Basin glacier substance balance analysis method and system based on big data

Abstract

The invention relates to the technical field of glacier research, and discloses a drainage basin glacier material balance analysis method and system based on big data, and the method comprises the steps: collecting remote sensing images, ground observation and meteorological data of a glacier drainage basin, carrying out the standardization processing and fusion of multi-source data, constructing a random forest model, analyzing the relevance between glacier changes and climate factors, and carrying out the analysis of the relevance. And determining a key driving factor. The dynamic response of glacier substances is further predicted by using a long-short-term memory network, high-sensitivity climate factors are determined through sensitivity analysis, and the prediction model is optimized. A dynamic response database is also established for long-term monitoring, and an automatic early warning mechanism is set to evaluate the abnormal change risk. According to the method, accurate analysis and prediction of glacier change and climate factor interaction are realized, and an effective tool is provided for glacier change research and environmental risk assessment.

Description

Technical Field

[0001] The present invention relates to the technical field of glacier research, and in particular, to a method and system for analyzing the mass balance of basin glaciers based on big data. Background Art

[0002] Currently, as an important part of the Earth's climate system, glaciers play a crucial indicative role in global water resources, ecological balance, and climate change. Their dynamic changes not only reflect the evolution trend of the natural environment but also directly affect the water use safety and ecological stability in downstream areas. Therefore, it is particularly crucial to study the response mechanism of glaciers. However, current research methods mostly rely on traditional observation means and simple statistical models, making it difficult to comprehensively capture the complexity of glacier changes and the impact of multi-factor coupling, and the prediction accuracy and comprehensive evaluation ability are significantly limited.

[0003] In the existing technology, the field of glacier research faces significant technical challenges. First and foremost is how to effectively integrate the observational data of glacier basins with multi-dimensional environmental parameters to form a complete data set that can reflect the true environmental characteristics. Due to the scattered data sources and complex dimensions, the lack of data integration directly leads to an insufficient understanding of the driving factors of glacier changes during model construction. Furthermore, this lack of understanding makes it extremely difficult to build a prediction model that can accurately simulate the dynamic response of glacier mass, especially when facing the complex interaction of climate factors, the model often fails to accurately reveal the sensitivity of glaciers to environmental changes.

[0004] Therefore, how to build a prediction model that can accurately simulate the dynamic response of glacier mass and analyze its sensitivity to climate factors based on the fusion of multi-source data has become a key problem that needs to be solved urgently in current research. Summary of the Invention

[0005] The present invention provides a method and system for analyzing the mass balance of basin glaciers based on big data to accurately simulate the dynamics of glaciers.

[0006] In a first aspect, to solve the above technical problems, the present invention provides a method for analyzing the mass balance of basin glaciers based on big data, including: Obtaining the original data set of the glacier and obtaining an initial data set in a unified format; Successively obtaining a comprehensive data set, a set of key driving factors, a prediction result for simulating the dynamic response of glacier mass, and a sensitivity distribution characteristic according to the initial data set; Obtaining an intuitive analysis result according to the prediction result, constructing a dynamic response database, and regularly updating the prediction result to obtain long-term monitoring data; According to the long-term monitoring data, if the change amplitude within a certain time period is detected to exceed the preset amplitude threshold, potential abnormal points of environmental parameters are judged, and a corresponding risk assessment report is generated.

[0007] Preferably, the steps of sequentially obtaining the comprehensive data set, the set of key driving factors, the prediction results of the simulated glacier mass dynamics response, and the sensitivity distribution characteristics from the initial data set include: Performing spatio-temporal alignment and feature extraction on the initial data set through data fusion technology, and performing data dimensionality reduction and integration through the weighted average method and the principal component analysis method to obtain the fused comprehensive data set; Based on the glacier changes and climate factors in the comprehensive data set, constructing a random forest model based on machine learning, and determining the set of key driving factors through model training and feature importance ranking; Based on the set of key driving factors, combining the mass dynamics characteristics of glacier changes, constructing a prediction model based on the long short-term memory network to obtain the prediction results of the simulated glacier mass dynamics response; According to the prediction results, performing perturbation tests on the prediction model, calculating the influence weights of the climate factors on the glacier changes, and determining the sensitivity distribution characteristics.

[0008] Preferably, the steps of obtaining the original data set of the glacier and obtaining the initial data set in a unified format include: Obtaining the original data covering glacier changes, climate factors, and environmental parameters from remote sensing images, ground observations, and meteorological data platforms, and performing preliminary format adjustment on the original data to obtain a preliminary sorted data set; According to the preliminary sorted data set, using image processing software to uniformly adjust the resolution of the remote sensing images, and aligning the sampling frequencies of the ground observations and the meteorological data platforms to obtain an intermediate data set with consistent resolution; Performing standardization processing on the intermediate data set, using data cleaning tools to remove outliers and missing values. If the detected missing values in the intermediate data set exceed the preset missing threshold, interpolation methods are used for filling to obtain a cleaned standard data set; According to the standard data set, using data integration tools to fuse data from different sources in a unified format to generate an initial data set containing glacier changes, climate factors, and environmental parameters.

[0009] Preferably, the steps of obtaining the comprehensive data set include: Obtaining the data source differences of the glacier changes and the climate factors in the initial data set, and using data alignment tools to perform calibration processing on the spatial distribution and time series to obtain a calibrated aligned data set; Use a feature extraction tool to separate the spatial eigenvalues corresponding to the glacier changes and the temporal eigenvalues corresponding to the climate factors in the calibrated alignment dataset, and determine the extracted feature dataset; If the eigenvalue distribution in the feature dataset is unbalanced, use a weighted average tool to balance and adjust the spatial eigenvalues and the temporal eigenvalues to obtain a balanced feature dataset; According to the balanced feature dataset, use a principal component analysis tool to reduce the dimensionality and integrate the eigenvalues to obtain the final comprehensive dataset.

[0010] Preferably, obtaining the set of key driving factors includes: Obtain the correlation analysis between the glacier changes and the climate factors from the comprehensive dataset, use a data cleaning tool to process the missing values and outliers in the comprehensive dataset, obtain the cleaned complete dataset, and determine the variable range in the complete dataset; Through the cleaned complete dataset, obtain variable screening and feature importance evaluation, use a correlation calculation tool to compare the correlation coefficients between the glacier changes and the climate factors to obtain a variable combination with a higher correlation; According to the variable combination with a higher correlation and the preliminary judgment of the driving factors, use a sorting tool to sort the variable combination by importance to obtain a sorted set of key variables.

[0011] Preferably, obtaining the prediction result of simulating the dynamic response of glacier mass includes: According to the set of key driving factors and the dynamic characteristics of glacier mass, use a data integration tool to standardize the environmental variables in the time series to obtain an organized time series data group, and determine the distribution range of the time series data group; Through the organized time series data group, use a feature extraction tool to perform hierarchical decomposition on the time series data group to obtain separated feature units; Judge whether the feature unit reaches the preset feature threshold standard. If it reaches, use a prediction tool to perform parameterization processing on the feature unit to obtain a generated prediction parameter group; Use a mapping tool to match and integrate the prediction parameter group with the spatio-temporal distribution to obtain the final trend distribution map.

