographical coordinate boundary range, and the vertical integration boundary is determined based on the pressure level distribution of the wind field data. The horizontal integration boundary and the vertical integration boundary are combined to form the spatial integration range for energy calculation. The horizontal wind speed vector field of each pressure level is extracted from the wind field data within the spatial integration range. The kinetic energy density distribution of the horizontal wind speed vector field is calculated, and the horizontal divergence field of kinetic energy flux is calculated along the horizontal direction. The horizontal divergence field of kinetic energy flux is then integrated within the horizontal integration boundary and layered within the vertical integration boundary to obtain the horizontal kinetic energy flux convergence. Vertical wind speed components for each pressure level are extracted from the wind field data within the spatial integration range. A vertical kinetic energy term is constructed by calculating the advection transport effect of the vertical wind speed components on the horizontal wind speed vector field. The vertical kinetic energy term is then used to perform gradient calculations for each pressure level along the vertical direction to form a vertical kinetic energy transport gradient field. The vertical kinetic energy transport gradient field is then subjected to surface integral operation within the horizontal integration boundary and integral operation within the vertical integration boundary to obtain the horizontal kinetic energy flux convergence, which characterizes the intensity of kinetic energy input.

[0009] Based on the evolution trend information, the time series direction of energy balance analysis is determined, and the latent heat release rate and sensible heat accumulation rate are calculated based on the temperature field data and the humidity field data, including: The spatial movement path of the abnormal change region is determined based on the movement direction vector, and the arrival time sequence of the abnormal change region along the spatial movement path is calculated based on the movement speed scalar. The time sequence of the arrival time sequence is used as the time series direction of energy balance analysis. Temperature and humidity profiles at each arrival time are extracted from the temperature and humidity data within the abnormal change region according to the time series direction. The saturation difference distribution between the saturated vapor pressure and the actual vapor pressure at each pressure level is calculated using the temperature and humidity profiles. The amount of water vapor condensation is calculated based on the saturation difference distribution, and the latent heat release per unit volume at each pressure level is calculated by combining the latent heat coefficient of water vapor condensation. The time derivative of the latent heat release per unit volume is calculated along the time series direction to obtain the latent heat release rate. The temperature field distribution at each arrival time is extracted from the temperature field data within the abnormal change region according to the time series direction. The product of the temperature field distribution and the air pressure field distribution is calculated as the sensible thermal energy density distribution. The sensible thermal energy density distribution is subjected to volume integral operation within the spatial integration range to obtain the total sensible thermal energy. The time derivative of the total sensible thermal energy is calculated along the time series direction to obtain the sensible thermal energy accumulation rate.

[0010] By comparing the combined thermodynamic driving intensity of the latent heat energy release rate and the sensible heat energy accumulation rate with the combined dynamic input intensity of the horizontal kinetic energy flux convergence and the vertical kinetic energy transport, the net energy conversion efficiency is calculated. When the net energy conversion efficiency is greater than zero, the energy conversion is determined to be in a state of positive feedback enhancement, including: The horizontal kinetic energy flux convergence is calculated by multiplying the horizontal kinetic energy flux convergence of each pressure level within the spatial integration range with the layer thickness weighting factor of the corresponding pressure level as a weighted horizontal kinetic energy input component. The weighted horizontal kinetic energy input component of all pressure levels is vertically integrated to obtain the total horizontal energy input and is vector-combined with the vertical kinetic energy transport. The magnitude of the composite vector is calculated as the combined input intensity. The horizontal kinetic energy flux convergence is calculated by multiplying the horizontal kinetic energy flux convergence of each pressure level within the spatial integration range with the layer thickness weighting factor of the corresponding pressure level as a weighted horizontal kinetic energy input component. The weighted horizontal kinetic energy input component of all pressure levels is vertically integrated to obtain the total horizontal kinetic energy input and is vector-combined with the vertical kinetic energy transport. The magnitude of the composite vector is calculated as the combined dynamic input intensity. The difference between the combined thermodynamic driving intensity and the combined dynamic input intensity is calculated as the net energy conversion efficiency. When the net energy conversion efficiency is greater than zero, the energy conversion state is marked as a positive feedback enhancement state.

[0011] The spatiotemporal evolution path of the monsoon composite index field distribution along the time series direction within the spatial integration range is coupled with the positive feedback enhancement state of energy balance to generate the following rainstorm diagnosis results for the South China Sea: Extract the spatial distribution pattern of the monsoon composite index field at each arrival time within the spatial integration range, calculate the centroid position offset vector of the monsoon composite index field distribution between adjacent arrival times based on the spatial distribution pattern, and connect the centroid position offset vectors according to the time series direction to form a spatiotemporal evolution path. The path direction vector is calculated for the spatiotemporal evolution path, and the movement direction vector in the evolution trend information is extracted as a reference direction vector. The cosine value of the angle between the path direction vector and the reference direction vector is calculated. When the cosine value of the angle is greater than a preset consistency threshold, it is determined that the evolution direction of the spatiotemporal evolution path is consistent with the direction represented by the evolution trend information. The consistency determination result of the spatiotemporal evolution path is logically ANDed with the positive feedback enhancement state of energy balance. When the consistency determination result of the evolution direction is true and the positive feedback enhancement state of energy balance is true, a rainstorm diagnosis result containing the geographical location information of the abnormal change area, the path trajectory information of the spatiotemporal evolution path, and the net energy balance is constructed.

[0012] A second aspect of the present invention provides a marine rainstorm diagnostic system based on a multi-source remote sensing South China Sea summer monsoon index, comprising: The data acquisition unit is used to acquire multi-source remote sensing observation data of the South China Sea, including wind field data, temperature field data, and humidity field data. The index analysis unit is used to perform vertical layered integral calculation on the wind field data and combine the temperature field data and humidity field data to perform multi-factor coupling to obtain the monsoon comprehensive index field distribution and perform spatiotemporal gradient analysis. After identifying abnormal change areas that meet the preset gradient threshold conditions, the corresponding spatial location information and evolution trend information are extracted. The energy calculation unit is used to determine the spatial integration range of energy calculation based on the spatial location information and to calculate the horizontal kinetic energy flux convergence and vertical kinetic energy transport based on the wind field data. It also determines the time series direction of energy budget analysis based on the evolution trend information and calculates the latent heat energy release rate and sensible heat energy accumulation rate based on the temperature field data and humidity field data. The income and expenditure determination unit is used to compare the combined thermal driving intensity of the latent heat energy release rate and the sensible heat energy accumulation rate with the combined power input intensity of the horizontal kinetic energy flux convergence and the vertical kinetic energy transport, calculate the net energy conversion efficiency, and determine that the energy conversion is in a positive feedback enhancement state when the net energy conversion efficiency is greater than zero. The diagnostic generation unit is used to couple the spatiotemporal evolution path of the monsoon composite index field distribution along the time series direction within the spatial integration range with the positive feedback enhancement state of energy balance to generate the rainstorm diagnostic results in the South China Sea.

[0013] A third aspect of the embodiments of the present invention, An electronic device is provided, comprising: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0014] Fourth aspect of the embodiments of the present invention, A computer-readable storage medium is provided, having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0015] The beneficial effects of this application are as follows: This method enables refined diagnosis of marine heavy rainfall processes under the background of the South China Sea summer monsoon. By fusing multi-source remote sensing data, a comprehensive monsoon index field containing wind, temperature, and humidity information is constructed. This index field can more comprehensively characterize the three-dimensional thermodynamic structure of the monsoon system, overcoming the shortcomings of single indices in representing complex circulation systems, and significantly improving the accuracy of monitoring monsoon intensity and evolution characteristics.

[0016] Spatiotemporal gradient analysis of the monsoon composite index field can automatically and objectively identify anomalous change areas in monsoon activity and accurately extract their spatial location and evolution trend. This process effectively filters out stable signals in the background field, directly focusing on key disturbance systems that may trigger rainstorms, providing clear target areas and time clues for subsequent targeted energy analysis, and significantly improving the directionality and efficiency of diagnosis.

[0017] Based on the spatial location and evolution trend of the anomalous region, the energy calculation range was dynamically delineated and the analysis time sequence was determined, ensuring that the energy budget analysis and the evolution of the weather system were strictly synchronized in time and space. By calculating the horizontal kinetic energy flux convergence, vertical kinetic energy transport, latent heat energy release, and sensible heat energy accumulation respectively, the energy input and output processes of the rainstorm system were quantified and analyzed in detail, revealing the details of energy conversion and distribution in different dimensions and physical processes.

