Big data methods and application platform for climate change monitoring in the Third Pole region
Through data fusion and prediction models, the problems of multi-source data capture differences and insufficient prediction reliability in climate monitoring in the third pole region are solved, and more accurate climate change predictions are achieved.
Patent Information
- Application Number
- CN202411310544.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-20
- Publication Date
- 2024-12-06
- Estimated Expiration
- 2044-09-20
AI Technical Summary
It is difficult to accurately and comprehensively capture the subtle differences and dynamic evolution of multi-source data in climate monitoring in the third pole region, resulting in insufficient accuracy and reliability of climate change predictions.
Technical means of data fusion, data correction and construction of climate change prediction models are used to collect ground observation stations and satellite remote sensing data, extract terrain characteristics, perform weighted fusion and quality evaluation, and use machine learning and deep learning models to make predictions.
The accuracy and reliability of climate change prediction in the third pole region have been improved, and a consistent and reliable climate monitoring system has been formed.
Smart Images

Figure CN118821074B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field related to climate data management, and specifically to a big data method and application platform for climate change monitoring in the Third Pole region. Background Art
[0002] The Third Pole refers to the plateau and its surrounding mountainous areas. As an important part of the Earth system, it has become a hot spot for global climate change research due to its unique climate characteristics and complex geographical environment. Climate change in the Third Pole has a profound impact on global climate change. It is particularly important to conduct accurate climate monitoring and data integration management in the Third Pole. However, the high altitude, vastness, extreme climate conditions, glaciers and permafrost areas in the Third Pole have brought unprecedented challenges to climate monitoring. Its vast territory and complex terrain conditions make it difficult for traditional ground observation stations to fully cover, resulting in spatial and temporal limitations in data acquisition. The extreme Climate conditions, such as severe cold, strong winds, and high radiation, place higher demands on the stability of observation equipment and data quality. Glacier retreat and permafrost thawing have an important impact on the climate and environment of the Third Pole region. Traditional climate monitoring methods mainly rely on data from ground observation stations, but are limited by the distribution density and observation conditions of the observation stations. It is difficult to fully reflect the complex and changeable climate characteristics of the Third Pole region. Climate data from different sources (such as ground observation stations, satellite remote sensing, etc.) differ in accuracy, resolution, and timeliness. How to effectively integrate these multi-source data to form a consistent and reliable climate monitoring system and make accurate and reliable climate change predictions has become an urgent problem to be solved.
[0003] Therefore, at the current stage, there are technical problems in the technologies related to climate monitoring and management in the Third Pole region, such as difficulty in accurately and comprehensively capturing the subtle differences and dynamic evolution of multi-source data in the Third Pole region, which in turn leads to insufficient accuracy and reliability in climate change predictions. Summary of the invention
[0004] This application provides a big data method and application platform for climate change monitoring in the Third Pole region, and adopts technical means such as data fusion, data correction, and construction of a climate change prediction model. It solves the technical problem that it is difficult to accurately and comprehensively capture the subtle differences and dynamic evolution of multi-source data in the Third Pole region, which in turn leads to insufficient accuracy and reliability in climate change predictions in existing climate monitoring and management, thereby achieving the technical effect of improving the accuracy and reliability of climate change predictions in the Third Pole region.
[0005] The present application provides a big data method for monitoring climate change in the Third Pole region, the method comprising: collecting real-time data from ground observation stations and satellite remote sensing data to obtain multi-source climate data; extracting elevation information, slope information and landform information from different regions in the Third Pole region to obtain the terrain features of the Third Pole region; weightedly fusing the multi-source climate data based on a data fusion strategy for the terrain features to generate climate monitoring data; performing real-time quality assessment on the climate monitoring data to obtain abnormal data, and correcting the abnormal data using neighboring site data and historical climate monitoring data to generate corrected data; using a climate change prediction model to perform climate change prediction on the corrected data to generate predicted climate change results; displaying the predicted climate change results through a visualization window, and generating a climate report.
[0006] In a possible implementation, real-time data from ground observation stations and satellite remote sensing data are collected to obtain multi-source climate data, and the following processing is performed: the format of the real-time data is converted to generate standardized real-time data; the standardized real-time data is aligned in time and space to generate time-space consistent real-time data; noise data and invalid data in the time-space consistent real-time data are filtered through signal processing to generate filtered real-time data; low-frequency access data in the filtered real-time data is compressed and stored using a compression algorithm to generate compressed real-time data and add it to the multi-source climate data.
[0007] In a possible implementation, the data fusion strategy based on the terrain features performs weighted fusion on the multi-source climate data to generate climate monitoring data, and performs the following processing: using a machine learning model to identify the correlation between the multi-source climate data and establish preliminary fusion strategies for different time nodes; adaptively adjusting the data source weights in the preliminary fusion strategy according to the terrain features to generate fusion weights; dynamically adjusting the preliminary fusion strategy using a time series analysis method to generate a data fusion strategy; combining a preferred portion of the multi-source climate data based on the data fusion strategy and the fusion weights to generate preliminary climate monitoring data; performing accuracy verification on the preliminary climate monitoring data, adjusting the data fusion strategy according to the accuracy verification results, and fusing to obtain the climate monitoring data.
[0008] In a possible implementation, the climate monitoring data is subjected to real-time quality assessment to obtain abnormal data, the abnormal data is corrected using neighborhood site data and historical climate monitoring data to generate corrected data, and the following processing is performed: a quality assessment index system is set, the quality assessment index system includes a data integrity index, a spatiotemporal consistency index, and an outlier detection index; the climate monitoring data is monitored in real time based on the quality assessment index system to identify the abnormal data; the abnormal data is analyzed using a climate monitoring model constructed based on the historical climate monitoring data to obtain an abnormal analysis result; the abnormal data is corrected according to the abnormal analysis result using the neighborhood site data to generate the corrected data.