[0012] Preferably, obtaining the sensitivity distribution characteristics includes: According to the prediction result, use a data organization tool to normalize the input parameters to obtain an organized parameter dataset; Determine whether the parameter data set reaches a preset threshold standard. If it reaches, perform multiple rounds of transformation on the parameter data set to obtain a transformed perturbation parameter combination; Perform an association process on the perturbation parameter combination and environmental variables to obtain the weight data of the climate factors; Integrate the weight data with the change trend to obtain the final distribution characteristic image.

[0013] Preferably, obtaining an intuitive analysis result includes: Obtain environmental variable data and climate factor data corresponding to the glacier change from a pre-established database and perform classification processing to obtain a classified data set; If the data set contains characteristic information of the change trend, match the classified data set with a preset distribution characteristic template to obtain a mapped data combination; Integrate the data combination with the visual style of the change trend to obtain rendered graphic data; Fuse the graphic data with the sensitivity distribution characteristics to obtain the final chart data; Constructing a dynamic response database and regularly updating the prediction results to obtain long-term monitoring data includes: Obtain the original records related to the glacier change, combine the historical data of the climate factors to form an initial comprehensive data set, and obtain a structured data grouping; Extract the characteristics of the structured data grouping, match the glacier change with the characteristic values sensitive to the factors. If the characteristic values exceed the preset characteristic threshold, mark the data grouping as highly sensitive data and determine the association characteristics with the dynamic response; Integrate the highly sensitive data with the prediction results to generate an updated version of the dynamic response database, regularly compare the interaction characteristics in the dynamic response database, and obtain the latest trend tracking records; Fuse the trend tracking records with the target data of long-term monitoring to generate a distribution chart reflecting the relationship between glacier change and climate factors.

[0014] Preferably, according to the long-term monitoring data, if it is detected that the change amplitude within a certain time period exceeds the preset amplitude threshold, judge potential environmental parameter abnormal points and generate a corresponding risk assessment report, including: Obtain the real-time record of the change amplitude during the monitoring period. If the change amplitude exceeds the preset change threshold, trigger an automatic warning mechanism to obtain a preliminary abnormal fluctuation identifier; Match the abnormal fluctuation identifier with the records of historical environmental parameters, mine the correlation characteristics of the historical environmental parameters, and determine the potential distribution of abnormal points; Comprehensively compare the correlation characteristics between the distribution of abnormal points and the historical environmental parameters, obtain the corresponding risk levels, and judge the distribution range of high-risk areas; Integrate the distribution range of the high-risk areas with the abnormal fluctuation identifier to generate a detailed report including the risk levels and the distribution of abnormal points.

[0015] In a second aspect, the present invention provides a big data-based analysis system for the mass balance of basin glaciers, including: A first acquisition module for acquiring the original data set of the glacier and obtaining an initial data set in a unified format; A second acquisition module for successively obtaining a comprehensive data set, a set of key driving factors, a prediction result of simulating the dynamic response of glacier mass, and a sensitivity distribution characteristic according to the initial data set; An update module for obtaining an intuitive analysis result according to the prediction result, constructing a dynamic response database, regularly updating the prediction result, and obtaining long-term monitoring data; A generation module for judging potential abnormal points of environmental parameters and generating a corresponding risk assessment report according to the long-term monitoring data if it is detected that the change amplitude within a certain time period exceeds a preset amplitude threshold.

[0016] Compared with the prior art, the present invention provides a big data-based analysis method and system for the mass balance of basin glaciers, which collect remote sensing images, ground observations and meteorological data of glacier basins, perform standardized processing and fusion on multi-source data, construct a random forest model to analyze the correlation between glacier changes and climate factors, determine key driving factors, use a long short-term memory network to predict the dynamic response of glacier mass, and determine highly sensitive climate factors through sensitivity analysis to optimize the prediction model. The present invention also establishes a dynamic response database for long-term monitoring and sets up an automatic early warning mechanism to evaluate the risk of abnormal changes. This method realizes the accurate analysis and prediction of the interaction between glacier changes and climate factors, and provides an effective tool for glacier change research and environmental risk assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is a flowchart of a big data-based analysis method for the mass balance of basin glaciers provided by the first embodiment of the present invention; Figure 2 is a schematic structural diagram of a big data-based analysis system for the mass balance of basin glaciers provided by the second embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0018] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0019] Referring to Figure 1 , the first embodiment of the present invention provides a flowchart of a method for analyzing the mass balance of basin glaciers based on big data, including the following steps: S11, obtaining the original data set of the glacier and obtaining the initial data set in a unified format; S12, successively obtaining a comprehensive data set, a set of key driving factors, a prediction result for simulating the dynamic response of glacier mass, and a sensitivity distribution characteristic according to the initial data set; S13, obtaining an intuitive analysis result according to the prediction result, constructing a dynamic response database, regularly updating the prediction result, and obtaining long-term monitoring data; S14, according to the long-term monitoring data, if it is detected that the change amplitude within a certain time period exceeds the preset amplitude threshold, then judge the potential environmental parameter anomaly points and generate a corresponding risk assessment report.

[0020] In step S11, the original data set of the glacier is obtained and the initial data set in a unified format is obtained.

[0021] The obtaining of the original data set of the glacier and obtaining the initial data set in a unified format includes: Obtaining the original data covering glacier changes, climate factors, and environmental parameters from remote sensing images, ground observations, and meteorological data platforms, and performing preliminary format adjustment on the original data to obtain a preliminarily sorted data set; According to the preliminarily sorted data set, using image processing software to uniformly adjust the resolution of the remote sensing image, and aligning the sampling frequencies of the ground observations and the meteorological data platforms to obtain an intermediate data set with consistent resolution; Performing standardization processing on the intermediate data set, using a data cleaning tool to remove outliers and missing values. If it is detected that the missing values in the intermediate data set exceed the preset missing threshold, then use the interpolation method to fill them to obtain a cleaned and standardized data set; According to the standardized data set, using a data integration tool to fuse the data from different sources in a unified format to generate an initial data set containing glacier changes, climate factors, and environmental parameters.

[0022] For example, when processing data related to glacier changes, the original data may come from remote sensing images, ground observation stations, and meteorological data platforms, with significant differences in data formats.