[0018] By comparing the combined energy input intensity and combined output intensity, it is possible to scientifically determine whether the system's energy balance is in a state of positive feedback enhancement. This criterion directly reveals from an energy perspective whether a rainstorm system possesses the intrinsic power for self-sustaining and continuous development, providing crucial physical evidence for judging the intensity changes and duration of rainstorms. This allows the diagnostic conclusions to transcend mere phenomenological descriptions and possess solid dynamic support.

[0019] By coupling the spatiotemporal evolution path of the monsoon index field with the positive feedback state of energy balance, this method achieves an organic combination of large-scale monsoon circulation background and mesoscale rainstorm system energy processes. This approach can not only diagnose whether rainstorms will occur, but also reveal the specific physical mechanisms and evolution paths of their occurrence and development. The generated diagnostic results have clear physical significance and higher forecasting and early warning reference value. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating the marine rainstorm diagnosis method based on the multi-source remote sensing South China Sea summer monsoon index, as an embodiment of the present invention. Figure 2 A schematic diagram illustrating the process of constructing a comprehensive monsoon index field and identifying areas of abnormal change. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] The technical solution of the present invention will be described in detail below with reference to specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0023] Figure 1 This is a flowchart illustrating the marine rainstorm diagnosis method based on the multi-source remote sensing South China Sea summer monsoon index, as described in an embodiment of the present invention. Figure 1 As shown, the method includes: Acquire multi-source remote sensing observation data of the South China Sea, including wind field data, temperature field data, and humidity field data; Vertical layered integration is performed on the wind field data, and multi-element coupling is performed by combining the temperature field data and the humidity field data to obtain the monsoon comprehensive index field distribution and perform spatiotemporal gradient analysis. After identifying abnormal change areas that meet the preset gradient threshold conditions, the corresponding spatial location information and evolution trend information are extracted. Based on the spatial location information, the spatial integration range of energy calculation is determined, and the horizontal kinetic energy flux convergence and vertical kinetic energy transport are calculated based on the wind field data. Based on the evolution trend information, the time series direction of energy budget analysis is determined, and the latent heat release rate and sensible heat accumulation rate are calculated based on the temperature field data and humidity field data. By comparing the combined thermal driving intensity of the latent heat energy release rate and the sensible heat energy accumulation rate with the combined power input intensity of the horizontal kinetic energy flux convergence and the vertical kinetic energy transport, the net energy conversion efficiency is calculated. When the net energy conversion efficiency is greater than zero, the energy conversion is determined to be in a positive feedback enhancement state. The spatiotemporal evolution path of the monsoon composite index field distribution along the time series direction within the spatial integration range is coupled with the positive feedback enhancement state of energy balance to generate rainstorm diagnosis results for the South China Sea.

[0024] In one optional implementation, the wind field data is subjected to vertical stratified integration and combined with the temperature field data and humidity field data for multi-factor coupling to obtain the monsoon comprehensive index field distribution and perform spatiotemporal gradient analysis, including: The wind field data is vertically divided according to pressure levels. The horizontal divergence field and vorticity field of the wind field vector in each pressure level are calculated and integrated in the vertical direction to obtain the vertical integral of the wind field characterizing the vertical circulation structure. The temperature field data is extracted vertically according to the pressure level, and the equivalent potential temperature field of each pressure level is calculated by combining the humidity field data. Based on the distribution characteristics of the equivalent potential temperature field in the vertical direction, thermodynamic structural parameters characterizing the thermal stability of the atmosphere are constructed. Based on the ratio of the spatial distribution variance of the vertical integral of the wind field to the spatial distribution variance of the thermal structure parameters, the weighting coefficients are dynamically allocated, and the vertical integral of the wind field and the thermal structure parameters are weighted and coupled to obtain the monsoon comprehensive index field distribution that characterizes the three-dimensional structural features of the monsoon system. The temporal derivative field is calculated along the time dimension and the spatial gradient field is calculated along the spatial dimension for the distribution of the monsoon composite index field. By extracting the joint distribution pattern of the temporal derivative field and the spatial gradient field and identifying the regions in the joint distribution pattern where the values ​​of the temporal derivative field and the spatial gradient field increase synchronously, the abnormal change regions that meet the preset gradient threshold conditions are obtained.

[0025] Wind field data is stratified according to atmospheric pressure levels, typically selecting multiple isobaric levels from near-surface to the tropopause, such as typical pressure layers like 925 hPa, 850 hPa, 700 hPa, 500 hPa, and 300 hPa, to ensure complete vertical coverage of the wind field structure across the entire troposphere. Within each pressure layer, the horizontal divergence and vorticity fields of the wind field vectors are calculated. The horizontal divergence field reflects the convergence and divergence characteristics of the airflow, while the vorticity field characterizes the rotational properties of the airflow; together, they describe the dynamic characteristics of each level. Subsequently, the divergence and vorticity fields of each pressure layer are integrated vertically, with the integration range covering all selected pressure levels, yielding the vertical integral of the wind field characterizing the vertical circulation structure. . By integrating the cumulative effects of airflow convergence, divergence, and rotation at different altitudes throughout the troposphere, it can more comprehensively reflect the three-dimensional dynamic characteristics of the South China Sea monsoon system and avoid the problem of insufficient representativeness caused by single-layer analysis.

[0026] Temperature field data were extracted vertically along the same pressure levels, and the equivalent potential temperature field was calculated on each pressure level by combining it with humidity field data. Equivalent potential temperature is a comprehensive thermodynamic variable that simultaneously incorporates temperature and latent heat of water vapor. Its calculation process converts the latent heat energy carried by the humidity zone into an equivalent temperature contribution, thereby making... It can simultaneously reflect the dry and wet thermodynamic states of the atmosphere. Based on each pressure layer. The vertical distribution characteristics are used to construct thermodynamic structural parameters characterizing atmospheric thermal stability. . Through analysis The stability of atmospheric stratification can be determined by the gradient of atmospheric stratification with altitude: when As altitude increases, the atmospheric stratification remains stable; when As altitude decreases, the atmospheric stratification is in a convective unstable state, which is conducive to the development of severe convective weather. The construction of this system enables thermal factors to participate in subsequent multi-factor coupling analysis in a quantitative form.

[0027] Integrating the vertical wind field With thermodynamic structural parameters Before weighted coupling, the weighting coefficients need to be dynamically determined based on the spatial distribution variances of both components. Specifically, the calculation... Spatial distribution variance within the study area as well as Spatial distribution variance The ratio between the two is used as the basis for weight allocation. (Dynamic weight coefficient) With thermal weighting coefficient satisfy , This ensures that the sum of the two weights is always 1, and that elements with larger variances receive higher weight contributions. This dynamic weight allocation method based on variance ratios can adaptively reflect the relative importance of dynamic and thermal factors in different weather processes: when the dynamic signal of the monsoon system is strong, Larger, dynamic weight The corresponding increase; when the thermal instability signal is more prominent, Larger, thermal weight The corresponding increase. The distribution of the monsoon composite index field obtained after weighted coupling. Expressed as , The formation of a continuous scalar field in space, with its high-value regions corresponding to areas where the dynamics and thermal forces of the monsoon system are enhanced, is a key basis for subsequent identification of potential areas for heavy rain.

[0028] get After determining the spatial field distribution, spatiotemporal gradient analysis is performed on it. This is done along the time dimension. Calculate the time derivative field This derivative field describes the rate of change of the monsoon composite index over time; a positive value indicates that the monsoon system is in a strengthening evolution phase at that location, while a negative value indicates that the system is weakening. Calculated along the spatial dimension... Spatial gradient field The magnitude of the spatial gradient field reflects the intensity of horizontal spatial variation in the monsoon composite index. High gradient regions typically correspond to monsoon fronts or the boundary zones of strong convective systems. After independently calculating the time derivative field and the spatial gradient field, their joint distribution pattern is extracted, focusing on identifying regions where the values ​​of the time derivative field and the spatial gradient field increase synchronously. Synchronous increase implies that the monsoon composite index in this region not only exhibits strong horizontal differences in space but also shows a rapid strengthening trend in time. This spatiotemporal co-enhancement is a typical precursor signal of the rapid development of strong convective systems. The condition where the time derivative field value exceeds a preset threshold is considered a prerequisite. Furthermore, the magnitude of the spatial gradient field exceeds the preset threshold of the spatial gradient. The regions identified as anomalous change areas are calibrated using two preset thresholds based on statistical analysis of historical rainstorm samples to ensure statistical significance. After identifying these anomalous change areas, their corresponding spatial location information (including latitude and longitude coordinates) and evolution trend information (including the magnitude and sign of the time derivative field) are extracted. This provides a basis for determining the spatial integration range and time series direction in subsequent energy budget analysis.