[0009] In a possible implementation, the abnormal data is corrected according to the abnormal analysis result through the neighborhood site data to generate the corrected data, and the following processing is also performed: the neighborhood site data is extracted to calculate the correction value of the abnormal analysis result to generate the correction value; the abnormal data is interpolated and reconstructed for adjacent points based on time and space to generate supplementary data; and a spatiotemporal consistency check is performed on the supplementary data. If the consistency check result meets the consistency check threshold, the supplementary data is added to the corrected data.
[0010] In a possible implementation, the climate change prediction of the corrected data is performed using a climate change prediction model to generate a predicted climate change result, and the following processing is also performed: historical climate change data, historical glacier retreat data, and historical permafrost thaw data are extracted based on the historical corrected data; a deep learning model is trained and feedback trained based on the historical climate change data, historical glacier retreat data, and historical permafrost thaw data to generate the climate change prediction model, wherein the climate change prediction model includes a first prediction model and a second prediction model; the corrected data is input into the first prediction model to obtain glacier retreat prediction data and permafrost thaw prediction data, and the glacier retreat prediction data and permafrost thaw prediction data are input into the second prediction model; and the predicted temperature parameters, predicted precipitation parameters, predicted wind speed parameters, and predicted snow depth parameters obtained by the second prediction model are added to the predicted climate change result.
[0011] In a possible implementation, the Third Pole regional climate change monitoring big data method also performs the following processing: allocates computing tasks between local processing and cloud analysis to obtain preliminary allocation results; introduces a dynamic resource scheduling mechanism to adjust the preliminary allocation results according to real-time climate conditions to obtain allocation results; and monitors the data transmission and processing of the allocation results in real time through blockchain technology. If the monitoring verification results meet the integrity threshold and security threshold, the computing task is executed.
[0012] The present application also provides a big data application platform for monitoring climate change in the Third Pole region, including: a multi-source climate data acquisition module, which is used to collect real-time data from ground observation stations and satellite remote sensing data to obtain multi-source climate data; a Third Pole region terrain feature acquisition module, which is used to extract elevation information, slope information and landform information of different regions in the Third Pole region to obtain the terrain features of the Third Pole region; a climate monitoring data generation module, which is used to perform weighted fusion of the multi-source climate data based on the data fusion strategy of the terrain features to generate climate monitoring data; a correction data generation module, which is used to perform real-time quality assessment of the climate monitoring data, obtain abnormal data, and use neighboring site data and historical climate monitoring data to correct the abnormal data to generate corrected data; a predicted climate change result generation module, which is used to use a climate change prediction model to perform climate change prediction on the corrected data to generate predicted climate change results; a result visualization display module, which is used to display the predicted climate change results through a visualization window and generate a climate report.
[0013] Through the big data method and application platform for climate change monitoring in the Third Pole region proposed in this application, real-time data from ground observation stations and satellite remote sensing data are collected to obtain multi-source climate data; elevation information, slope information and landform information of different regions in the Third Pole region are extracted to obtain the terrain characteristics of the Third Pole region; the multi-source climate data are weighted and fused based on the data fusion strategy of the terrain characteristics to generate climate monitoring data; real-time quality assessment of the climate monitoring data is performed to obtain abnormal data, and the abnormal data is corrected using the data of neighboring sites and historical climate monitoring data to generate corrected data; climate change prediction of the corrected data is performed using a climate change prediction model to generate predicted climate change results; the predicted climate change results are displayed through a visualization window and a climate report is generated. This solves the technical problem that it is difficult to accurately and comprehensively capture the subtle differences and dynamic evolution of multi-source data in the Third Pole region, which leads to insufficient accuracy and reliability of climate change predictions in existing climate monitoring management, and achieves the technical effect of improving the accuracy and reliability of climate change predictions in the Third Pole region. BRIEF DESCRIPTION OF THE DRAWINGS
[0014] In order to more clearly illustrate the technical solution of the embodiment of the present disclosure, the accompanying drawings of the embodiment of the present disclosure are briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the platform according to the embodiment of the present application. It should be understood that the previous or following operations are not necessarily performed precisely in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.
[0015] Figure 1A schematic diagram of a big data flow chart of a method for monitoring climate change in the Third Pole region provided in an embodiment of the present application;
[0016] Figure 2 A schematic diagram of the structure of the big data application platform for climate change monitoring in the Third Pole region provided in an embodiment of the present application.
[0017] Explanation of the reference numerals: multi-source climate data acquisition module 10, Third Pole region terrain feature acquisition module 20, climate monitoring data generation module 30, correction data generation module 40, climate change prediction result generation module 50, result visualization display module 60. DETAILED DESCRIPTION
[0018] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.
[0019] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.
[0020] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, platform, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.