[0023] For example, in the resolution unified adjustment process, there may be inconsistencies in the spatial resolution of remote sensing images. Some images have a resolution of 10 meters, while others have a resolution of 30 meters. Through image processing software such as ENVI, the low-resolution images can be upsampled to a unified 10-meter resolution to ensure the consistency of spatial information. At the same time, the sampling frequencies of ground observations and meteorological data may be different. For example, the ground observation data is once a day, while the meteorological data is once an hour. By using the time averaging method, the meteorological data can be aligned to the daily average to match the ground observation frequency, forming an intermediate dataset with consistent resolution. This alignment operation helps to avoid time or space misalignment during subsequent data fusion and improves the analysis accuracy.

[0024] For example, in the standardization processing and data cleaning stage, the intermediate dataset may contain outliers and missing values. Suppose in the temperature data of a meteorological station, a certain day is recorded as -50 degrees, which is significantly deviated from the normal range. Through data cleaning tools such as the Pandas library in Python, it can be marked as an outlier and removed. If the proportion of missing values exceeds a preset threshold, such as 10%, the linear interpolation method can be used to fill in the missing data.

[0025] For example, if the data of a glacier thickness observation point is missing for three consecutive days, interpolation estimation can be performed based on the data trends of the previous and next two days. This cleaning and filling operation can ensure the integrity and reliability of the data, laying a foundation for subsequent analysis.

[0026] For example, in the data integration stage, for the cleaned and standardized dataset, data integration tools such as ArcGIS can be used to fuse the data of glacier changes, climate factors, and environmental parameters according to unified time and space coordinates. Suppose the glacier area data comes from remote sensing images, the temperature and precipitation data come from meteorological stations, and the soil moisture data comes from ground observations. These data can be matched by year and geographical location to generate an initial dataset containing multi-dimensional information. This fusion method can comprehensively reflect the correlation between glacier changes and environmental factors and provide data support for studying the relationship between glacier retreat and climate change.

[0027] It should be noted that the technical processing of each of the above links can significantly improve the data quality.

[0028] For example, format unification avoids data reading errors, resolution alignment reduces spatial and time errors, cleaning and interpolation ensure data integrity, and integration provides convenience for comprehensive analysis. These steps work together to ensure that the final dataset is both accurate and practical, providing a reliable basis for glacier monitoring and climate research.

[0029] In step S12, a comprehensive dataset, a set of key driving factors, a prediction result of simulating the dynamic response of glacier mass, and a sensitivity distribution characteristic are sequentially obtained according to the initial dataset.

[0030] The obtaining of the comprehensive dataset includes: Obtain the data source differences between the glacier changes and the climate factors in the initial dataset, and use a data alignment tool to calibrate the spatial distribution and time series to obtain a calibrated aligned dataset; Use a feature extraction tool to separate the spatial feature values corresponding to the glacier changes and the time feature values corresponding to the climate factors in the calibrated aligned dataset, and determine the extracted feature dataset; If the eigenvalue distribution in the feature dataset is unbalanced, balance the spatial feature values and the time feature values through a weighted average tool to obtain a balanced feature dataset; According to the balanced feature dataset, use a principal component analysis tool to reduce the dimensionality and integrate the eigenvalues to obtain the final comprehensive dataset.

[0031] For example, when processing the initial dataset, for the data source differences between glacier changes and climate factors, the alignment of spatial distribution and time series is particularly important. Spatial distribution alignment can be understood as matching data from different sources according to a unified geographic coordinate system, while time series alignment is to ensure the consistency of data in the time dimension. Suppose the glacier area data is sourced from remote sensing images, covering the longitude and latitude grid of a certain basin, and climate factors such as temperature data are sourced from weather stations, recording the hourly changes at specific points. When using a data alignment tool, the weather station data can be extended to the same grid range as the remote sensing image through spatial interpolation methods. For example, the point temperature data can be interpolated to each square kilometer grid cell, and at the same time, the hourly data is summarized by daily mean to align with the daily update frequency of the glacier data.

[0032] For example, in the feature extraction stage, for the calibrated aligned dataset, separating the spatial feature values related to glacier changes and the time feature values related to climate factors is a key step. Spatial feature values may include the distribution density of glacier coverage area or the amplitude of boundary changes, while time feature values may involve seasonal fluctuations of temperature or precipitation. Suppose in a certain glacier basin, remote sensing images show that the glacier edge has retreated about 50 meters per year in the past 10 years, which can be used as a spatial feature value; while meteorological data shows that the average temperature rises by 0.5 degrees every summer, which can be used as a time feature value. The feature extraction tool will separate these data from the original dataset to form an independent feature dataset for subsequent analysis.

[0033] For example, if the eigenvalue distribution in the feature dataset is unbalanced, such as the proportion of glacier area change data in the spatial eigenvalues being too high while the precipitation data in the temporal eigenvalues is relatively sparse, then a weighted average tool is needed for balance adjustment. The specific implementation method can be to assign a lower weight, such as 0.4, to the spatial eigenvalues and a higher weight, such as 0.6, to the temporal eigenvalues according to the importance and coverage of the data, so as to avoid the excessive influence of a certain type of feature on subsequent analysis. The adjusted balanced feature dataset can more comprehensively reflect the relationship between glaciers and climate factors.

[0034] For example, in the dimensionality reduction and integration stage, the principal component analysis tool is used to process the balanced feature dataset, aiming to compress the high-dimensional eigenvalues into a few comprehensive indicators. Suppose the original feature dataset contains multiple variables such as glacier area, thickness, temperature, and precipitation. Through principal component analysis, these variables can be integrated into 2 - 3 main components, reducing the data complexity while retaining most of the information. Such a comprehensive dataset is more suitable for the research on the correlation between glacier changes and climate factors, facilitating subsequent modeling.

[0035] It should be noted that the processing of each of the above links is closely carried out around the fields of glacier monitoring and climate research, ensuring that the data is gradually optimized from alignment to feature extraction, then to balance adjustment and dimensionality reduction and integration, laying a foundation for in-depth analysis of the relationship between glacier retreat and environmental changes. Especially in feature extraction and balance adjustment, aiming at the particularity of glacier basin data, reasonably allocating weights and separating features can effectively improve the representativeness and analysis accuracy of the data.

[0036] The obtaining of the set of key driving factors includes: Obtain the correlation analysis between the glacier changes and the climate factors from the comprehensive dataset, use a data cleaning tool to process the missing values and outliers in the comprehensive dataset, obtain a complete dataset after cleaning, and determine the variable range in the complete dataset; Through the complete dataset after cleaning, obtain variable screening and feature importance assessment, use a correlation calculation tool to compare the correlation coefficients between the glacier changes and the climate factors, and obtain a variable combination with a relatively high correlation; According to the variable combination with a relatively high correlation and the preliminary judgment of the driving factors, use a sorting tool to sort the importance of the variable combination and obtain a sorted set of key variables.