[0029] The aforementioned complete process of vertical stratified integration, equivalent potential temperature calculation, dynamic weight coupling, and spatiotemporal gradient joint analysis constitutes a technical chain from multi-source remote sensing observation data to the distribution of the monsoon composite index field and then to the identification of anomalous change areas. Each step is specifically designed for the physical characteristics of the South China Sea monsoon system: vertical integration covers the entire troposphere to capture deep convection characteristics; equivalent potential temperature incorporates latent heat of water vapor to reflect the thermodynamic effect; dynamic weight allocation adapts to the switching of the dominant dynamic and thermodynamic factors under different weather backgrounds; and spatiotemporal joint gradient analysis simultaneously captures the synergistic signals of spatial structural anomalies and temporal evolution anomalies, thereby significantly improving the diagnostic capability of the monsoon composite index field in characterizing the precursory features of rainstorms.

[0030] In one optional implementation, after identifying abnormal change regions that meet preset gradient threshold conditions, extracting the corresponding spatial location information and evolution trend information includes: Extract the geographic coordinate boundary range and center coordinates of the abnormal change area as the spatial location information; A multi-moment snapshot sequence is constructed along the time dimension for the distribution of the monsoon composite index field within the abnormal change region. The current moment snapshot in the multi-moment snapshot sequence is used as the reference snapshot and the next moment snapshot is used as the target snapshot. A sliding window is set within the spatial range of the reference snapshot and sliding matching is performed within the spatial range of the target snapshot. For each sliding position, the spatial correlation coefficient of the monsoon composite index field values ​​of the reference snapshot and the target snapshot within the coverage area of ​​the sliding window is calculated. All spatial correlation coefficients corresponding to the sliding positions are distributed and mapped according to the spatial coordinates of the sliding positions to form a spatial correlation coefficient field. The maximum position trajectory of the spatial correlation coefficient field is extracted, and the movement direction vector and movement speed scalar of the maximum position trajectory are combined as the evolution trend information.

[0031] After identifying anomalous change regions that meet the preset gradient threshold conditions, two types of key information need to be extracted from these regions: first, location information describing the spatial extent of the anomalous region; and second, trend information describing the evolution of this region over time. These two types of information together constitute the basic input for subsequent energy budget analysis and rainstorm diagnosis.

[0032] To extract spatial location information, the outer contour of the anomalous change area is traced to obtain its geographic coordinate boundary range, including the minimum and maximum values ​​in the longitude and latitude directions, thus determining a rectangular geographic framework that completely encloses the anomalous area. Based on this, the latitude and longitude weighted average of all grid points within this boundary range that meet the threshold conditions is calculated, using the value of the monsoon composite index field distribution as the weight, to obtain the center coordinates of the anomalous change area. This center coordinate represents the energy accumulation point of the anomalous signal in space, rather than a simple geometric center, and therefore has stronger physical significance for determining the subsequent spatial integration range. The geographic coordinate boundary range and the center coordinate together constitute complete spatial location information, used for subsequent determination of the spatial integration range for energy calculation.

[0033] Extracting evolution trend information relies on dynamically tracking the distribution of the monsoon composite index field over time. Within areas of anomalous change, the spatial distribution of the monsoon composite index field at multiple moments is sequentially extracted along the time dimension, forming a multi-moment snapshot sequence. Each moment corresponds to a two-dimensional spatial distribution image, called a snapshot. The snapshot at the current moment in the sequence is designated as the reference snapshot, and the snapshot at the next moment immediately following it is designated as the target snapshot. The time interval between the reference snapshot and the target snapshot should be consistent with the temporal resolution of the remote sensing data, typically the time step corresponding to hourly or 6-hourly satellite inversion data products.

[0034] A rectangular sliding window is set within the spatial range of the reference snapshot. The window size is determined based on the spatial scale of the anomalous change region, typically taking 1 / 3 to 1 / 2 of the spatial span along the minor axis of the anomalous region as the window side length to ensure that the window can capture local structural features without excessively smoothing spatial details. This sliding window is then slid grid-by-grid within the spatial range of the target snapshot at a preset step size. The step size can be 1 to 2 times the resolution of a single grid point to balance computational efficiency and matching accuracy. For each sliding position, the monsoon composite index field numerical sequence within the area covered by the sliding window in the reference snapshot and the corresponding monsoon composite index field numerical sequence within the area covered by the sliding position in the target snapshot are extracted, and the spatial correlation coefficient between the two sets of numerical sequences is calculated.

[0035] The spatial correlation coefficient is calculated as follows: Assume there are a total of [number] instances within the reference snapshot window. The grid point, the first The monsoon composite index value for each grid point in the reference snapshot is... The value within the area covered by the sliding position corresponding to the target snapshot is Then the spatial correlation coefficient for: ,in To reference the average monsoon composite index values ​​of all grid points within the snapshot window, The mean of the monsoon composite index values ​​for all grid points within the area covered by the sliding position corresponding to the target snapshot. This represents the total number of grid points covered by the sliding window. This correlation coefficient reflects the similarity of spatial structural features in the reference snapshot to different locations in the target snapshot; a higher value indicates that the spatial distribution of the target snapshot at that sliding position is closer to that of the reference snapshot.

[0036] After traversing all valid sliding positions within the target snapshot's spatial range, the spatial correlation coefficient value corresponding to each sliding position is mapped according to the spatial coordinates (longitude and latitude) of that sliding position, forming a two-dimensional distribution map corresponding to the target snapshot's spatial range, i.e., the spatial correlation coefficient field. The spatial distribution of this coefficient field reflects the degree of matching between the monsoon composite index field structure in the reference snapshot and different regions of the target snapshot, with high-value regions indicating the most likely location of the original spatial structure in the next moment.

[0037] The location of the maximum value in the spatial correlation coefficient field is extracted, and the spatial coordinates of this location represent the optimal matching position of the spatial structure corresponding to the reference snapshot at the time of the target snapshot. This sliding matching process is repeated for each pair of adjacent snapshots in the multi-time snapshot sequence to sequentially obtain the maximum value locations at each time. These locations are then connected in chronological order to form the maximum value location trajectory. This trajectory spatially depicts the movement path of the monsoon composite index field anomaly structure, and its temporal continuity reflects the overall propagation characteristics of the anomaly system.

[0038] Based on the trajectory of the maximum value location, the latitude and longitude difference between the maximum value locations at adjacent times is calculated and converted into actual displacement distance in kilometers. This is then combined with the corresponding time interval to calculate the scalar velocity, in meters per second or kilometers per hour. Simultaneously, the direction of movement is determined based on the longitude and latitude components of the displacement. Using true north as a reference, the direction angle is measured clockwise to obtain the direction vector. The direction vector and the velocity scalar are combined to form the evolution trend information. The direction vector describes the propagation orientation of the anomalous system, while the velocity scalar describes its propagation speed. The combination of both can comprehensively characterize the dynamic evolution of the monsoon composite index field anomaly region, providing a quantitative basis for determining the time series direction in subsequent energy budget analysis.

[0039] In practical applications, when the low-level wind field convergence intensifies and the mid-tropospheric water vapor flux significantly increases during the active South China Sea summer monsoon, a rapidly strengthening anomalous center often appears in the central to northern South China Sea region of the monsoon composite index field. By constructing the aforementioned multi-moment snapshot sequence and using the sliding matching method, the movement path of this anomalous center can be accurately tracked, and its evolutionary trend towards the north-northeast can be identified, thus providing reliable kinematic constraint information for determining the spatiotemporal transfer of areas affected by heavy rainfall.