[0021] The present application embodiment provides a big data method for monitoring climate change in the Third Pole region, such as Figure 1 As shown, the method includes:
[0022] Step S100, collect real-time data from ground observation stations and satellite remote sensing data to obtain multi-source climate data. Collecting real-time data from ground observation stations and satellite remote sensing data refers to collecting real-time climate data from different sources in the Third Pole region (plateau and surrounding mountainous areas) through ground observation stations and satellite remote sensing to obtain multi-source climate data. Specifically, ground observation station data is data obtained by continuously monitoring and recording climate elements using professional instruments at fixed observation stations set up on the ground. It usually includes but is not limited to meteorological elements (such as temperature, humidity, air pressure, wind speed, wind direction, precipitation, etc.), radiation observations (such as solar radiation, surface reflected radiation, long-wave radiation, etc.), and soil parameters (such as soil temperature, humidity, heat flux, etc.). The advantage of ground observation station data is its high accuracy and continuity, and it can provide detailed information on local climate change. However, due to the limited distribution of ground observation stations and the difficulty of setting up stations at high altitudes and remote areas, its coverage is relatively limited; satellite remote sensing data is obtained through Through non-contact observation of the surface by sensors on satellites, large-scale, high-resolution climate-related information in the Third Pole region is obtained, usually including multispectral / hyperspectral data (used to monitor surface vegetation cover, land use changes, water distribution, etc.), microwave remote sensing data (used to monitor soil moisture, precipitation, snow accumulation, etc., and can penetrate clouds to obtain surface information), atmospheric observation data (such as temperature, humidity, wind field, cloud cover, etc., obtained through atmospheric detection instruments on satellites). The advantage of satellite remote sensing data is its wide coverage and fast update speed, which can make up for the shortcomings of ground observation stations in spatial distribution. However, since satellite observations are affected by many factors (such as cloud cover, sensor performance, etc.), there may be certain uncertainties in the data quality and accuracy. Finally, these real-time climate monitoring data are combined into a multi-source climate data set to provide data support for climate change prediction.
[0023] In a possible implementation, step S100 further includes step S110, performing format conversion of the real-time data to generate standardized real-time data. Performing format conversion on real-time data from different sources such as ground observation stations and satellite remote sensing, for example, converting different data formats (such as CSV, NetCDF, HDF5, etc.) into a standardized format that the system can recognize and process, and generating standardized real-time data. Step S120, aligning the standardized real-time data in time and space to generate time-space consistent real-time data. Adjusting the time and space resolution of the data so that all standardized real-time data are aligned in time (i.e., ensuring that the timestamps of all data points are consistent) and aligned in space (i.e., ensuring that the geographic area and resolution covered by the data are consistent), for example, if the data collection frequency of one data source is once an hour, and the data of another data source is once a minute, interpolation or resampling is required to unify the time scale; spatial alignment may be resampling images or raster data of different resolutions to the same resolution, generating time-space consistent real-time data, so that the spatial and time dimensions of the data are consistent.
[0024] Step S130, filtering the noise data and invalid data in the temporally consistent real-time data by signal processing to generate filtered real-time data. Apply signal processing technology (such as filtering, denoising, smoothing, etc.) to process the temporally consistent real-time data (including noise data and invalid data). Specifically, remove and correct these noise data and invalid data by signal processing technology to generate higher quality filtered real-time data, wherein the noise data may be caused by sensor failure, environmental interference or data transmission error, and the invalid data may include null values, extreme values or abnormal values beyond a reasonable range. Step S140, using a compression algorithm to compress and store the low-frequency access data in the filtered real-time data, generate compressed real-time data and add it to the multi-source climate data. Identify which data in the filtered real-time data belongs to low-frequency access data (for example, long-term archived data or data not often used for daily analysis), and apply a compression algorithm to compress these data, including lossless compression (such as ZIP, LZMA). The compressed data will take up less storage space and be added back to the multi-source climate data set, optimizing the storage resource utilization of the system.
[0025] Step S200, extracting elevation information, slope information and landform information of different regions in the Third Pole region to obtain the terrain features of the Third Pole region. Extracting elevation information, slope information and landform information of different regions in the Third Pole region. Specifically, elevation information refers to the vertical distance of a point on the surface relative to the sea level. In the Third Pole region, the extraction of elevation information mainly relies on satellite remote sensing data (such as SRTM elevation data, ASTER GDEM, etc.) and ground mapping data. Through these data, a high-precision digital elevation model (DEM) can be generated to analyze the ups and downs and changing trends of the terrain in the region and obtain the elevation information of the Third Pole region; slope information reflects the degree of inclination of the surface in the horizontal direction and is an important parameter in terrain analysis. In the Third Pole region, the extraction of slope information is usually based on DEM data. Through DEM data, the slope value of each point on the surface can be calculated to analyze the steepness and changing characteristics of the terrain in the region; landform information refers to the comprehensive expression of surface morphology and characteristics, including various types. In the Third Pole region, DEM data is used to analyze characteristic parameters such as surface undulation and cut degree, and remote sensing image data is combined to identify surface cover types (such as vegetation, water bodies, bare soil, etc.) and geological structure characteristics. By comparing and analyzing the landform characteristics of different regions, different landform types are divided, and their causes and evolution processes are explored, such as periglacial landforms, alpine meadows, desertified land, and other types. By extracting the elevation information, slope information and landform information of different areas in the Third Pole region, the terrain characteristics of the Third Pole region are obtained.
[0026] Step S300, weighted fusion of the multi-source climate data is performed based on the data fusion strategy of the terrain features to generate climate monitoring data. Data fusion refers to the comprehensive processing of data from different sources, different formats, and different time or spatial resolutions to generate more comprehensive and accurate information. In the field of climate monitoring, multi-source climate data includes ground observation station data, satellite remote sensing data, meteorological model prediction data, etc. Specifically, quality control is performed on multi-source climate data, including denoising, outlier detection, data correction, etc., and terrain feature data is extracted and sorted to ensure consistency with climate data in time and space. Then, the impact mechanism of terrain features on climate elements is analyzed, such as the impact of elevation on temperature, the impact of slope on precipitation distribution, etc., and climate data from different sources are allocated according to terrain features. Different weights. During the fusion process, the climate data of each source are weighted and summed according to their weights to generate fused climate monitoring data, thereby improving the accuracy and reliability of climate monitoring data. In the climate monitoring of the Third Pole region, ground observation station data and satellite remote sensing data are weightedly fused based on terrain features (such as elevation, slope, and landform type). For example, in high-altitude areas, since the temperature decreases with increasing altitude, higher weights can be assigned to satellite remote sensing data in high-altitude areas when fusing temperature data; and in areas with larger slopes, since precipitation distribution may be affected by terrain, the weight allocation of slope factors can be considered when fusing precipitation data.
