Heterogeneous multi-source frozen circle big data exploration and analysis method and system based on space-time segmentation

Abstract

The invention discloses a heterogeneous multi-source frozen circle big data exploration and analysis method and system based on space-time segmentation, and belongs to the technical field of data analysis. The method comprises the following steps: collecting and preprocessing freezing ring data; dividing the research area into space-time grid units; performing heterogeneous data fusion on the frozen circle data in each time space grid unit; and identifying modes and anomalies in the space grid units based on the fused frozen circle data, and / or predicting the future change trend of the frozen circle. According to the invention, high-efficiency exploration and analysis of frozen circle big data are realized.

Description

Technical Field

[0001] The present invention relates to the field of data analysis technology, and specifically to a heterogeneous multi-source cryosphere big data exploration and analysis method and system based on time and space segmentation, which is suitable for processing and analyzing multi-source data such as temperature, precipitation, glacier thickness, and snow cover. Background Art

[0002] Cryosphere research requires processing and analyzing a large amount of heterogeneous data from different data sources, such as meteorological data, remote sensing data, geographic information data, etc. These data have significant differences in space and time. Traditional data analysis methods are difficult to effectively integrate and process these data, resulting in low accuracy and reliability of analysis results. Therefore, it is of great significance to propose a method that can effectively integrate and analyze heterogeneous multi-source data for cryosphere research. Summary of the invention

[0003] The present invention aims to solve the problems of data integration difficulties, low analysis efficiency and insufficient prediction accuracy in the prior art, and proposes a heterogeneous multi-source cryosphere big data exploration and analysis method and system based on spatiotemporal segmentation. The method constructs a spatiotemporal network by performing spatiotemporal segmentation and fusion on the data, thus realizing efficient exploration and analysis of cryosphere big data.

[0004] To achieve the above-mentioned object of the invention, the technical solution of the present invention includes the following contents.

[0005] A heterogeneous multi-source cryosphere big data exploration and analysis method based on time-space segmentation, the method comprising:

[0006] Collect and pre-process cryosphere data, including meteorological data, remote sensing data and geographic information data;

[0007] The study area is divided into space-time grid units, each of which contains cryosphere data within a specific time period and spatial range.

[0008] Fusion of heterogeneous cryosphere data within each spatial and temporal grid unit;

[0009] Identify patterns and anomalies in spatial and temporal grid cells based on fused cryosphere data, and / or predict future trends in the cryosphere.

[0010] Furthermore, the preprocessing of cryosphere data includes:

[0011] Performing data cleaning on the cryosphere data;

[0012] Filling missing values ​​in the cryosphere data using a mean-filling algorithm, a median-filling algorithm, an interpolation algorithm, or a machine learning-based prediction model;

[0013] Standardizing the cryosphere data to convert data of different magnitudes to the same standard; wherein the algorithm for standardizing the cryosphere data includes: a Z-score standardization algorithm, a minimum-maximum standardization algorithm, or a logarithmic transformation algorithm;

[0014] The cryosphere data are denoised using filters and smoothing techniques.

[0015] Furthermore, the fusion of heterogeneous data of cryosphere data in each space-time grid unit includes:

[0016] Align cryosphere data from different sources and types in terms of time, space and resolution to obtain spatially registered cryosphere data;

[0017] According to the fusion purpose and fusion level, the spatially registered cryosphere data are organically synthesized to obtain a representation or estimation that meets the research objectives; wherein the fusion level includes: pixel-level fusion, feature-level fusion and decision-level fusion.

[0018] Furthermore, the spatially registered cryosphere data are organically synthesized according to the fusion purpose and fusion level to obtain a representation or estimation that meets the research objectives, including:

[0019] The meteorological data and remote sensing data are fused and processed based on the data fusion algorithm to make up for the fact that meteorological data is difficult to depict the impact on fine scales; wherein the data fusion algorithm includes: principal component analysis, Bruettmann model, band ratio method, discrete wavelet transform and deep learning method;

[0020] and / or,

[0021] Based on the data fusion algorithm, geographic information data is fused with meteorological data / remote sensing data to provide geographic element information for meteorological data and remote sensing data.