[0040] In one optional implementation, determining the spatial integration range for energy calculation based on the spatial location information and calculating the horizontal kinetic energy flux convergence and vertical kinetic energy transport based on the wind field data includes: The horizontal integration boundary is determined based on the geographical coordinate boundary range, and the vertical integration boundary is determined based on the pressure level distribution of the wind field data. The horizontal integration boundary and the vertical integration boundary are combined to form the spatial integration range for energy calculation. The horizontal wind speed vector field of each pressure level is extracted from the wind field data within the spatial integration range. The kinetic energy density distribution of the horizontal wind speed vector field is calculated, and the horizontal divergence field of kinetic energy flux is calculated along the horizontal direction. The horizontal divergence field of kinetic energy flux is then integrated within the horizontal integration boundary and layered within the vertical integration boundary to obtain the horizontal kinetic energy flux convergence. Vertical wind speed components for each pressure level are extracted from the wind field data within the spatial integration range. A vertical kinetic energy term is constructed by calculating the advection transport effect of the vertical wind speed components on the horizontal wind speed vector field. The vertical kinetic energy term is then used to perform gradient calculations for each pressure level along the vertical direction to form a vertical kinetic energy transport gradient field. The vertical kinetic energy transport gradient field is then subjected to surface integral operation within the horizontal integration boundary and integral operation within the vertical integration boundary to obtain the horizontal kinetic energy flux convergence, which characterizes the intensity of kinetic energy input.

[0041] Based on the geographic coordinate boundaries determined by spatial location information, the latitude and longitude coordinates of the boundary of the anomalous change area are extracted along the latitudinal and meridional directions respectively. These boundary coordinates are then mapped to the grid system of remote sensing observation data to obtain the horizontal integration boundary. In the vertical direction, based on the pressure level distribution contained in the wind field data, effective pressure levels are identified layer by layer from the near-surface layer (usually around 850 hPa) to the upper troposphere (usually around 200 hPa). The lowest and highest effective pressure levels are used as the upper and lower limits of the vertical integration boundary. Combining the horizontal and vertical integration boundaries constitutes a three-dimensional spatial integration range. This range covers both the horizontal distribution of the monsoon anomaly region and the main dynamically active layers within the troposphere, providing a complete integration domain for subsequent kinetic energy calculations.

[0042] After determining the spatial integration range, the horizontal wind speed vector field for each pressure level is extracted layer by layer from the wind field data. For each pressure level, the horizontal wind speed vector field consists of the zonal wind speed component. With meridional wind speed component Based on this composition, the kinetic energy density distribution of this level can be calculated. Kinetic energy density. The calculation expression is as follows ,in The wind speed is in the zonal direction. For meridional wind speed, The unit is After obtaining the kinetic energy density at each grid point, it is combined with the atmospheric density field of that pressure level. Construct a kinetic flux vector field, kinetic flux vector Defined as ,in Let be the horizontal wind speed vector. Calculate the divergence of the kinetic flux vector field in the horizontal plane to obtain the horizontal divergence field of the kinetic flux. A region with a negative divergence value indicates that kinetic energy converges and accumulates at that location, while a region with a positive divergence value indicates that kinetic energy dissipates outward.

[0043] The kinetic flux horizontal divergence field is integrally integrated over the region bounded by the horizontal integral boundary. The area integral uses a grid-point area weighting method, with each grid point having a specific area weight. The value is determined by the grid spacing of the latitude and longitude corresponding to the grid point and the radius of curvature of the Earth, and the specific expression is as follows: ,in The average radius of the Earth For the first The latitude of the grid points and These represent the meridional and latitudinal grid spacings (in radians), respectively. After performing surface integration for each pressure level, layer integration is then performed along the vertical direction in the pressure coordinate system for each layer. The layer integration uses the pressure difference. Weighted and divided by gravitational acceleration Converted to the dimension of energy per unit area, the horizontal kinetic energy flux convergence is finally obtained. . A negative value indicates net kinetic energy convergence in the region, meaning strong horizontal dynamic input, which is conducive to the maintenance and enhancement of the rainstorm system.

[0044] The calculation of vertical kinetic energy transport is based on the vertical wind speed components at each pressure level. Vertical wind speed components In the pressure coordinate system The unit is used to characterize the velocity of the air parcel in the vertical direction. Vertical translational kinetic energy term. It describes the transfer of horizontal kinetic energy by vertical motion, and is defined as follows: This is the product of the vertical wind speed component and the kinetic energy density along the vertical gradient of the pressure coordinate. Physically, it measures the ability of airflow to transfer kinetic energy from a pressure layer to an adjacent layer through vertical motion. (Corresponding to upward motion) and When kinetic energy increases with increasing pressure, meaning the kinetic energy of lower levels is stronger than that of higher levels, A negative value indicates that kinetic energy is transported from lower to higher levels, which is beneficial for the accumulation of kinetic energy in the upper troposphere.

[0045] The vertical translational kinetic energy term was calculated at each grid point of each pressure level. Then, along the vertical direction Calculate the pressure gradient to obtain the vertical kinetic energy transport gradient field. The gradient field spatially reflects the variation of vertical kinetic energy transport intensity with height. Positive gradient regions correspond to the kinetic energy transport divergence layer, while negative regions correspond to the kinetic energy transport convergence layer. A surface integral operation is performed on the vertical kinetic energy transport gradient field within the horizontal integration boundary. The surface integral method is consistent with the grid-point area weighting method used in the calculation of horizontal kinetic energy flux convergence, i.e., using... We then perform a weighted summation of the values ​​at each grid point, using the pressure difference as the weight. After completing the area integral, we integrate the area integral results for each pressure level within the vertical integration boundary, again using the pressure difference. Weighted and divided by gravitational acceleration Finally, the vertical kinetic energy transfer quantity, which characterizes the intensity of vertical kinetic energy input, is obtained. .

[0046] Horizontal kinetic flux convergence With vertical kinetic energy transfer Together they constitute a quantitative index of power input intensity, both expressed in the dimension of energy flux ( Expressed in units (or equivalent units), this facilitates comparison and analysis with thermal driving intensity within the same dimensional framework. In practical calculations, the grid resolution of wind field data directly affects the accuracy of kinetic energy density distribution and divergence field. It is recommended to use remote sensing fusion wind field products with a horizontal resolution of no less than 0.25°×0.25° and a vertical layering of no less than 10 standard pressure layers to ensure the representativeness of the integration results. For missing grid points, bilinear interpolation is used to fill them in, avoiding large integration errors due to local missing measurements. In addition, before performing area integration calculations, spatial smoothing filters are applied to the horizontal divergence field of kinetic flux and the vertical kinetic energy transport gradient field, respectively, to filter out high-frequency noise with a spatial scale smaller than 1° and retain mesoscale and larger signals that match the scale of the rainstorm system. This ensures that the final horizontal kinetic energy flux convergence and vertical kinetic energy transport can robustly characterize the dynamic input characteristics of the South China Sea rainstorm system.

[0047] In one optional implementation, determining the time series direction of energy budget analysis based on the evolution trend information and calculating the latent heat release rate and sensible heat accumulation rate based on the temperature field data and the humidity field data includes: The spatial movement path of the abnormal change region is determined based on the movement direction vector, and the arrival time sequence of the abnormal change region along the spatial movement path is calculated based on the movement speed scalar. The time sequence of the arrival time sequence is used as the time series direction of energy balance analysis. Temperature and humidity profiles at each arrival time are extracted from the temperature and humidity data within the abnormal change region according to the time series direction. The saturation difference distribution between the saturated vapor pressure and the actual vapor pressure at each pressure level is calculated using the temperature and humidity profiles. The amount of water vapor condensation is calculated based on the saturation difference distribution, and the latent heat release per unit volume at each pressure level is calculated by combining the latent heat coefficient of water vapor condensation. The time derivative of the latent heat release per unit volume is calculated along the time series direction to obtain the latent heat release rate. The temperature field distribution at each arrival time is extracted from the temperature field data within the abnormal change region according to the time series direction. The product of the temperature field distribution and the air pressure field distribution is calculated as the sensible thermal energy density distribution. The sensible thermal energy density distribution is subjected to volume integral operation within the spatial integration range to obtain the total sensible thermal energy. The time derivative of the total sensible thermal energy is calculated along the time series direction to obtain the sensible thermal energy accumulation rate.