[0027] In a possible implementation, step S300 further includes step S310, using a machine learning model to identify the correlation between the multi-source climate data and establish a preliminary fusion strategy for different time nodes. Use a machine learning model (such as random forest, support vector machine, cluster analysis, etc.) to analyze and identify the correlation between multi-source climate data. For example, the model may identify the correlation between ground temperature data and satellite measurement data, or similar climate patterns between different meteorological stations within a certain period of time. Based on these analysis results, a preliminary fusion strategy will be established for different time nodes. The preliminary fusion strategy defines how to combine data from different data sources at different time points to provide consistent monitoring results.
[0028] Step S320, adaptively adjust the weight of the data source in the preliminary fusion strategy according to the terrain features to generate fusion weights. Based on the preliminary fusion strategy, the weight of the data source is adjusted according to the extracted terrain features (such as elevation, slope, etc.). For example, in mountainous areas, ground observation data may be more reliable than satellite data, so it is necessary to increase the weight of ground observation data. Then, the adaptive algorithm is used to dynamically adjust these weights according to the actual terrain conditions to ensure that the fusion results can be more accurate under different terrain conditions, and finally generate fusion weights.
[0029] Step S330: dynamically adjust the preliminary fusion strategy using a time series analysis method to generate a data fusion strategy. Analyze the changing trend of climate data over time based on a time series analysis method (such as ARIMA model, LSTM, etc.), and dynamically adjust the preliminary fusion strategy based on the analysis results. For example, in a specific season, the reliability of certain data sources may increase or decrease. The fusion strategy of these data sources can be adjusted according to the time series analysis results, thereby generating a more accurate data fusion strategy to cope with the periodic or trend changes in climate change.
[0030] Step S340, based on the data fusion strategy and the fusion weight, the preferred part of the multi-source climate data is combined to generate preliminary climate monitoring data. After the data fusion strategy and the fusion weight of each data source are determined, the optimal data part is selected from different data sources according to these strategies and weights. These preferred parts may be those verified, high-quality, and most relevant data. By combining these data, preliminary climate monitoring data is generated.
[0031] Step S350, verify the accuracy of the preliminary climate monitoring data, adjust the data fusion strategy according to the accuracy verification result, and fuse to obtain the climate monitoring data. The accuracy verification of the preliminary climate monitoring data usually includes comparison with known high-precision data or actual observation results. If it is found that the preliminary data deviates from the actual situation, the data fusion strategy is adjusted according to the verification result, such as adjusting the weight of the data source or recombining the data. Finally, the optimized fusion strategy will be used to fuse and generate the final high-precision climate monitoring data.
[0032] Step S400, perform real-time quality assessment on the climate monitoring data, obtain abnormal data, use the neighborhood site data and historical climate monitoring data to correct the abnormal data, and generate corrected data. Performing real-time quality assessment on climate monitoring data refers to real-time monitoring and assessment of the quality of climate monitoring data, identifying abnormal conditions in the data, such as extreme values, noise, missing values, etc. Specifically, continuous quality checks are performed on the received climate monitoring data, and pre-set quality assessment indicators (such as data integrity, spatiotemporal consistency, etc.) are used to identify and mark abnormal points in the data. These abnormal points may be due to equipment failure, sensor error, extreme weather or other reasons. Data anomalies. During the quality assessment process, data points identified as not meeting quality standards will be marked as abnormal data. These abnormal data may deviate from normal climate patterns and appear as unreasonable extreme values or sudden changes. Neighborhood site data refers to data collected by other meteorological sites with similar geographical locations. Since the climate conditions in neighboring areas are usually similar, the data from these sites can be used to infer the reasonable range of abnormal data. Historical climate monitoring data refers to climate data under the same or similar conditions in the past. In climate monitoring in the Third Pole region, snow accumulation and permafrost are often used to monitor the climate of the surrounding areas. Regional data is extremely important. At the same time, the data in these areas are easily affected by weather changes and become abnormal. Through the method of neighborhood analysis, the data and historical trends of neighboring stations are used to supplement and correct the data in these areas in real time to ensure the accuracy and continuity of the data. In addition, when extreme weather events occur, the system will dynamically adjust the data sampling frequency, centrally process key data, and use models to supplement the reduced data volume to ensure the accuracy and reliability of the overall data processing. For example, if the data of a certain monitoring point is abnormal, the reasonable correction value of the abnormal value is inferred by referring to the data of the monitoring point under similar conditions in history, and the neighborhood station data and historical climate monitoring data are used for correction, that is, the abnormal data points are corrected by interpolation or model calculation through the neighborhood station data and historical climate data to make them closer to the actual situation. The corrected data will replace the original abnormal data to generate more accurate and continuous climate monitoring data, and finally obtain corrected climate monitoring data that meets quality requirements, that is, generate corrected data.