[0037] For example, when processing comprehensive datasets, data cleaning is a crucial step in the correlation analysis of glacier changes and climate factors. Data cleaning tools are mainly used to handle missing values and outliers in the dataset to ensure the accuracy of subsequent analyses. Suppose in a comprehensive dataset of a certain glacier basin, the glacier area data for some years is missing due to cloud cover in remote sensing images, and there are extremely high values in some climate factors such as precipitation data, which may be recording errors. The cleaning tool can fill in the missing glacier area data through interpolation methods, such as estimating the data for missing years based on the average values of the previous and subsequent years; for outliers, a reasonable range can be set. For example, if the precipitation exceeds twice the historical maximum value, it is considered abnormal and excluded, obtaining a complete cleaned dataset.

[0038] For example, in the variable range determination and screening stage, a correlation calculation tool is used to evaluate the strength of the association between glacier changes and climate factors. Suppose the cleaned dataset contains variables such as glacier area, glacier thickness, annual average temperature, and annual precipitation. Through correlation calculation, it can be found that the correlation coefficient between glacier area and annual average temperature is as high as 0.8, while the correlation coefficient with precipitation is only 0.3. This indicates that temperature may be an important variable affecting glacier changes, and then a variable combination with a relatively high correlation, such as the combination of glacier area and temperature, is selected to provide a key direction for subsequent analysis.

[0039] For example, for the preliminary judgment of driving factors, a sorting tool ranks the importance of variable combinations to form a set of key variables. Suppose in the above combination, the influence weight of temperature on glacier area change is evaluated as 0.7, while the influence weight of other variables such as wind speed is only 0.2. Through the sorting tool, temperature can be listed as the primary key variable, and a threshold requirement can be set, such as the weight needs to be greater than 0.5. If all the variables in the set of key variables meet this threshold, it proceeds to the next step of analysis. This approach helps to focus on the factors that have the most significant impact on glacier changes.

[0040] For example, in environmental impact and time series analysis, a data visualization tool is used to draw charts of the spatial distribution and time series characteristics of key variables. Suppose for the association between temperature and glacier area, a spatial distribution map is drawn within a certain basin, which can visually show the overlap between areas with higher temperatures and areas with decreasing glacier area; at the same time, a time series chart can present the trend of increasing temperature year by year and decreasing glacier area year by year over the past 20 years. This way of chart drawing facilitates researchers to quickly grasp the association distribution between variables and provides an intuitive basis for further research. Such a processing method can effectively improve the efficiency of data interpretation and lay a foundation for the analysis of the driving mechanism of glacier changes.

[0041] The obtained prediction results of simulating the dynamic response of glacier mass include: According to the set of key driving factors and the dynamic characteristics of glacier mass, a data integration tool is used to standardize the environmental variables in the time series, obtain an organized time series data set, and determine the distribution range of the time series data set; Through the organized time series data set, a feature extraction tool is used to perform hierarchical decomposition on the time series data set to obtain separated feature units; Determine whether the feature unit reaches a preset feature threshold standard. If it reaches, a prediction tool is used to perform parameterization processing on the feature unit to obtain a generated prediction parameter set; A mapping tool is used to match and integrate the prediction parameter set with the spatio-temporal distribution to obtain a final trend distribution map.

[0042] For example, when studying the relationship between the dynamic characteristics of glacier mass and key driving factors, the standardization processing of the data integration tool is particularly important. The standardization processing aims to eliminate the dimensional differences of different environmental variables and ensure that the time series data can be compared on the same scale. Suppose in a time series data set of a glacier basin, variables such as temperature, precipitation, and glacier volume change are included. The temperature is in degrees Celsius and ranges from -5 to 10, while the precipitation is in millimeters and ranges from 200 to 800. The data integration tool can convert these variables into standard values with a mean of 0 and a standard deviation of 1 for subsequent analysis. This processing method helps to unify the data distribution and avoid biases caused by different dimensions.

[0043] For example, for non-linear feature separation, the hierarchical decomposition of the feature extraction tool is a key link. Non-linear features often hide in complex data and are difficult to directly observe. Through hierarchical decomposition, the data set can be split into multiple feature units, each representing a different change pattern. Suppose in the above basin data, the change in glacier volume is affected by both temperature and precipitation. The feature extraction tool can separate the rapid changes driven by temperature and the long-term trends driven by precipitation to form two feature units. If the preset threshold standard is that the explanatory power of the feature unit needs to reach 80% of the total variance, then it is judged whether the condition is met. This separation method helps to more clearly identify the action mechanisms of various factors.

[0044] For example, in the correlation modeling of driving factors, the parameterization processing of the prediction tool is used to quantify the relationship between the feature unit and glacier change. Suppose the separated temperature-driven feature unit is selected as the main analysis object. The prediction tool will generate a set of prediction parameters based on historical data. For example, for every 1-degree increase in temperature, the glacier volume may decrease by 0.5%. Subsequently, by comparing the historical records with the prediction results, it is judged whether the parameter set is applicable to the current basin. This parameterization method provides a reliable basis for subsequent trend simulation.

[0045] For example, for trend simulation, the mapping tool matches and integrates the prediction parameter groups with the spatio-temporal distribution to form a trend distribution map. Suppose in the above-mentioned basin, the distribution map generated based on the temperature parameter shows that the trend of glacier volume reduction in the southern region is more significant, which is consistent with the actually observed glacier retreat area. If the distribution map has a high degree of coincidence with the historical trend, it is considered to meet the requirements of response simulation. This matching and integration helps to visually present the spatial differences in glacier changes and provides a reference for further research.

[0046] For example, in the above-mentioned various links, standardization processing, hierarchical decomposition, parameterization processing, and mapping integration form a complete analysis chain. From data collation to trend presentation, each step closely focuses on the relationship between glacier changes and environmental factors, ensuring the logic and consistency of the analysis. This systematic processing method can effectively reveal the driving mechanism behind glacier dynamics and provide support for research in related fields.

[0047] The obtaining of the sensitivity distribution characteristics includes: According to the prediction results, use a data collation tool to normalize the input parameters to obtain a collated parameter data set; Judge whether the parameter data set reaches a preset threshold standard. If it reaches, perform multiple rounds of transformation on the parameter data set to obtain a transformed perturbation parameter combination; Perform an association process on the perturbation parameter combination and the environmental variables to obtain the weight data of the climate factors; Integrate the weight data with the change trend to obtain the final distribution characteristic image.

[0048] For example, when studying the interaction between climate factors and glacier changes, the normalization processing of the data collation tool is particularly crucial. The purpose of normalization processing is to unify input parameters with different dimensions to a comparable scale for subsequent analysis. Suppose in the study of a glacier basin, the input parameters include annual average temperature, precipitation, and wind speed, etc. The temperature range is from -8 to 5 degrees Celsius, the precipitation range is from 300 to 900 millimeters, and the wind speed range is from 2 to 10 meters per second. Through the data collation tool, these parameters can be transformed into standardized data with a mean of 0 and a standard deviation of 1, ensuring that different variables will not cause deviations due to dimensional differences in the analysis. This processing method lays a foundation for subsequent judgment of whether the parameter data set reaches the preset threshold standard.