[0048] The evolution trend information includes two key attributes: the direction vector of movement of the anomalous change region and the scalar of movement velocity. The direction vector is determined by the spatial centroid drift direction of the monsoon composite index field distribution over time, specifically obtained by vector synthesis of the differences in the geometric center coordinates of the anomalous change region at adjacent times. The scalar of movement velocity is calculated by dividing the spherical distance between the geometric centers of the anomalous change region at adjacent times by the corresponding time interval. Based on the direction vector of movement, the spatial movement path of the anomalous change region can be depicted in the latitude and longitude coordinate system of the South China Sea. This path starts from the geometric center of the anomalous change region at the initial time and extends sequentially along the direction vector of movement, forming a continuous trajectory curve.

[0049] Based on the movement speed scalar and the spatial movement path, the arrival times of the abnormal change region along the path to each spatial node are calculated, forming an arrival time sequence. Specifically, let the nth node on the path be... The spherical arc length between each node and the starting node is... The movement speed scalar is Then the first Arrival times corresponding to each node satisfy ,in The initial time is used as the reference point. Arranging the arrival time sequence in chronological order yields the time series direction for energy budget analysis. The physical significance of this approach is that tracing energy budget changes along the actual movement trajectory of the anomalous change area avoids spurious time change signals caused by system movement in fixed-site analysis, allowing the calculation of latent heat release rate and sensible heat accumulation rate to more accurately reflect the intrinsic thermodynamic evolution of the rainstorm system.

[0050] After determining the time series direction, temperature and humidity profiles are extracted from the temperature and humidity data within the regions of anomalous change, according to this time series direction. The temperature profile includes the vertical distribution of atmospheric temperature at each standard pressure level from the ground to the tropopause, while the humidity profile includes the vertical distribution of specific humidity or relative humidity at each corresponding pressure level. Based on the temperature profiles, the saturated vapor pressure at each pressure level is calculated using the Clausius-Clapeyron equation. This equation expresses saturated vapor pressure as an exponential function of temperature; based on the humidity profile and the current pressure value of the pressure layer, the actual vapor pressure is calculated. The saturation difference distribution is defined as follows: It reflects the degree of deviation of the atmosphere from the saturation state at each pressure level. The smaller the saturation difference, the closer the atmosphere is to saturation, and the easier it is for water vapor to condense.

[0051] Calculate water vapor condensation based on saturation difference distribution, assuming that when the saturation difference... When the temperature drops to zero or below, all water vapor exceeding the saturation point condenses into liquid water or solid ice crystals. (Water vapor condensation rate per unit volume) The concentration is obtained by converting the portion of the actual water vapor pressure exceeding the saturated water vapor pressure into a mass concentration. The specific conversion requires consideration of the current atmospheric pressure and temperature conditions, and must be combined with the latent heat of condensation. (Liquid water condensation corresponds to approximately) J / kg, ice crystal sublimation corresponds to approximately J / kg (in actual calculations, the phase state is determined based on the layer temperature), latent heat release per unit volume at each pressure level. Depend on Given, among which Atmospheric density (consistent with the definition in claim 4, and will not be given a new meaning here). Latent heat release per unit volume. Calculate the time derivative along the time series direction, that is, between adjacent arrival times. Dividing the difference by the corresponding time interval yields the latent heat release rate for each pressure level. The unit is W / m³.

[0052] The calculation of the sensible thermal energy accumulation rate is based on temperature field data. The temperature field distribution at each arrival time is extracted from the temperature field data within the region of anomalous change, following the time series direction. This temperature field distribution includes the temperature values ​​of all grid points within the region of anomalous change at each pressure level. The calculation of the sensible thermal energy density distribution uses the product of the temperature field distribution and the air pressure field distribution as the core quantity. The physical basis for this is that, under the dry adiabatic approximation, the atmospheric sensible thermal energy density is closely related to both temperature and air pressure, and the product of temperature and air pressure is proportional to the atmospheric internal energy density under certain approximation conditions. Let the... Line number Grid point at the The temperature of the pressure layer is The corresponding air pressure value is Then the sensible thermal energy density of that lattice point in that layer Represented as ,in The specific heat capacity of dry air at constant pressure. The gas constant for dry air. Using a reference pressure (usually 1000 hPa), this expression is dimensionally consistent with the sensible thermal energy per unit volume.

[0053] thermal energy density distribution The total sensible thermal energy is obtained by performing a volume integral operation within the spatial integration range. The volume integral operation is performed in three dimensions along the longitudinal, latitudinal, and vertical directions. The method for handling the grid area weights is consistent with the calculation of kinetic flux convergence in claim 4. The area of ​​each latitude grid point is cosine-weighted and corrected, while the vertical direction is adjusted according to the pressure layer spacing. With gravitational acceleration The ratio is converted into a mass weight. The total sensible thermal energy is expressed as:

[0054] Total sensible thermal energy Calculate the time derivative along the time series direction, i.e., for adjacent arrival times. and Between Difference divided by time interval The rate of accumulation of sensible thermal energy is obtained. The unit is W. When When this occurs, it indicates that the sensible thermal energy in the abnormal change region increases over time, and the atmospheric thermodynamic state tends to be unstable, which is conducive to convection triggering; when When this occurs, it indicates that the sensible thermal energy is dissipated and the thermal driving force is weakened.

[0055] Latent heat release rate With the rate of accumulation of sensible heat energy Together, these two factors constitute the combined thermodynamic driving force, and they are physically interconnected: latent heat release heats the atmosphere, leading to an increase in sensible thermal energy; the accumulation of sensible thermal energy destabilizes the atmospheric stratification, further promoting water vapor uplift and condensation, releasing even more latent heat. This positive feedback coupling relationship is the core thermodynamic mechanism for the maintenance and intensification of rainstorm systems under the South China Sea summer monsoon background. By calculating these two rates using a Lagrange tracking framework along the moving trajectory, the dynamic evolution characteristics of the rainstorm system's thermodynamic energy during its movement can be accurately captured, providing a reliable thermodynamic input for subsequent determination of net energy conversion efficiency.

[0056] In one optional implementation, the net energy conversion efficiency is calculated by comparing the combined thermodynamic driving intensity of the latent heat energy release rate and the sensible heat energy accumulation rate with the combined dynamic input intensity of the horizontal kinetic energy flux convergence and the vertical kinetic energy transport. When the net energy conversion efficiency is greater than zero, the energy conversion is determined to be in a state of positive feedback enhancement, including: The latent heat release rate and the sensible heat accumulation rate are integrated over time along the time series direction to obtain the cumulative latent heat release amount and the cumulative sensible heat increment. The ratio of the cumulative latent heat release amount to the cumulative sensible heat increment is calculated as the thermal driving structure coefficient. Based on the thermal driving structure coefficient, the latent heat release rate and the sensible heat accumulation rate are nonlinearly coupled and calculated to obtain the combined thermal driving intensity. The horizontal kinetic energy flux convergence is calculated by multiplying the horizontal kinetic energy flux convergence of each pressure level within the spatial integration range with the layer thickness weighting factor of the corresponding pressure level as a weighted horizontal kinetic energy input component. The weighted horizontal kinetic energy input component of all pressure levels is vertically integrated to obtain the total horizontal kinetic energy input and is vector-combined with the vertical kinetic energy transport. The magnitude of the composite vector is calculated as the combined dynamic input intensity. The difference between the combined thermodynamic driving intensity and the combined dynamic input intensity is calculated as the net energy conversion efficiency. When the net energy conversion efficiency is greater than zero, the energy conversion state is marked as a positive feedback enhancement state.

[0057] After independently calculating the latent heat release rate and the sensible heat accumulation rate, they need to be integrated into a unified thermodynamic driving intensity index for quantitative comparison with the power input intensity. Regarding the latent heat release rate... Along the time series direction during the integration period The cumulative latent heat release is obtained by integrating over time. ,Right now Similarly, the rate of accumulation of sensible thermal energy... By integrating over the same period of time, the cumulative sensible heat increment is obtained. ,Right now The cumulative latent heat release reflects the total amount of heat released into the atmosphere by water vapor condensation over the entire time series; the cumulative sensible heat increment reflects the total amount of heat energy accumulated by the atmospheric column through sensible means during that period. The ratio of the two is defined as the thermally driven structure coefficient. ,Right now The physical meaning of the thermodynamically driven structure coefficient lies in: when When the latent heat is relatively large, it indicates that latent heat release dominates the thermal driving force, and convective condensation heating is the core energy source driving atmospheric motion; when When the value approaches 1, the contributions of latent heat and sensible heat are roughly equal; when When the temperature is relatively low, sensible heating is more significant, and atmospheric stratification is mainly based on stable thermal accumulation.