[0033] In a possible implementation, step S400 further includes step S410, setting a quality assessment index system, which includes a data integrity index, a spatiotemporal consistency index, and an outlier detection index. The data integrity index is used to evaluate the integrity of the data, such as whether the data is missing, whether there are null values or invalid values; the spatiotemporal consistency index is used to check the consistency of the data in the time and space dimensions, for example, whether there are mutations in the data of consecutive time points, whether the data of adjacent sites in space have reasonable correlation, which helps to identify anomalies caused by sensor failure or data transmission errors; the outlier detection index is used to identify abnormal data points that are beyond the normal range, for example, the temperature value suddenly soars to an unreasonable high temperature, or the rainfall changes extremely in a short period of time; the quality assessment index system is set based on the data integrity index, the spatiotemporal consistency index, and the outlier detection index.
[0034] Step S420: Monitor the climate monitoring data in real time based on the quality assessment index system and identify the abnormal data. Monitor the climate monitoring data in real time and evaluate the data according to the quality assessment index. For example, whenever new monitoring data arrives, check its integrity, spatiotemporal consistency with previous and subsequent data, and whether there are abnormal values. If a data point does not meet the set quality standards (for example, data is missing, spatiotemporal inconsistent, or detected as an abnormal value), mark it as abnormal data.
[0035] Step S430, using the climate monitoring model constructed based on the historical climate monitoring data to analyze the abnormal data and obtain abnormal analysis results. Based on a large amount of historical climate monitoring data, a climate monitoring model is constructed to predict or simulate normal climate data behavior, and then the identified abnormal data is input into the climate monitoring model, and the difference between the abnormal data and the normal value predicted by the model is compared, and the abnormal analysis results are obtained according to the climate monitoring model analysis, that is, it is determined whether the abnormal data is caused by equipment failure, data transmission error, extreme weather events, or other reasons.
[0036] Step S440, the abnormal data is corrected according to the abnormal analysis result by using the neighboring site data to generate the corrected data. The abnormal data is corrected by using the data of other sites with similar geographical locations to make it closer to the true value, so as to ensure the integrity and accuracy of the data set. Specifically, according to the abnormal analysis result, a suitable correction method is selected. For example, if the abnormality is due to data loss caused by sensor failure, spatial interpolation may be required. If it is an instantaneous abnormal peak, smoothing may be required. Then, the abnormal data is adjusted by using the data of the neighboring sites. For example, a weighted average method is used to calculate a substitute value based on the data of the neighboring sites, or a more reasonable value is predicted through a time series model. The corrected data replaces the original abnormal data to form a new, corrected climate monitoring data set.
[0037] In a possible implementation, step S440 further includes step S441, extracting the neighboring site data to calculate the correction value of the abnormal analysis result to generate the correction value. Relevant data is extracted from other meteorological sites close to the location of the abnormal data, and the correction value is calculated using statistical methods (such as weighted average, linear regression) or machine learning models (such as KNN, nearest neighbor interpolation). For example, if the temperature data of a certain site is abnormal, the system will refer to the temperature data of the surrounding sites to calculate a reasonable correction value, thereby ensuring the accuracy and continuity of the data.
[0038] Step S420, interpolate and reconstruct the abnormal data for adjacent points based on time and space to generate supplementary data. If the abnormal data is missing or abnormal in the time series, interpolate based on the data of the previous and next time points. For example, if the temperature data of a certain time point is lost, use the temperature data of the previous and next time points to interpolate and reconstruct the missing temperature value; for abnormal data in spatial distribution, interpolate using the spatial data of adjacent sites, for example, use the data of geographically adjacent sites to estimate the value of the abnormal site to ensure that the site data is consistent with the data of the surrounding sites in space; combine time and space interpolation technology to reconstruct the abnormal data and generate complete data points, that is, fill in the blank part of the data, or replace those modified abnormal data, and finally generate supplementary data.
[0039] Step S430, perform a spatiotemporal consistency check on the supplemented data. If the consistency check result meets the consistency check threshold, add the supplemented data to the corrected data. The consistency check threshold is a preset consistency check threshold. Only when the check result of the supplemented data is within the threshold range, the data is considered qualified. Specifically, based on time series analysis and space series analysis, the supplemented data are checked to ensure that they are consistent with adjacent data points in time and space dimensions, that is, to check whether the supplemented data is consistent with the adjacent actual observation data or historical data in terms of trend and range. If the supplemented data passes the consistency check, the supplemented data is added to the corrected climate monitoring data set to replace the original abnormal data or supplement the missing data to ensure the integrity and consistency of the entire data set.
[0040] Step S500: Use a climate change prediction model to perform climate change prediction on the modified data to generate a predicted climate change result.
[0041] Furthermore, the step S500 of using the climate change prediction model to predict climate change for the corrected data to generate predicted climate change results also includes: step S510, extracting historical climate change data, historical glacier retreat data and historical permafrost thaw data based on historical corrected data; step S520, training and feedback training the deep learning model based on the historical climate change data, historical glacier retreat data and historical permafrost thaw data to generate the climate change prediction model, wherein the climate change prediction model includes a first prediction model and a second prediction model; step S530, inputting the corrected data into the first prediction model to obtain glacier retreat prediction data and permafrost thaw prediction data, and inputting the glacier retreat prediction data and permafrost thaw prediction data into the second prediction model; step S540, adding the predicted temperature parameters, predicted precipitation parameters, predicted wind speed parameters and predicted snow depth parameters obtained by the second prediction model to the predicted climate change results.