[0022] Further, the identifying patterns and anomalies in the space-time grid cells based on the fused cryosphere data includes:

[0023] Select key factors from the fused cryosphere data; where the research target is glacier change or sea ice change, the key factors include: temperature, precipitation, sea ice thickness and glacier coverage;

[0024] Select a clustering algorithm based on the cryosphere data size and research objectives;

[0025] Based on the key factor clustering algorithm, the fused cryosphere data are clustered into multiple clusters, and the patterns and anomalies in the space-time grid cells are determined by analyzing the characteristics of each cluster.

[0026] Further, the identifying patterns and anomalies in the space-time grid cells based on the fused cryosphere data includes:

[0027] Based on the fused cryosphere data, the spatiotemporal sequence data of the research target is generated, and the key spatiotemporal features of the spatiotemporal sequence data are extracted; where the research target is the relationship between temperature rise and glacier retreat, the spatiotemporal sequence data includes: glacier area and sea ice concentration;

[0028] Selecting an association rule algorithm according to the research objective and the data scale of the fused cryosphere data, wherein the association rule algorithm includes: an Apriori algorithm or an FP-growth algorithm;

[0029] Applying the association rule algorithm based on the key spatiotemporal features to obtain patterns in the spatiotemporal grid cells whose occurrence times are greater than a set threshold;

[0030] The key spatiotemporal patterns are screened for the patterns in the spatiotemporal grid cells whose occurrence times are greater than a set threshold.

[0031] Furthermore, the prediction of future changes in the cryosphere based on the fused cryosphere data includes:

[0032] Generate training data set and test data set;

[0033] Selecting a spatiotemporal regression model based on the research objective; wherein the spatiotemporal regression model is constructed using a multivariate linear regression algorithm or a geographically weighted regression algorithm;

[0034] Defining dependent variables and independent variables of the spatiotemporal regression model; wherein the dependent variable includes: glacier area, and the independent variables include: temperature and precipitation;

[0035] estimating parameters of a spatiotemporal regression model using statistical methods on the training data set;

[0036] Validate and tune the parameters of the spatiotemporal regression model on the test dataset;

[0037] The fused cryosphere data are input into the adjusted spatiotemporal regression model to obtain the future change trend of the cryosphere.

[0038] Furthermore, the prediction of future changes in the cryosphere based on the fused cryosphere data includes:

[0039] Generate training data set and test data set;

[0040] Selecting a spatiotemporal neural network based on the research objective; wherein the spatiotemporal regression model is constructed using a convolutional neural network, a recurrent neural network algorithm, or a long short-term memory network;

[0041] Training the spatiotemporal neural network on the training data set;

[0042] Verify and tune the parameters of the spatiotemporal neural network on the test dataset;

[0043] The fused cryosphere data are input into the adjusted spatiotemporal neural network to obtain the future change trend of the cryosphere.

[0044] Further, after identifying patterns and anomalies in the space-time grid cells based on the fused cryosphere data, and / or predicting future trends of the cryosphere, the method further includes:

[0045] Display the distribution and changes of cryosphere data in space-time grid cells based on heat maps, space-time trajectory maps and three-dimensional maps;

[0046] and / or,

[0047] Based on the patterns and anomaly identification results in the space-time grid units, as well as the future changes in the cryosphere, scientific decision support is provided; among which, the scientific decisions include: glacier retreat prediction, snow cover change trend, and assessment of the impact of climate change on the cryosphere.

[0048] A heterogeneous multi-source cryosphere big data exploration and analysis system based on time and space segmentation, the system comprising:

[0049] A preprocessing module, used for collecting and preprocessing cryosphere data, wherein the cryosphere data includes: meteorological data, remote sensing data and geographic information data;

[0050] The space-time partitioning module is used to divide the study area into space-time grid cells; each space-time grid cell contains cryosphere data within a specific time period and spatial range;

[0051] Heterogeneous data fusion module, used to fuse heterogeneous cryosphere data in each space-time grid unit;

[0052] The spatiotemporal analysis and modeling module is used to identify patterns and anomalies in spatiotemporal grid cells based on the fused cryosphere data and / or predict future changes in the cryosphere.