[0058] Based on the thermally driven structural coefficient rate of latent heat release With the rate of accumulation of sensible heat energy Nonlinear coupling calculations were performed to obtain the combined thermodynamic driving intensity. The nonlinear coupling method is as follows: Introduced as a power-law weight in the coupling expression, the specific form is as follows: The advantage of this power-law coupling method is that when the latent heat contribution is much greater than the sensible heat contribution (i.e., (After truncating the values ​​to a reasonable upper limit and normalizing them), the combined thermal driving intensity will reflect the magnitude of the latent heat release rate more accurately; conversely, when the sensible heat contribution is relatively strong, the combined result will be more affected by the sensible heat energy accumulation rate. This nonlinear structure can more realistically capture the composite characteristics of thermal driving during the South China Sea monsoon rainstorm process, avoiding the ambiguity of physical meaning caused by simple linear superposition. In actual calculations, to ensure numerical stability, the combined thermal driving intensity is further truncated. and All positive values ​​are used in the calculation, while negative values ​​(i.e., the net energy loss stage) are marked and processed separately and are not included in the positive feedback judgment process.

[0059] After calculating the combined thermodynamic driving intensity, the next step is to quantitatively evaluate the dynamic input intensity, including horizontal kinetic energy flux convergence. The independent calculations for each pressure layer have been completed. It is necessary to introduce a layer thickness weighting factor to perform a weighted integration of the contribution of each pressure layer. Let the first layer be... The thickness weighting factor corresponding to the pressure layer is It is defined as the ratio of the pressure thickness of this layer in the atmospheric column to the total pressure thickness, i.e. ,in For the first Interlayer spacing of pressure layers. Convergence of horizontal kinetic energy fluxes across pressure layers. With corresponding layer thickness weighting factor Multiplying them together yields the weighted horizontal kinetic energy input components for that layer. ,Right now The total horizontal kinetic energy input is obtained by vertically integrating the weighted horizontal kinetic energy input components of all pressure layers. ,Right now The introduction of the layer thickness weighting factor ensures that the contribution of the middle troposphere (around 500 hPa) and the lower troposphere (around 850 hPa) to the total kinetic energy input can be reasonably allocated according to the actual pressure thickness, avoiding the problem of overestimation of the contribution of the thin pressure layer near the top of the troposphere due to equal weight summation.

[0060] The total kinetic energy input in the horizontal direction With vertical kinetic energy transfer Perform vector composition operations. Because... Represents the intensity of kinetic energy convergence in the horizontal direction. The two represent the kinetic energy transfer intensity in the vertical direction. In physical space, they correspond to the energy input contributions in two orthogonal directions, horizontal and vertical, respectively. Therefore, vector synthesis uses a modulus calculation method in a Cartesian coordinate system to combine the kinetic input intensity. Defined as The physical basis for this synthesis method lies in the fact that the driving force of the South China Sea summer storm and rain system includes both the horizontal kinetic energy flux convergence formed by low-level convergence and high-level divergence, and the vertical component of kinetic energy transported to the upper levels by rising air currents. These two factors together determine the total power input of the entire system. Using the modulus length rather than an algebraic sum ensures that regardless of the relative magnitudes of the horizontal and vertical components, the combined power input intensity remains non-negative and possesses a clear physical dimension.

[0061] After completing the combined thermal driving intensity Combined dynamic input strength After independent calculation, the net energy conversion efficiency is defined. The difference between the two is, . The physical meaning of is: at the current moment, the net energy transferred from the atmospheric thermodynamic process to the dynamic process exceeds the degree to which the dynamic system itself consumes energy. When When the intensity of thermal driving exceeds the intensity of power input consumption, there is a net energy surplus within the system, the thermal-dynamic positive feedback mechanism is enhanced, and the rainstorm system possesses the energy conditions for continued development; when When this occurs, it indicates that the power input consumption has caught up with or exceeded the thermal drive supply, the system energy conversion has entered a balance or negative feedback stage, and the development of the rainstorm system tends to weaken or remain stable.

[0062] In the actual diagnostic process, it is calculated once for each time step. And record its time series. If within multiple consecutive time steps If all values ​​remain positive, the energy conversion state during that time period is marked as a positive feedback enhancement state, serving as one of the key input conditions for generating subsequent rainstorm diagnostic results. In cases where the value fluctuates frequently around zero, a smoothing window is introduced to perform a moving average on the time series to filter out short-term numerical noise that interferes with state judgment, ensuring that the identification of positive feedback-enhanced states has physical consistency and temporal continuity. This process for calculating net energy conversion efficiency and determining state integrates multidimensional information from the thermodynamic and dynamic fields into a single scalar criterion, providing a quantitative basis for the objective diagnosis of torrential rains in the South China Sea.

[0063] In one optional implementation, the spatiotemporal evolution path of the monsoon composite index field distribution along the time series direction within the spatial integration range is coupled with the positive feedback enhancement state of the energy balance to generate rainstorm diagnosis results for the South China Sea, including: Extract the spatial distribution pattern of the monsoon composite index field at each arrival time within the spatial integration range, calculate the centroid position offset vector of the monsoon composite index field distribution between adjacent arrival times based on the spatial distribution pattern, and connect the centroid position offset vectors according to the time series direction to form a spatiotemporal evolution path. The path direction vector is calculated for the spatiotemporal evolution path, and the movement direction vector in the evolution trend information is extracted as a reference direction vector. The cosine value of the angle between the path direction vector and the reference direction vector is calculated. When the cosine value of the angle is greater than a preset consistency threshold, it is determined that the evolution direction of the spatiotemporal evolution path is consistent with the direction represented by the evolution trend information. The consistency determination result of the spatiotemporal evolution path is logically ANDed with the positive feedback enhancement state of energy balance. When the consistency determination result of the evolution direction is true and the positive feedback enhancement state of energy balance is true, a rainstorm diagnosis result containing the geographical location information of the abnormal change area, the path trajectory information of the spatiotemporal evolution path, and the net energy balance is constructed.

[0064] After obtaining the distribution of the monsoon composite index field and the determination of the enhanced positive feedback state of the energy balance, it is necessary to further couple the two to generate diagnostic results for the South China Sea rainstorm with spatial positioning significance and physical interpretation capabilities. The core of the coupling analysis lies in extracting structured path information from the spatiotemporal evolution of the monsoon composite index field and verifying its logical consistency with the energy balance state, thereby ensuring that the diagnostic results have both dynamic and thermodynamic support.

[0065] For each arrival time within the spatial integration range The corresponding two-dimensional distribution matrix of the monsoon composite index field is used, with the monsoon composite index values ​​of the grid points as weights, to calculate the weighted centroid coordinates of the field distribution at that moment. Let the arrival time be... At that time, there are a total of within the spatial integration range The effective grid point, the first The monsoon composite index value of each grid point is The corresponding longitude coordinates are Latitude coordinates are Then the longitude component of the centroid at that moment With latitude components They are determined by the following formulas respectively: , ,in The index number of the grid points within the spatial integration range. This represents the total number of valid grid points participating in the centroid calculation. By performing the above calculation sequentially for all arrival times, a set of centroid coordinate sequences arranged with the time series is obtained. Adjacent arrival times and The centroid position offset vector between Defined from point to The spherical displacement vector, whose longitude and latitude components are respectively... and Connecting the centroid offset vectors of all adjacent moments sequentially along the time series direction forms a complete spatiotemporal evolution path. This path appears as a broken line trajectory in geographic space, intuitively reflecting the shifting pattern of the centroid of the monsoon composite index field distribution over time.

[0066] After obtaining the spatiotemporal evolution path, it is necessary to quantify and represent the overall direction of the path. Path direction vector. Take the centroid coordinates of the starting node of the path Pointing to the centroid coordinates of the path termination node , where The total number of path nodes is reduced by one. This end-to-end direction definition method effectively suppresses directional jitter caused by local disturbances in the middle section of the path, preserving the main movement trend of the path. Simultaneously, the movement direction vectors of abnormally changing areas are extracted from the evolution trend information as reference direction vectors. The reference direction vector is recorded synchronously when identifying anomalous change regions in the preceding spatiotemporal gradient analysis step, reflecting the dominant movement direction of the anomalous change regions within the gradient analysis time window.