[0049] For example, regarding the judgment of whether a parameter dataset reaches a preset threshold standard, a specific standard can be set. For instance, the normalized variance of the dataset needs to reach more than 85% of the total variance. Suppose in the above-mentioned basin data, after normalization, the variance interpretation rate of the dataset reaches 88%, then it is considered to meet the conditions. This judgment method helps to screen out representative parameter sets and provides reliable inputs for subsequent perturbation response analysis.

[0050] For example, in the adjustment test of parameter perturbation response, the adjustment test tool can perform multiple rounds of transformations on the parameter dataset to generate perturbed parameter combinations. Suppose the air temperature is the main perturbation factor, and its perturbation range is set to ±2 degrees Celsius. By simulating the changing trends of glacier volume under different perturbations with the tool, a set of perturbed parameter combinations is obtained. This method can help analyze the degree of influence of perturbations on glacier changes and provides data support for further research.

[0051] For example, regarding the analysis of influencing weights, the data matching tool can correlate the perturbed parameter combinations with environmental variables to obtain the weight data of each climate factor. Suppose in the above data, it is found through analysis that the weight of air temperature on glacier changes is 0.6, precipitation is 0.3, and wind speed is 0.1. Then it can be judged that the weight data reflects the dynamic response characteristics. This correlation process helps to clarify the role sizes of each factor and provides a basis for subsequent analysis.

[0052] For example, during the generation of the sensitivity distribution, the spatial mapping tool integrates the weight data with the changing trends to generate a distribution characteristic image. Suppose in the above-mentioned basin, the image generated based on the weight data shows that the northern region is more sensitive to air temperature changes, which is consistent with the actually observed glacier retreat distribution. Then it is considered that the image reflects the characteristics of the response mechanism. This integration method visually presents the spatial differences in glacier changes and provides important references for research.

[0053] It should be noted that the above-mentioned links from data collation to the generation of distribution characteristic images form a complete analysis chain. Each link closely focuses on the interaction between climate factors and glacier changes to ensure strict logic. Through normalization processing, perturbation testing, weight analysis, and spatial mapping, the driving mechanism of glacier changes can be gradually revealed, providing support for research in related fields. The beneficial effect brought by this systematic analysis lies in improving the accuracy of data processing and the visualization degree of results.

[0054] In step S13, an intuitive analysis result is obtained based on the prediction result, a dynamic response database is constructed, and the prediction result is updated regularly to obtain long-term monitoring data.

[0055] The obtaining of the intuitive analysis result includes: Obtain environmental variable data and climate factor data corresponding to the glacier change from a pre-established database and perform classification processing to obtain a classified data set; If the data set contains characteristic information of the change trend, match the classified data set with a preset distribution characteristic template to obtain a mapped data combination; Integrate the data combination with the visual style of the change trend to obtain rendered graphic data; Fuse the graphic data with the sensitivity distribution characteristics to obtain the final chart data; The construction of the dynamic response database, regularly update the prediction results to obtain long-term monitoring data, including: Obtain the original records related to the glacier change, combine the historical data of the climate factors to form an initial comprehensive data set, and obtain a structured data grouping; Extract the characteristics of the structured data grouping, match the glacier change with the characteristic values sensitive to the factors. If the characteristic values exceed the preset characteristic threshold, mark the data grouping as highly sensitive data and determine the association characteristics with the dynamic response; Integrate the highly sensitive data with the prediction results to generate an updated version of the dynamic response database, regularly compare the interaction characteristics in the dynamic response database to obtain the latest trend tracking records; Fuse the trend tracking records with the target data of long-term monitoring to generate a distribution chart reflecting the relationship between glacier change and climate factors.

[0056] According to the association between the sensitivity distribution and the climate factors, use a data screening tool to classify multiple climate factors, obtain a set of highly sensitive factors that meet the preset threshold criteria, and determine whether the set of highly sensitive factors is associated with the environmental variables. If the set of highly sensitive factors is associated with the environmental variables, for the prediction optimization process, use a parameter calibration tool to adjust the parameters of the set of highly sensitive factors to obtain an adjusted parameter data group, and judge whether the parameter data group meets the response characteristic criteria. Through the adjusted parameter data group, for the adjustment of weight allocation, use a data mapping tool to integrate the parameter data group with the change trend to obtain an optimized weight allocation combination, and determine whether the weight allocation combination reflects the distribution image characteristics. According to the optimized weight allocation combination, for the optimization process of highly sensitive factors, use a data fusion tool to match the weight allocation combination with a pre-established prediction template to obtain the final prediction optimization data, and judge whether the prediction optimization data reflects the sensitivity distribution characteristics.

[0057] If the influence weight of one of the climate factors exceeds the preset weight threshold, it is marked as a highly sensitive factor, and the highly sensitive factor is input into the prediction model for secondary optimization to adjust the parameters and weight allocation of the prediction model, and an optimized prediction result is obtained.

[0058] For example, when studying the dynamic response characteristics of glacier changes, the application of data collection tools is the key first step. By extracting environmental variable data and climate factor data from a pre-established database, a foundation can be laid for subsequent analysis. Suppose in a glacier basin study, the environmental variables include terrain height and aspect, and the climate factors cover annual average temperature and rainfall. The data collection tools preliminarily organize these data according to time series and spatial distribution to form an original data set. This process ensures the comprehensiveness and systematicness of the data and provides reliable input for classification processing.

[0059] For example, for the classification processing of data, the data can be divided according to the degree of influence through a grouping method.

[0060] In a possible implementation, assume that the annual average temperature has a significant impact on glacier melting, and rainfall plays an important role in glacier recharge. The classification processing will divide the data into a main influence group and a secondary influence group. In the classified data set, the annual average temperature data may show an increasing trend year by year, while the rainfall data shows seasonal fluctuations. When judging whether it contains information on the characteristics of the change trend, if the temperature data has continuously increased by more than 2 degrees Celsius in the past 10 years, it is considered to have an obvious trend characteristic, providing a basis for subsequent matching.

[0061] For example, in the use of data mapping tools, the classified data set is matched with a preset distribution characteristic template. Suppose the template is constructed based on the historical glacier retreat rate, and the mapping process associates the temperature increase trend with the reduction of glacier area. If the mapped data combination shows that for every 1 degree Celsius increase in temperature, the glacier area decreases by about 5%, it is considered to reflect the relevance of the influence weight. This matching method helps to reveal the key driving factors of glacier changes.

[0062] For example, for the generation of a trend chart, a graphic rendering tool integrates the data combination with a visual style. Suppose a line chart is used to show the relationship between temperature change and glacier area. In the rendered graphic data, the temperature curve and the area reduction curve show a synchronous change trend. If the graphic data clearly shows the correlation between the two, it is considered to reflect the associated characteristics of the dynamic response. This visual presentation method facilitates researchers to intuitively understand the relationship between the data.