[0053] Compared with the prior art, the present invention has at least the following beneficial effects.

[0054] 1) Spatiotemporal segmentation and grid management: Through spatiotemporal segmentation and grid management, complex heterogeneous data can be systematized and structured to facilitate unified processing and analysis.

[0055] 2) Heterogeneous data fusion: Interpolation and resampling techniques are used to achieve seamless fusion of data from different sources and types, ensuring data consistency and integrity.

[0056] 3) Spatiotemporal pattern recognition and prediction: Combining cluster analysis and spatiotemporal prediction models, we can deeply explore the spatiotemporal variation patterns and future trends of cryosphere data and provide data-driven scientific predictions.

[0057] 4) Visual decision support: Display data analysis results through a variety of visualization methods to improve the intuitiveness and comprehensibility of data and support scientific decision-making. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0059] The present invention is further described in detail below in conjunction with specific embodiments. The given examples are only for illustrating the present invention, but not for limiting the scope of the present invention.

[0060] The technical solution of the present invention is as follows Figure 1 As shown, it includes the following five steps:

[0061] The first step is data collection and preprocessing

[0062] Data sources: Cryosphere data comes from a variety of sources, including meteorological data (such as temperature and precipitation), remote sensing data (such as ice

[0063] The present invention collects data on temperature, precipitation, glacier thickness, snow cover, and geographic information data (such as topography and land use), which are of great significance for the study of climate change, environmental monitoring, and natural disaster warning. This embodiment collects data on temperature, precipitation, glacier thickness, snow cover, etc. from different channels such as meteorological bureaus, remote sensing satellites, and geographic information systems.

[0064] Data preprocessing: Clean the collected data, fill in missing values, standardize the data, and remove noise to ensure the data is

[0065] According to quality and consistency.

[0066] ■Data cleaning: This mainly includes deleting irrelevant data, correcting erroneous data, and identifying and processing outliers. In cryosphere data, this involves correcting data points that deviate significantly from the norm based on geographic or seasonal factors, or deleting erroneous data caused by equipment failure.

[0067] ■ Missing value filling: Cryosphere data often have missing values, especially under extreme climate conditions. Various techniques are used to fill these missing values:

[0068] ◆Mean filling: For relatively consistent data, use the overall or group mean for filling.

[0069] ◆Median filling: For skewed distributed data, use the median to fill.

[0070] ◆Interpolation method: Use linear interpolation or polynomial interpolation of time series data.

[0071] ◆Advanced methods: Use machine learning based prediction models such as K-nearest neighbors (KNN) to predict missing values.

[0072] ■Data standardization: Data standardization is a key step in processing data from different sources and scales. Its purpose is to convert data of different magnitudes to the same standard. The present invention adopts data standardization methods such as Z-score standardization, minimum-maximum standardization, and logarithmic transformation.

[0073] ■ Denoising: Denoising is an important step to improve data quality, especially in cryosphere data containing a lot of background noise. Filters (high-pass, low-pass or band-pass filters), smoothing techniques, etc. are used in the present invention.

[0074] Step 2: Spatiotemporal segmentation

[0075] ●Time segmentation: split the data by time dimension, such as year, season, month, day, to capture and analyze the patterns of time changes.

[0076] Spatial segmentation: Split the data by spatial dimensions, such as by administrative regions or geographic units (such as river basins or mountain ranges) to capture spatial variation patterns.

[0077] ● Space-time gridding: Divide the study area into space-time grid units, each unit contains multi-source data within a specific time period and spatial range, to achieve systematic and structured management of space-time data.

[0078] The third step is to integrate heterogeneous data.

[0079] Data alignment: Align data from different sources and types in time and space to ensure data consistency within the same space-time grid.

[0080] Data Fusion:

[0081] Data alignment: Align data from different sources and types in time, space, and resolution to ensure data consistency within the same space-time grid.

[0082] ■ Time alignment: cryosphere data may come from different sensors and observation stations, and the collection time of these data may be expressed differently. In the present invention, all data points are converted to the same time standard (UTC).

[0083] ■Spatial alignment: To ensure that the datasets match accurately in geographical location, all datasets are transformed into the same map projection system and coordinate system (WGS 84).