[0042] Preferably, historical corrected data refers to a reliable data set obtained after quality control and anomaly correction of historical climate data, which may come from ground-based meteorological observation stations, satellite remote sensing data, climate model outputs, etc. over the past few decades or longer, and have undergone rigorous quality assessment and correction processing, from which historical climate change data, historical glacier retreat data, and historical permafrost thaw data are extracted, wherein historical climate change data refers to climate data observed within a certain time range in the past, including parameters such as temperature, precipitation, and wind speed; historical glacier retreat data records changes in glaciers over a certain period in the past, such as a reduction in glacier area, changes in thickness, etc., reflecting the dynamic changes of glaciers; historical permafrost thaw data refers to the historical thawing of permafrost, including thawing depth, area changes, etc.
[0043] Preferably, a prediction model is constructed based on a deep learning model (such as a convolutional neural network (CNN), a long short-term memory network (LSTM), etc.), and then the historical climate change data, the historical glacier retreat data, and the historical permafrost thawing data are standardized and normalized, and the processed data is used as input data and input into the constructed prediction model for preliminary training. The prediction model learns the patterns and trends in the data through repeated iterations, and adjusts internal parameters to improve the accuracy of the prediction. The model after preliminary training is then evaluated to analyze its prediction accuracy on historical data. According to the evaluation results, the parameters, number of layers or structure of the model are adjusted, and further training (i.e., feedback training) is performed to correct the model. The shortcomings of the model in the initial training are improved, and its prediction accuracy is improved. Finally, a climate change prediction model that can accurately predict future climate change, glacier retreat and permafrost thawing is generated, including a first prediction model and a second prediction model. The first prediction model is used to predict glacier retreat and permafrost thawing. By learning the patterns in historical glacier and permafrost data, the prediction data of glacier retreat and permafrost thawing in the future are generated; the second prediction model uses the output of the first prediction model (i.e., glacier retreat prediction data and permafrost thawing prediction data) to further predict related climate parameters, such as future temperature, precipitation, wind speed and snow depth, and can combine the changes in glaciers and permafrost to give more comprehensive climate prediction results.
[0044] Preferably, the generated corrected data are input into the first prediction model of the climate change prediction model to generate glacier retreat prediction data and permafrost thawing prediction data, which reflect the dynamic change trend of glaciers and permafrost in the future. The glacier retreat and permafrost thawing prediction data generated by the first prediction model are then input into the second prediction model. The second model will further predict key climate parameters such as future temperature, precipitation, wind speed and snow depth based on these input data, and then add the obtained predicted temperature parameters, predicted precipitation parameters, predicted wind speed parameters and predicted snow depth parameters to the predicted climate change results, describing in detail the climate change trends in the future and the dynamic changes of glaciers and permafrost, reflecting the overall picture of future climate change in the Third Pole region.
[0045] In a possible implementation, the big data method for monitoring climate change in the Third Pole region also includes, step S591, allocating computing tasks between local processing and cloud analysis to obtain preliminary allocation results. The current computing tasks are classified. For example, tasks with high real-time requirements can be processed locally first, while large analysis tasks that require a large amount of computing resources can be allocated to the cloud. Specifically, according to factors such as the real-time nature of the task, computing resource requirements, and data volume, a preliminary task allocation strategy is automatically generated. For example, edge devices (such as meteorological stations in the Third Pole region) can process some preliminary analysis of real-time monitoring data, while more complex climate model calculations can be completed on cloud servers. Tasks are allocated to local and cloud servers according to the above strategy, and preliminary task allocation results are generated to clarify which computing tasks will be processed locally and which will be handed over to cloud analysis to optimize resource utilization and improve computing efficiency.
[0046] Step S592, introduce a dynamic resource scheduling mechanism, adjust the preliminary allocation result according to the real-time climate conditions, and obtain the allocation result. Introduce a dynamic resource scheduling mechanism, wherein the dynamic resource scheduling mechanism refers to a mechanism that adjusts the allocation and use of computing resources in real time according to actual needs and environmental changes. In a complex computing environment, such as a climate monitoring and analysis system, the demand for computing tasks may change significantly over time or external conditions (such as climate change). The dynamic resource scheduling mechanism can flexibly adjust the allocation of resources to ensure that the system can operate efficiently and stably under various circumstances. Specifically, the current climate conditions are continuously monitored and their impact on computing tasks is evaluated. For example, extreme weather such as blizzards may increase the urgency of certain key tasks, such as the calculation of meteorological warnings. The preliminary allocated tasks are dynamically adjusted based on real-time climate monitoring data. For example, if a blizzard is detected to be approaching, the local computing resource allocation for the meteorological warning task is increased, and the computing power for related data analysis tasks is increased in the cloud to ensure that these tasks can be completed quickly, and the final task allocation result is generated to ensure that the system can flexibly respond to real-time changing climate conditions, especially in extreme weather, and computing resources are optimally configured.
[0047] Step S593, the data transmission and processing process of the allocation result is monitored in real time through blockchain technology. If the monitoring verification result meets the integrity threshold and security threshold, the computing task is executed. In the process of task allocation and execution, blockchain technology is used to record and monitor the data transmission and processing of each step, and the integrity and security of data transmission are verified through blockchain, for example, to check whether the data is tampered with or lost during transmission, or whether there is unauthorized access during processing. If the monitoring verification result meets the integrity threshold and security threshold, it indicates that the data transmission and processing process is safe and reliable, and the computing task can be executed; if it does not meet the threshold, the task execution is stopped or an alarm is issued. Through real-time monitoring and verification of blockchain technology, the data security and integrity of the computing task during the allocation and execution process are ensured, so that the task can be completed reliably.