[0063] For example, in the generation of a heat map, the visualization drawing tool integrates graphic data with the distribution characteristics of influence weights. Suppose the heat map uses the depth of color to represent the degree of influence of temperature on different glacier regions. In the chart data, the color is darker near the low-altitude area, indicating a greater temperature influence. If the chart data is consistent with the actual glacier melting distribution, it is considered to reflect the correlation characteristics of environmental variables. This method of illustration provides an intuitive reference for spatial distribution for researchers and helps to identify key impact areas.

[0064] For example, in the logical progression from the core solution to the extended solution, the above-mentioned links form a complete data processing and visualization chain. The core solution focuses on data collection and classification, while the extended solution gradually deepens the understanding of the correlation between glacier changes and climate factors through mapping, rendering, and fusion. Suppose in a certain glacier study, through this chain, it is found that the influence weight of temperature on the marginal area of the glacier is higher than that of the central area, which provides a direction for the focus of subsequent research. This way of multi-link cooperation not only improves the depth of data analysis but also provides a multi-dimensional perspective for the study of glacier dynamic response.

[0065] In step S14, according to the long-term monitoring data, if the change amplitude within a certain time period is detected to exceed the preset amplitude threshold, potential environmental parameter abnormal points are judged, and a corresponding risk assessment report is generated.

[0066] The step of judging potential environmental parameter abnormal points and generating a corresponding risk assessment report according to the long-term monitoring data, if the change amplitude within a certain time period is detected to exceed the preset amplitude threshold, includes: Obtain a real-time record of the change amplitude during the monitoring period. If the change amplitude exceeds the preset change threshold, trigger an automatic warning mechanism to obtain a preliminary abnormal fluctuation identifier; Match the abnormal fluctuation identifier with the records of historical environmental parameters, mine the correlation characteristics of the historical environmental parameters, and determine the distribution of potential abnormal points; Comprehensively compare the distribution of abnormal points with the correlation characteristics of the historical environmental parameters, obtain the corresponding risk level, and judge the distribution range of high-risk areas; Integrate the distribution range of high-risk areas with the abnormal fluctuation identifier to generate a detailed report including the risk level and the distribution of abnormal points.

[0067] For example, when constructing a data acquisition framework, raw records related to glacier changes can be obtained from multiple environmental variable sources. Suppose in a glacier monitoring project, the data sources include satellite remote sensing images, ground meteorological station records, and historical archive data. These sources provide information such as glacier area changes, surface temperature fluctuations, and precipitation. By integrating these data, a comprehensive dataset containing time series and spatial distributions is formed. This approach ensures data diversity and coverage, laying a solid foundation for subsequent analysis.

[0068] For example, for the classification and sorting of the dataset, the comprehensive dataset can be grouped according to the types of environmental variables and climate factors.

[0069] In a possible implementation, assume that the dataset contains variables such as altitude, slope, average annual temperature, and annual precipitation. The classification and sorting will divide this data into a terrain-related group and a climate-related group. The terrain group data may show that the glacier distribution is more concentrated in high-altitude areas, while the climate group data reflects that the temperature changes more significantly in low-altitude areas. This grouping method helps to sort out the hierarchical relationships between the data and facilitates subsequent feature extraction.

[0070] For example, in feature extraction and matching analysis, the data processing tool can perform sensitivity analysis on structured data groups. Suppose the extracted feature values include the temperature change rate and the glacier area reduction rate. If the temperature change rate of a certain group of data exceeds the preset threshold of 0.5 degrees Celsius per year, it is marked as highly sensitive data. This marking method can highlight the key influencing factors, clarify the dynamic relationship between glacier changes and climate factors, and provide a targeted basis for further research.

[0071] For example, by integrating highly sensitive data and prediction results through a data storage tool, an updated version of the dynamic response database can be generated. Suppose in the database update, the newly added data shows that a certain glacier area has decreased by 10% in the past 5 years and is highly correlated with the rising temperature trend. Regularly comparing these data can track trend changes and update the monitoring focus in a timely manner. This continuous update method helps to maintain the timeliness of the data and provides support for long-term research.

[0072] For example, when using a visualization drawing tool to generate distribution charts, the trend tracking records can be integrated with long-term monitoring data. Suppose the chart shows the relationship between glacier changes and temperature in different regions with the depth of color, and the darker areas indicate more significant temperature effects. If the chart shows that the low-altitude areas are darker, which is consistent with the actual monitoring data, it indicates that it well reflects the interactive characteristics of the comprehensive dataset. This intuitive presentation method facilitates researchers to quickly identify key areas and improve analysis efficiency.

[0073] For example, during the long-term monitoring of glacier changes, the identification and handling of abnormal fluctuations are particularly important. The real-time recording function of data collection tools can help capture the change amplitude of glacier area or thickness. Suppose in a monitoring project, the preset threshold for glacier area change is 5% per year. If the area of a certain region decreases by 3% within half a year, the data collection tool will immediately record this change and trigger an automatic warning mechanism to generate a preliminary abnormal fluctuation identification. This real-time monitoring method can quickly respond to potential problems and provide timely basis for subsequent analysis.

[0074] For example, for the matching of abnormal fluctuation identifications with historical environmental parameters, data comparison tools can play a key role. Suppose historical records show that the temperature in a certain region has been continuously rising in the past 10 years, with an average annual increase of 0.3 degrees Celsius, while the precipitation has decreased by 20%. By comparing the abnormal fluctuation identifications with these parameters, it can be found that there is a strong correlation between the decrease in glacier area and the increase in temperature. After further exploring the correlation characteristics between parameters using trend analysis tools, this region is determined as a potential abnormal point. This method helps to screen out key regions from massive data and improve the pertinence of analysis.

[0075] For example, in the risk assessment stage, the risk assessment generation tool comprehensively compares the correlation characteristics between the distribution of abnormal points and environmental parameters. Suppose the glacier in a high-altitude region has a large change amplitude and is highly correlated with temperature and precipitation changes. The tool will classify it as a high-risk level according to the preset criteria. After further analysis, it is found that the coverage area of this region accounts for about 15% of the total monitoring area and is located in the core area of the glacier, making it an object of key concern. This risk level classification can clearly define the distribution range of high-risk regions and provide important reference for subsequent decision-making.

[0076] For example, by using the report generation tool to integrate the distribution range of high-risk regions and abnormal fluctuation identifications, a detailed risk assessment report can be generated. Suppose the report indicates that a certain region has the highest risk level, and the abnormal points are concentrated in the area above 4000 meters above sea level, along with a glacier change trend chart and environmental parameter fluctuation records for the past 5 years. This integration method ensures the comprehensiveness and intuitiveness of the report content, facilitating researchers to quickly understand the situation and formulate countermeasures. Potential influencing factors can also be marked in the report, such as the possible exacerbation of risks caused by continuous temperature rise, providing support for long-term planning.