[0067] Calculate path direction vector With reference direction vector cosine value of the angle between The standard vector inner product formula is used:

[0068] in This is the angle between the path direction vector and the reference direction vector. When Greater than the preset consistency threshold When determining whether the direction of the spatiotemporal evolution path aligns with the direction represented by the evolution trend information, the consistency determination result is recorded as true; otherwise, it is recorded as false. A preset consistency threshold is set. Typically, the setting is based on the tolerance for directional deviation in the actual application scenario. In a typical application scenario for diagnosing summer storms and heavy rainfall in the South China Sea, it can be... The value is set to between 0.7 and 0.9, corresponding to an allowable directional deviation range of approximately 25° to 45°, to balance the strictness of path tracking with robustness to local disturbances in the atmospheric flow field.

[0069] The logical operation of determining the consistency of the evolution direction and the positive feedback enhancement state of the energy budget is a crucial step in generating the final rainstorm diagnosis result. Only when the consistency of the evolution direction is true and the positive feedback enhancement state of the energy budget (i.e., net energy conversion efficiency) is the result generated. The logical AND operation of the two conditions is true only when both are greater than zero, triggering the process of constructing the rainstorm diagnosis result. The physical significance of this dual verification mechanism lies in the fact that the spatiotemporal evolution path of the monsoon composite index field is consistent with the movement direction of the anomalous change region, indicating that the dynamic guiding role of the large-scale monsoon circulation remains effective; simultaneously, the energy balance shows a positive feedback enhancement state, indicating that thermal drive and dynamic input jointly maintain and strengthen the atmospheric energy accumulation process. When both conditions are met simultaneously, sufficient physical conditions are met to trigger a marine rainstorm.

[0070] When the logical AND operation yields a true result, a rainstorm diagnostic result containing three core types of information is constructed. The geographic location information of the anomalous change area includes the latitude and longitude boundaries of the spatial integration range, the centroid coordinates of the anomalous change area, and its spatial coverage at each arrival time. This information is stored in the form of a grid list or polygon boundaries, providing precise spatial positioning data for subsequent maritime operation early warnings. The path trajectory information of the spatiotemporal evolution path includes the centroid coordinate sequence corresponding to each arrival time. Centroid offset vector between adjacent nodes Path direction vector and the cosine value of the angle between it and the reference direction vector It fully records the spatiotemporal movement trajectory of the centroid of the monsoon composite index field distribution. The net energy balance information includes net energy conversion efficiency. The specific values ​​and their trends over time provide a quantitative reference for assessing the potential intensity of rainstorms.

[0071] The rainstorm diagnosis results formed by combining the above three types of information geospatially pinpoint the location and movement trend of potential rainstorm areas, provide an assessment of the thermal and dynamic driving intensity of rainstorm triggering in terms of energy physics, and describe the development and evolution trajectory of the rainstorm system in the temporal dimension. Compared with the traditional single index threshold triggering method, this coupled diagnostic method, which combines the consistency of spatiotemporal evolution paths with the positive feedback state of energy balance, can effectively reduce the misjudgment rate caused by local short-term disturbances, while maintaining high identification sensitivity for marine rainstorm processes with clear movement paths and continuous energy supply. It is suitable for operational rainstorm diagnosis under the background of complex air-sea interactions during the active summer monsoon in the South China Sea.

[0072] A second aspect of this invention provides a marine storm diagnosis system based on multi-source remote sensing South China Sea summer monsoon index, comprising: The data acquisition unit is used to acquire multi-source remote sensing observation data of the South China Sea, including wind field data, temperature field data, and humidity field data. The index analysis unit is used to perform vertical layered integral calculation on the wind field data and combine the temperature field data and humidity field data to perform multi-factor coupling to obtain the monsoon comprehensive index field distribution and perform spatiotemporal gradient analysis. After identifying abnormal change areas that meet the preset gradient threshold conditions, the corresponding spatial location information and evolution trend information are extracted. The energy calculation unit is used to determine the spatial integration range of energy calculation based on the spatial location information and to calculate the horizontal kinetic energy flux convergence and vertical kinetic energy transport based on the wind field data. It also determines the time series direction of energy budget analysis based on the evolution trend information and calculates the latent heat energy release rate and sensible heat energy accumulation rate based on the temperature field data and humidity field data. The income and expenditure determination unit is used to compare the combined thermal driving intensity of the latent heat energy release rate and the sensible heat energy accumulation rate with the combined power input intensity of the horizontal kinetic energy flux convergence and the vertical kinetic energy transport, calculate the net energy conversion efficiency, and determine that the energy conversion is in a positive feedback enhancement state when the net energy conversion efficiency is greater than zero. The diagnostic generation unit is used to couple the spatiotemporal evolution path of the monsoon composite index field distribution along the time series direction within the spatial integration range with the positive feedback enhancement state of energy balance to generate the rainstorm diagnostic results in the South China Sea.

[0073] A third aspect of the present invention provides an electronic device, comprising: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the aforementioned method.

[0074] A fourth aspect of the present invention provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the aforementioned method.

[0075] This invention can be a method, apparatus, system, and / or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for performing various aspects of the invention.

[0076] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for diagnosing marine heavy rainfall based on the South China Sea summer monsoon index using multi-source remote sensing, characterized in that... include: Acquire multi-source remote sensing observation data of the South China Sea, including wind field data, temperature field data, and humidity field data; Vertical layered integration is performed on the wind field data, and multi-element coupling is performed by combining the temperature field data and the humidity field data to obtain the monsoon comprehensive index field distribution and perform spatiotemporal gradient analysis. After identifying abnormal change areas that meet the preset gradient threshold conditions, the corresponding spatial location information and evolution trend information are extracted. Based on the spatial location information, the spatial integration range of energy calculation is determined, and the horizontal kinetic energy flux convergence and vertical kinetic energy transport are calculated based on the wind field data. Based on the evolution trend information, the time series direction of energy budget analysis is determined, and the latent heat release rate and sensible heat accumulation rate are calculated based on the temperature field data and humidity field data. By comparing the combined thermal driving intensity of the latent heat energy release rate and the sensible heat energy accumulation rate with the combined power input intensity of the horizontal kinetic energy flux convergence and the vertical kinetic energy transport, the net energy conversion efficiency is calculated. When the net energy conversion efficiency is greater than zero, the energy conversion is determined to be in a positive feedback enhancement state. The spatiotemporal evolution path of the monsoon composite index field distribution along the time series direction within the spatial integration range is coupled with the positive feedback enhancement state of energy balance to generate rainstorm diagnosis results for the South China Sea.

2. The method according to claim 1, characterized in that, Vertical stratified integration is performed on the wind field data, and multi-factor coupling is performed by combining the temperature field data and the humidity field data to obtain the monsoon comprehensive index field distribution and perform spatiotemporal gradient analysis, including: The wind field data is vertically divided according to pressure levels. The horizontal divergence field and vorticity field of the wind field vector in each pressure level are calculated and integrated in the vertical direction to obtain the vertical integral of the wind field characterizing the vertical circulation structure. The temperature field data is extracted vertically according to the pressure level, and the equivalent potential temperature field of each pressure level is calculated by combining the humidity field data. Based on the distribution characteristics of the equivalent potential temperature field in the vertical direction, thermodynamic structural parameters characterizing the thermal stability of the atmosphere are constructed. Based on the ratio of the spatial distribution variance of the vertical integral of the wind field to the spatial distribution variance of the thermal structure parameters, the weighting coefficients are dynamically allocated, and the vertical integral of the wind field and the thermal structure parameters are weighted and coupled to obtain the monsoon comprehensive index field distribution that characterizes the three-dimensional structural features of the monsoon system. The temporal derivative field is calculated along the time dimension and the spatial gradient field is calculated along the spatial dimension for the distribution of the monsoon composite index field. By extracting the joint distribution pattern of the temporal derivative field and the spatial gradient field and identifying the regions in the joint distribution pattern where the values ​​of the temporal derivative field and the spatial gradient field increase synchronously, the abnormal change regions that meet the preset gradient threshold conditions are obtained.