[0048] Step S600, the predicted climate change results are displayed through a visualization window, and a climate report is generated. Using visualization technology, complex climate prediction data are displayed in a graphical manner, so that users can intuitively understand future climate change trends. For example, the temperature changes and precipitation distribution of a certain area in the future are displayed on a map, or the dynamic process of glacier retreat and permafrost thawing is displayed with a dynamic chart. The visualization window may also include interactive functions, such as zooming, filtering, timeline sliding, etc. Users can view climate prediction results at different times, locations or conditions through these functions. For example, users can select an area on the map to view the temperature change trend of the area in the next few years. A climate report is generated based on the predicted climate change results, which describes in detail the future climate change trends and their possible impacts.
[0049] In the above, refer to Figure 1 The third polar region climate change monitoring big data method according to an embodiment of the present invention is described in detail. Figure 2 The third polar region climate change monitoring big data application platform according to an embodiment of the present invention is described.
[0050] The big data application platform for monitoring climate change in the Third Pole region according to the embodiment of the present invention is used to solve the technical problem that it is difficult to accurately and comprehensively capture the subtle differences and dynamic evolution of multi-source data in the Third Pole region, which leads to insufficient accuracy and reliability of climate change predictions in existing climate monitoring management, and achieves the technical effect of improving the accuracy and reliability of climate change predictions in the Third Pole region. The big data application platform for monitoring climate change in the Third Pole region includes: a multi-source climate data acquisition module 10, a Third Pole region terrain feature acquisition module 20, a climate monitoring data generation module 30, a correction data generation module 40, a predicted climate change result generation module 50, and a result visualization display module 60.
[0051] A multi-source climate data acquisition module 10 is used to collect real-time data from ground observation stations and satellite remote sensing data to obtain multi-source climate data; a third pole region terrain feature acquisition module 20 is used to extract elevation information, slope information and landform information of different regions in the third pole region to obtain the terrain features of the third pole region; a climate monitoring data generation module 30 is used to perform weighted fusion of the multi-source climate data based on the data fusion strategy of the terrain features to generate climate monitoring data; a correction data generation module 40 is used to perform real-time quality assessment of the climate monitoring data, obtain abnormal data, and use neighboring site data and historical climate monitoring data to correct the abnormal data to generate corrected data; a predicted climate change result generation module 50 is used to use a climate change prediction model to perform climate change prediction on the corrected data to generate predicted climate change results; a result visualization display module 60 is used to display the predicted climate change results through a visualization window and generate a climate report.
[0052] The specific configuration of the multi-source climate data acquisition module 10 will be described in detail below. The multi-source climate data acquisition module 10 further includes: performing format conversion of the real-time data to generate standardized real-time data; performing time and space alignment on the standardized real-time data to generate time-space consistent real-time data; filtering noise data and invalid data in the time-space consistent real-time data through signal processing to generate filtered real-time data; using a compression algorithm to compress and store low-frequency access data in the filtered real-time data, generating compressed real-time data and adding it to the multi-source climate data.
[0053] The specific configuration of the climate monitoring data generation module 30 will be described in detail below. The climate monitoring data generation module 30 further includes: using a machine learning model to identify the correlation between the multi-source climate data, and establishing a preliminary fusion strategy for different time nodes; adaptively adjusting the data source weights in the preliminary fusion strategy according to the terrain characteristics to generate fusion weights; dynamically adjusting the preliminary fusion strategy using a time series analysis method to generate a data fusion strategy; combining the preferred part of the multi-source climate data based on the data fusion strategy and the fusion weights to generate preliminary climate monitoring data; performing accuracy verification on the preliminary climate monitoring data, adjusting the data fusion strategy according to the accuracy verification results, and fusing to obtain the climate monitoring data.
[0054] The specific configuration of the correction data generation module 40 will be described in detail below. The correction data generation module 40 may further include: setting a quality assessment index system, the quality assessment index system including a data integrity index, a spatiotemporal consistency index and an outlier detection index; real-time monitoring of the climate monitoring data based on the quality assessment index system to identify the abnormal data; analyzing the abnormal data using a climate monitoring model constructed based on the historical climate monitoring data to obtain an abnormal analysis result; and correcting the abnormal data according to the abnormal analysis result using the neighborhood site data to generate the corrected data.
[0055] The specific configuration of the correction data generation module 40 will be described in detail below. The correction data generation module 40 may further include: extracting the neighborhood site data to calculate the correction value of the abnormal analysis result to generate the correction value; interpolating and reconstructing the abnormal data for adjacent points based on time and space to generate supplementary data; performing a spatiotemporal consistency check on the supplementary data, and if the consistency check result meets the consistency check threshold, adding the supplementary data to the correction data.
[0056] The specific configuration of the predicted climate change result generation module 50 will be described in detail below. The predicted climate change result generation module 50 further includes: extracting historical climate change data, historical glacier retreat data and historical frozen soil thawing data according to historical correction data; training and feedback training the deep learning model based on the historical climate change data, historical glacier retreat data and historical frozen soil thawing data to generate the climate change prediction model, wherein the climate change prediction model includes a first prediction model and a second prediction model; inputting the correction data into the first prediction model to obtain glacier retreat prediction data and frozen soil thawing prediction data, and inputting the glacier retreat prediction data and frozen soil thawing prediction data into the second prediction model; adding the predicted temperature parameters, predicted precipitation parameters, predicted wind speed parameters and predicted snow depth parameters obtained by the second prediction model to the predicted climate change results.
[0057] The specific configuration of the big data method and application platform for monitoring climate change in the Third Pole region will be described in detail below. The big data method and application platform for monitoring climate change in the Third Pole region further include: allocating computing tasks between local processing and cloud analysis to obtain preliminary allocation results; introducing a dynamic resource scheduling mechanism to adjust the preliminary allocation results according to real-time climate conditions to obtain allocation results; and monitoring the data transmission and processing of the allocation results in real time through blockchain technology. If the monitoring verification results meet the integrity threshold and security threshold, the computing tasks are executed.