[0077] For example, throughout the process, the implementation methods of various technical themes are closely integrated, forming a complete chain from data collection to risk assessment. Whether it is the real-time recording of abnormal fluctuations, the matching analysis of historical parameters, or the division of risk levels and the generation of the final report, each link focuses on the interaction between glacier changes and environmental factors. This systematic processing method not only improves the monitoring efficiency but also provides reliable data support for glacier protection and climate change research, ensuring that key areas receive timely attention and effective management.

[0078] For those skilled in the art, it is obvious that the present application is not limited to the details of the above exemplary embodiments, and without departing from the spirit or basic characteristics of the present application, the present application can be implemented in other specific forms. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present application is defined by the appended claims rather than the above description. Therefore, all changes falling within the meaning and scope of the equivalent elements of the claims are intended to be included in the present application. Any reference signs in the claims should not be regarded as limiting the claims involved.

[0079] In summary, the present invention discloses a method for analyzing the mass balance of basin glaciers based on big data. By collecting remote sensing images, ground observations, and meteorological data of glacier basins, standardizing and fusing multi-source data, constructing a random forest model to analyze the correlation between glacier changes and climate factors, and determining key driving factors. Further, a long short-term memory network is used to predict the dynamic response of glacier mass, and high-sensitive climate factors are determined through sensitivity analysis to optimize the prediction model. The present invention also establishes a dynamic response database for long-term monitoring and sets up an automatic early warning mechanism to evaluate the risk of abnormal changes. This method realizes the accurate analysis and prediction of the interaction between glacier changes and climate factors, providing an effective tool for glacier change research and environmental risk assessment.

[0080] Refer to Figure 2 , the second embodiment of the present invention provides a structural diagram of a system for analyzing the mass balance of basin glaciers based on big data, including: The first acquisition module 201 is used to acquire the original data set of the glacier and obtain an initial data set in a unified format; The second acquisition module 202 is used to sequentially obtain a comprehensive data set, a set of key driving factors, a prediction result of simulating the dynamic response of glacier mass, and a sensitivity distribution characteristic according to the initial data set; The update module 203 is used to obtain an intuitive analysis result according to the prediction result, construct a dynamic response database, regularly update the prediction result, and obtain long-term monitoring data; A generation module 204 is configured to, based on the long-term monitoring data, if it detects that the change amplitude within a certain time period exceeds a preset amplitude threshold, determine potential environmental parameter anomaly points and generate corresponding risk assessment reports.

[0081] It should be noted that the system for analyzing the mass balance of basin glaciers based on big data provided in the embodiments of the present invention is used to execute all the process steps of the method for analyzing the mass balance of basin glaciers based on big data in the above embodiments. The working principles and beneficial effects of the two correspond one by one, so they will not be elaborated here.

[0082] Embodiments of the present invention also provide an electronic device. The electronic device includes: a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a program for analyzing the mass balance of basin glaciers based on big data. When the processor executes the computer program, it implements the steps in the above embodiments of the method for analyzing the mass balance of basin glaciers based on big data, such as Figure 1 Step S11 shown. Alternatively, when the processor executes the computer program, it implements the functions of each module / unit in the above device embodiments, such as the first acquisition module.

[0083] Exemplarily, the computer program can be divided into one or more modules / units. The one or more modules / units are stored in the memory and executed by the processor to complete the present invention. The one or more modules / units can be a series of computer program instruction segments capable of performing specific functions, and these instruction segments are used to describe the execution process of the computer program in the electronic device.

[0084] The electronic device can be a computing device such as a desktop computer, a notebook, a palm computer, and a smart tablet. The electronic device may include, but is not limited to, a processor and a memory. Those skilled in the art can understand that the above components are only examples of the electronic device and do not constitute a limitation to the electronic device. It may include more or fewer components than the above, or combine some components, or different components. For example, the electronic device may further include input / output devices, network access devices, a bus, etc.

[0085] The so-called processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The processor is the control center of the electronic device and connects various parts of the entire electronic device through various interfaces and circuits.

[0086] The memory can be used to store the computer programs and / or modules. By running or executing the computer programs and / or modules stored in the memory, and by calling the data stored in the memory, the processor realizes various functions of the electronic device. The memory mainly includes a program storage area and a data storage area. Among them, the program storage area can store an operating system, application programs required for at least one function (such as a sound playback function, an image playback function, etc.); the data storage area can store data created according to the use of the mobile phone (such as audio data, phone book, etc.). In addition, the memory can include high-speed random access memory, and can also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, at least one magnetic disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0087] Among them, if the modules / units integrated in the electronic device are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, to implement all or part of the processes in the above-described embodiment methods of the present invention, it can also be completed by a computer program instructing relevant hardware. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above-described various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file, or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc. It should be noted that the content included in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practice, the computer-readable medium does not include electrical carrier signals and telecommunication signals.

[0088] It should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed to multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. In addition, in the accompanying drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which can be specifically implemented as one or more communication buses or signal lines. Those of ordinary skill in the art can understand and implement it without creative effort.

[0089] The above-described specific embodiments have further elaborated on the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above is only the specific embodiments of the present invention and is not used to limit the protection scope of the present invention. It is particularly pointed out that for those skilled in the art, any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for analyzing the mass balance of basin glaciers based on big data, characterized in that, Including: Obtain the original dataset of the glacier and get the initial dataset in a unified format; Successively obtain the comprehensive dataset, the set of key driving factors, the prediction results of the simulated glacier mass dynamic response, and the sensitivity distribution characteristics according to the initial dataset; Obtain intuitive analysis results according to the prediction results, construct a dynamic response database, regularly update the prediction results, and obtain long-term monitoring data; According to the long-term monitoring data, if it is detected that the change amplitude within a certain period exceeds the preset amplitude threshold, then judge the potential environmental parameter anomaly points and generate the corresponding risk assessment report.

2. The method for analyzing the mass balance of basin glaciers based on big data according to claim 1, wherein The obtaining the comprehensive dataset, the set of key driving factors, the prediction results of the simulated glacier mass dynamic response, and the sensitivity distribution characteristics successively according to the initial dataset includes: Perform spatio-temporal alignment and feature extraction on the initial dataset through data fusion technology, and perform data dimensionality reduction and integration through the weighted average method and the principal component analysis method to obtain the fused comprehensive dataset; Construct a random forest model based on machine learning according to the glacier changes and climate factors in the comprehensive dataset, and determine the set of key driving factors through model training and feature importance ranking; According to the set of key driving factors, combine the mass dynamic characteristics of glacier changes to construct a prediction model based on the long short-term memory network, and obtain the prediction results of the simulated glacier mass dynamic response; According to the prediction results, conduct perturbation tests on the prediction model, calculate the influence weights of the climate factors on the glacier changes, and determine the sensitivity distribution characteristics.