3. The method according to claim 1, characterized in that, After identifying abnormal change regions that meet the preset gradient threshold conditions, the corresponding spatial location information and evolution trend information are extracted, including: Extract the geographic coordinate boundary range and center coordinates of the abnormal change area as the spatial location information; A multi-moment snapshot sequence is constructed along the time dimension for the distribution of the monsoon composite index field within the abnormal change region. The current moment snapshot in the multi-moment snapshot sequence is used as the reference snapshot and the next moment snapshot is used as the target snapshot. A sliding window is set within the spatial range of the reference snapshot and sliding matching is performed within the spatial range of the target snapshot. For each sliding position, the spatial correlation coefficient of the monsoon composite index field values ​​of the reference snapshot and the target snapshot within the coverage area of ​​the sliding window is calculated. All spatial correlation coefficients corresponding to the sliding positions are distributed and mapped according to the spatial coordinates of the sliding positions to form a spatial correlation coefficient field. The maximum position trajectory of the spatial correlation coefficient field is extracted, and the movement direction vector and movement speed scalar of the maximum position trajectory are combined as the evolution trend information.

4. The method according to claim 3, characterized in that, Based on the spatial location information, the spatial integration range for energy calculation is determined, and the horizontal kinetic energy flux convergence and vertical kinetic energy transport are calculated based on the wind field data, including: The horizontal integration boundary is determined based on the geographical coordinate boundary range, and the vertical integration boundary is determined based on the pressure level distribution of the wind field data. The horizontal integration boundary and the vertical integration boundary are combined to form the spatial integration range for energy calculation. The horizontal wind speed vector field of each pressure level is extracted from the wind field data within the spatial integration range. The kinetic energy density distribution of the horizontal wind speed vector field is calculated, and the horizontal divergence field of kinetic energy flux is calculated along the horizontal direction. The horizontal divergence field of kinetic energy flux is then integrated within the horizontal integration boundary and layered within the vertical integration boundary to obtain the horizontal kinetic energy flux convergence. Vertical wind speed components for each pressure level are extracted from the wind field data within the spatial integration range. A vertical kinetic energy term is constructed by calculating the advection transport effect of the vertical wind speed components on the horizontal wind speed vector field. The vertical kinetic energy term is then used to perform gradient calculations for each pressure level along the vertical direction to form a vertical kinetic energy transport gradient field. The vertical kinetic energy transport gradient field is then subjected to surface integral operation within the horizontal integration boundary and integral operation within the vertical integration boundary to obtain the horizontal kinetic energy flux convergence, which characterizes the intensity of kinetic energy input.

5. The method according to claim 3, characterized in that, Based on the evolution trend information, the time series direction of energy balance analysis is determined, and the latent heat release rate and sensible heat accumulation rate are calculated based on the temperature field data and the humidity field data, including: The spatial movement path of the abnormal change region is determined based on the movement direction vector, and the arrival time sequence of the abnormal change region along the spatial movement path is calculated based on the movement speed scalar. The time sequence of the arrival time sequence is used as the time series direction of energy balance analysis. Temperature and humidity profiles at each arrival time are extracted from the temperature and humidity data within the abnormal change region according to the time series direction. The saturation difference distribution between the saturated vapor pressure and the actual vapor pressure at each pressure level is calculated using the temperature and humidity profiles. The amount of water vapor condensation is calculated based on the saturation difference distribution, and the latent heat release per unit volume at each pressure level is calculated by combining the latent heat coefficient of water vapor condensation. The time derivative of the latent heat release per unit volume is calculated along the time series direction to obtain the latent heat release rate. The temperature field distribution at each arrival time is extracted from the temperature field data within the abnormal change region according to the time series direction. The product of the temperature field distribution and the air pressure field distribution is calculated as the sensible thermal energy density distribution. The sensible thermal energy density distribution is subjected to volume integral operation within the spatial integration range to obtain the total sensible thermal energy. The time derivative of the total sensible thermal energy is calculated along the time series direction to obtain the sensible thermal energy accumulation rate.

6. The method according to claim 1, characterized in that, By comparing the combined thermodynamic driving intensity of the latent heat energy release rate and the sensible heat energy accumulation rate with the combined dynamic input intensity of the horizontal kinetic energy flux convergence and the vertical kinetic energy transport, the net energy conversion efficiency is calculated. When the net energy conversion efficiency is greater than zero, the energy conversion is determined to be in a state of positive feedback enhancement, including: The latent heat release rate and the sensible heat accumulation rate are integrated over time along the time series direction to obtain the cumulative latent heat release amount and the cumulative sensible heat increment. The ratio of the cumulative latent heat release amount to the cumulative sensible heat increment is calculated as the thermal driving structure coefficient. Based on the thermal driving structure coefficient, the latent heat release rate and the sensible heat accumulation rate are nonlinearly coupled and calculated to obtain the combined thermal driving intensity. The horizontal kinetic energy flux convergence is calculated by multiplying the horizontal kinetic energy flux convergence of each pressure level within the spatial integration range with the layer thickness weighting factor of the corresponding pressure level as a weighted horizontal kinetic energy input component. The weighted horizontal kinetic energy input component of all pressure levels is vertically integrated to obtain the total horizontal kinetic energy input and is vector-combined with the vertical kinetic energy transport. The magnitude of the composite vector is calculated as the combined dynamic input intensity. The difference between the combined thermodynamic driving intensity and the combined dynamic input intensity is calculated as the net energy conversion efficiency. When the net energy conversion efficiency is greater than zero, the energy conversion state is marked as a positive feedback enhancement state.

7. The method according to claim 6, characterized in that, The spatiotemporal evolution path of the monsoon composite index field distribution along the time series direction within the spatial integration range is coupled with the positive feedback enhancement state of energy balance to generate the following rainstorm diagnosis results for the South China Sea: Extract the spatial distribution pattern of the monsoon composite index field at each arrival time within the spatial integration range, calculate the centroid position offset vector of the monsoon composite index field distribution between adjacent arrival times based on the spatial distribution pattern, and connect the centroid position offset vectors according to the time series direction to form a spatiotemporal evolution path. The path direction vector is calculated for the spatiotemporal evolution path, and the movement direction vector in the evolution trend information is extracted as a reference direction vector. The cosine value of the angle between the path direction vector and the reference direction vector is calculated. When the cosine value of the angle is greater than a preset consistency threshold, it is determined that the evolution direction of the spatiotemporal evolution path is consistent with the direction represented by the evolution trend information. The consistency determination result of the spatiotemporal evolution path is logically ANDed with the positive feedback enhancement state of energy balance. When the consistency determination result of the evolution direction is true and the positive feedback enhancement state of energy balance is true, a rainstorm diagnosis result containing the geographical location information of the abnormal change area, the path trajectory information of the spatiotemporal evolution path, and the net energy balance is constructed.

8. A marine rainstorm diagnostic system based on the South China Sea summer monsoon index using multi-source remote sensing, used to implement the method as described in any one of claims 1-7, characterized in that, include: The data acquisition unit is used to acquire multi-source remote sensing observation data of the South China Sea, including wind field data, temperature field data, and humidity field data. The index analysis unit is used to perform vertical layered integral calculation on the wind field data and combine the temperature field data and humidity field data to perform multi-factor coupling to obtain the monsoon comprehensive index field distribution and perform spatiotemporal gradient analysis. After identifying abnormal change areas that meet the preset gradient threshold conditions, the corresponding spatial location information and evolution trend information are extracted. The energy calculation unit is used to determine the spatial integration range of energy calculation based on the spatial location information and to calculate the horizontal kinetic energy flux convergence and vertical kinetic energy transport based on the wind field data. It also determines the time series direction of energy budget analysis based on the evolution trend information and calculates the latent heat energy release rate and sensible heat energy accumulation rate based on the temperature field data and humidity field data. The income and expenditure determination unit is used to compare the combined thermal driving intensity of the latent heat energy release rate and the sensible heat energy accumulation rate with the combined power input intensity of the horizontal kinetic energy flux convergence and the vertical kinetic energy transport, calculate the net energy conversion efficiency, and determine that the energy conversion is in a positive feedback enhancement state when the net energy conversion efficiency is greater than zero. The diagnostic generation unit is used to couple the spatiotemporal evolution path of the monsoon composite index field distribution along the time series direction within the spatial integration range with the positive feedback enhancement state of energy balance to generate the rainstorm diagnostic results in the South China Sea.

9. An electronic device, characterized in that, include: processor; Memory used to store processor-executable instructions; The processor is configured to invoke instructions stored in the memory to execute the method according to any one of claims 1 to 7.

10. A computer-readable storage medium having computer program instructions stored thereon, characterized in that, When the computer program instructions are executed by the processor, they implement the method described in any one of claims 1 to 7.

Images