[0058] The Third Pole region climate change monitoring big data application platform provided in the embodiment of the present invention can execute the Third Pole region climate change monitoring big data method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.
[0059] Although the present application makes various references to certain modules in the platform according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be implemented; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.
[0060] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.
Claims
1. A big data method for monitoring climate change in the Third Pole region, characterized in that: include: Collect real-time data from ground observation stations and satellite remote sensing data to obtain multi-source climate data; Extracting elevation information, slope information and landform information of different regions in the Third Pole region to obtain the topographic features of the Third Pole region; Perform weighted fusion on the multi-source climate data based on the data fusion strategy of the terrain features to generate climate monitoring data; Performing real-time quality assessment on the climate monitoring data to obtain abnormal data, and correcting the abnormal data using neighboring site data and historical climate monitoring data to generate corrected data, including: Setting a quality assessment indicator system, wherein the quality assessment indicator system includes a data integrity indicator, a spatiotemporal consistency indicator, and an outlier detection indicator; Based on the quality assessment indicator system, the climate monitoring data is monitored in real time to identify the abnormal data; Analyzing the abnormal data using a climate monitoring model constructed based on the historical climate monitoring data to obtain an abnormal analysis result; Correcting the abnormal data according to the abnormal analysis result using the neighboring site data to generate the corrected data; The climate change prediction model is used to predict the climate change of the modified data to generate predicted climate change results, including: Extract historical climate change data, historical glacier retreat data and historical permafrost thaw data based on historical revision data; Based on the historical climate change data, historical glacier retreat data and historical frozen soil thawing data, a deep learning model is trained and feedback trained to generate the climate change prediction model, wherein the climate change prediction model includes a first prediction model and a second prediction model; Inputting the corrected data into the first prediction model to obtain glacier retreat prediction data and frozen soil thawing prediction data, and inputting the glacier retreat prediction data and frozen soil thawing prediction data into the second prediction model; adding the predicted temperature parameter, the predicted precipitation parameter, the predicted wind speed parameter and the predicted snow depth parameter obtained by the second prediction model to the predicted climate change result; The predicted climate change results are displayed through a visualization window, and a climate report is generated.
2. The big data method for monitoring climate change in the third polar region according to claim 1, characterized in that: Collect real-time data from ground observation stations and satellite remote sensing data to obtain multi-source climate data, including: Performing format conversion of the real-time data to generate standardized real-time data; Aligning the standardized real-time data in time and space to generate time-space consistent real-time data; Filtering noise data and invalid data in the spatiotemporally consistent real-time data by signal processing to generate filtered real-time data; A compression algorithm is used to compress and store the low-frequency access data in the filtered real-time data, and compressed real-time data is generated and added to the multi-source climate data.
3. The big data method for monitoring climate change in the third polar region according to claim 1, characterized in that: The data fusion strategy based on the terrain features performs weighted fusion on the multi-source climate data to generate climate monitoring data, including: Using machine learning models to identify the correlation between the multi-source climate data and establish preliminary fusion strategies at different time nodes; Adaptively adjusting the data source weights in the preliminary fusion strategy according to the terrain features to generate fusion weights; Dynamically adjusting the preliminary fusion strategy using a time series analysis method to generate a data fusion strategy; Combining the preferred parts of the multi-source climate data based on the data fusion strategy and the fusion weights to generate preliminary climate monitoring data; The preliminary climate monitoring data is subjected to accuracy verification, the data fusion strategy is adjusted according to the accuracy verification result, and the climate monitoring data is obtained by fusion.
4. The big data method for monitoring climate change in the third polar region according to claim 1, characterized in that: The step of correcting the abnormal data according to the abnormal analysis result by using the neighboring site data to generate the corrected data includes: Extracting the neighboring site data, calculating a correction value for the abnormal analysis result, and generating a correction value; Interpolating and reconstructing the abnormal data for adjacent points based on time and space to generate supplementary data; Perform a spatiotemporal consistency check on the supplemented data, and if a consistency check result meets a consistency check threshold, add the supplemented data to the corrected data.
5. The big data method for monitoring climate change in the third polar region according to claim 1, characterized in that: Also includes: Allocate computing tasks between local processing and cloud analysis to obtain preliminary allocation results; Introducing a dynamic resource scheduling mechanism to adjust the preliminary allocation result according to real-time climate conditions to obtain an allocation result; The data transmission and processing process of the allocation result is monitored in real time through blockchain technology. If the monitoring verification result meets the integrity threshold and the security threshold, the computing task is executed.
6. The Third Pole Regional Climate Change Monitoring Big Data Application Platform is characterized by: The steps for implementing the big data method for monitoring climate change in the Third Polar Region as described in any one of claims 1 to 5 include: Multi-source climate data acquisition module, used to collect real-time data from ground observation stations and satellite remote sensing data to obtain multi-source climate data; A module for obtaining terrain features of the third pole region, used to extract elevation information, slope information and landform information of different regions in the third pole region, and obtain terrain features of the third pole region; A climate monitoring data generation module, used for performing weighted fusion on the multi-source climate data based on the data fusion strategy of the terrain characteristics to generate climate monitoring data; A correction data generation module is used to perform real-time quality assessment on the climate monitoring data, obtain abnormal data, correct the abnormal data using neighboring site data and historical climate monitoring data, and generate correction data; A predicted climate change result generating module is used to use a climate change prediction model to perform climate change prediction on the revised data and generate predicted climate change results; The result visualization display module is used to display the predicted climate change results through a visualization window and generate a climate report.