3. The method for analyzing the mass balance of basin glaciers based on big data according to claim 1 or 2, characterized in that, The obtaining the original dataset of the glacier and getting the initial dataset in a unified format includes: Obtain the original data covering glacier changes, climate factors, and environmental parameters from remote sensing images, ground observations, and meteorological data platforms, and perform preliminary format adjustment on the original data to obtain a preliminarily sorted data set; According to the preliminarily sorted data set, use image processing software to uniformly adjust the resolution of the remote sensing image, and align the sampling frequencies of the ground observations and the meteorological data platforms to obtain an intermediate data set with consistent resolution; Perform standardization processing on the intermediate data set, use data cleaning tools to remove outliers and missing values, and if it is detected that the missing values in the intermediate data set exceed the preset missing threshold, then use the interpolation method to fill them to obtain a cleaned and standardized data set; According to the standardized data set, use data integration tools to fuse the data from different sources in a unified format to generate an initial data set containing glacier changes, climate factors, and environmental parameters.

4. The method for analyzing the mass balance of basin glaciers based on big data according to claim 2, wherein, The obtaining the comprehensive dataset includes: Obtain the data source differences of the glacier changes and the climate factors in the initial dataset, and use data alignment tools to calibrate the spatial distribution and time series to obtain a calibrated alignment dataset; Use feature extraction tools to separate the spatial feature values corresponding to glacier changes and the time feature values corresponding to climate factors in the calibrated alignment dataset, and determine the extracted feature dataset; If the eigenvalue distribution in the feature dataset is unbalanced, balance and adjust the spatial eigenvalue and the temporal eigenvalue through a weighted average tool to obtain a balanced feature dataset; According to the balanced feature dataset, use a principal component analysis tool to perform dimensionality reduction and integration on the eigenvalues to obtain a final comprehensive dataset.

5. The method for analyzing the mass balance of basin glaciers based on big data according to claim 2, characterized in that, The obtaining of the set of key driving factors includes: Obtain the correlation analysis between the glacier change and the climate factor from the comprehensive dataset, use a data cleaning tool to process the missing values and outliers in the comprehensive dataset to obtain a complete cleaned dataset, and determine the variable range in the complete dataset; Through the complete cleaned dataset, obtain variable screening and feature importance evaluation, use a correlation calculation tool to compare the correlation coefficients between the glacier change and the climate factor to obtain a variable combination with a relatively high correlation; According to the variable combination with a relatively high correlation and the preliminary judgment of the driving factors, use a sorting tool to sort the importance of the variable combination to obtain a sorted set of key variables.

6. The method for analyzing the mass balance of basin glaciers based on big data according to claim 1, characterized in that The obtaining of the prediction result of the simulated dynamic response of glacier mass includes: According to the set of key driving factors and the dynamic characteristics of glacier mass, use a data integration tool to standardize the environmental variables in the time series to obtain an organized time series data group, and determine the distribution range of the time series data group; Through the organized time series data group, use a feature extraction tool to perform hierarchical decomposition on the time series data group to obtain separated feature units; Judge whether the feature unit reaches a preset feature threshold standard. If it reaches, use a prediction tool to perform parameterization processing on the feature unit to obtain a generated prediction parameter group; Use a mapping tool to match and integrate the prediction parameter group with the spatio-temporal distribution to obtain a final trend distribution map.

7. The method for analyzing the mass balance of basin glaciers based on big data according to claim 2, wherein, The obtaining of the sensitivity distribution characteristics includes: According to the prediction result, use a data organization tool to normalize the input parameters to obtain an organized parameter dataset; Judge whether the parameter dataset reaches a preset threshold standard. If it reaches, perform multiple rounds of transformation on the parameter dataset to obtain a transformed perturbation parameter combination; Perform an association process on the perturbation parameter combination and the environmental variables to obtain the weight data of the climate factor; Integrate the weight data with the change trend to obtain a final distribution characteristic image.

8. The method for analyzing the mass balance of basin glaciers based on big data according to claim 7, characterized in that, The obtaining of the intuitive analysis result includes: Obtain the environmental variable data and climate factor data corresponding to the glacier change from a pre-established database and perform classification processing to obtain a classified data set; If the data set contains the characteristic information of the change trend, match the classified data set with a preset distribution characteristic template to obtain a mapped data combination; Integrate the data combination with the visual style of the change trend to obtain rendered graphic data; Fuse the graphic data with the sensitivity distribution characteristics to obtain a final chart data; The construction of the dynamic response database, regularly update the prediction result to obtain long-term monitoring data, includes: Obtain the original records related to glacier changes, combine with the historical data of the climate factors to form an initial comprehensive dataset, and obtain a structured data grouping; Extract features from the structured data grouping, match the eigenvalues sensitive to glacier changes and factors, if the eigenvalues exceed the preset feature threshold, then mark the data grouping as highly sensitive data, and determine the associated characteristics with dynamic response; Integrate the highly sensitive data with the prediction results to generate an updated version of the dynamic response database, regularly compare the interaction characteristics in the dynamic response database, and obtain the latest trend tracking records; Integrate the trend tracking records with the target data of long-term monitoring to generate a distribution chart reflecting the relationship between glacier changes and climate factors.

9. The method for analyzing the mass balance of basin glaciers based on big data according to claim 1, wherein According to the long-term monitoring data, if it is detected that the change amplitude in a certain time period exceeds the preset amplitude threshold, then judge potential abnormal points of environmental parameters, and generate a corresponding risk assessment report, including: Obtain the real-time record of the change amplitude during the monitoring period, if the change amplitude exceeds the preset change threshold, then trigger an automatic warning mechanism to obtain a preliminary abnormal fluctuation identifier; Match the abnormal fluctuation identifier with the records of historical environmental parameters, mine the associated characteristics of the historical environmental parameters, and determine the distribution of potential abnormal points; Comprehensively compare the distribution of abnormal points with the associated characteristics of the historical environmental parameters, obtain the corresponding risk level, and judge the distribution range of high-risk areas; Integrate the distribution range of high-risk areas with the abnormal fluctuation identifier to generate a detailed report including the risk level and the distribution of abnormal points.

10. A system for analyzing the material balance of glaciers in a river basin based on big data, characterized in that, Including: The first acquisition module is used to acquire the original dataset of glaciers and obtain an initial dataset in a unified format; The second acquisition module is used to sequentially obtain a comprehensive dataset, a set of key driving factors, a prediction result of simulating the dynamic response of glacier mass, and a sensitivity distribution characteristic according to the initial dataset; The update module is used to obtain an intuitive analysis result according to the prediction result, construct a dynamic response database, regularly update the prediction result, and obtain long-term monitoring data; The generation module is used to judge potential abnormal points of environmental parameters and generate a corresponding risk assessment report according to the long-term monitoring data if it is detected that the change amplitude in a certain time period exceeds the preset amplitude threshold.

Images