[0084] ■ Resolution matching: Different data sources may have different temporal and spatial resolutions. Resolution matching involves adjusting the granularity of the data so that they can be effectively compared and integrated.

[0085] ◆ Downsampling: Reducing the resolution of high-resolution data to match lower-resolution data.

[0086] ◆Upsampling: Use interpolation techniques such as bilinear interpolation or cubic interpolation to increase the resolution of low-resolution data.

[0087] Data fusion: Interpolation, resampling and other methods are used to fuse heterogeneous data into a unified space-time grid for unified processing and analysis. Specifically, the present invention selects a suitable fusion algorithm according to the fusion purpose and fusion level, and organically synthesizes the spatially registered remote sensing data to obtain a more accurate representation or estimation that meets the target. The levels of data fusion include pixel-level fusion, feature-level fusion and decision-level fusion. Data fusion algorithms include principal component analysis (PCA), Brovey Transform, Gram-Schmidt Spectral Sharpening, discrete wavelet transform (DWT) and deep learning methods. Meteorological data and remote sensing data are fused to compensate for the difficulty of meteorological data in depicting fine-scale effects; geographic information data is fused with meteorological data / remote sensing data to provide richer geographic element information for meteorological and remote sensing data, so as to facilitate data extraction and intelligent mining.

[0088] Step 4: Spatiotemporal analysis and modeling

[0089] Spatiotemporal pattern recognition: Use cluster analysis, spatiotemporal association rule mining and other methods to identify patterns and anomalies in spatiotemporal networks

[0090] ● Spatiotemporal prediction model: Build spatiotemporal prediction models (such as spatiotemporal regression models, spatiotemporal neural networks) to predict future changes in the cryosphere and provide data-driven scientific predictions.

[0091] ● Spatiotemporal pattern recognition: Use cluster analysis, spatiotemporal association rule mining and other methods to identify patterns and anomalies in the spatiotemporal network and reveal the spatiotemporal variation of the cryosphere. For example, abnormal phenomena such as rapidly retreating glaciers and abnormally melting sea ice areas; patterns of seasonal melting and refreezing of glaciers and sea ice, trends such as glacier retreat, ice sheet thinning and annual reduction in sea ice area; glacier flow patterns that reflect the dynamic processes inside the ice body and the basement through glacier flow, etc.

[0092] ■ Cluster analysis steps

[0093] ◆Feature selection: In the data processed by the previous steps, select the key factors that affect glacier changes or sea ice changes, such as temperature, precipitation, sea ice thickness, glacier coverage area, etc.

[0094] ◆Algorithm selection: Choose an appropriate clustering algorithm based on the data size and research objectives. K-means is suitable for larger data sets, but requires a predefined number of clusters; DBSCAN, a density-based clustering algorithm, can automatically determine the number of clusters and is suitable for irregularly shaped spatial data.

[0095] Clustering execution: Apply clustering algorithms to cluster the fused cryosphere data into multiple clusters; analyze the characteristics of each cluster to determine which clusters show abnormal or significant change trends.

[0096] ◆Result interpretation: Identify areas of significant change, such as rapidly retreating glaciers and abnormally melting sea ice areas.

[0097] ■Application of spatiotemporal association rules, taking the association between temperature rise and glacier retreat as an example

[0098] ◆Define and prepare spatiotemporal datasets: Extract key spatiotemporal features such as temperature, glacier area, and sea ice thickness from time series data.

[0099] ◆Select association rule algorithms: Apriori: mines frequent item sets, suitable for small but complex association analysis; FP-growth: suitable for large-scale data sets and more efficient.

[0100] ◆Perform association rule mining: Use the above algorithm to find frequently occurring patterns, such as the change in glacier area under a specific temperature; evaluate the support and confidence of the rules to ensure the reliability of the rules.

[0101] ◆Analyze and apply discovered rules: Identify key spatiotemporal patterns, such as the acceleration of sea ice melt in specific seasons; use the discovered rules to predict future glacier and sea ice changes.

[0102] ● Spatiotemporal prediction model: Build spatiotemporal prediction models (such as spatiotemporal regression models, spatiotemporal neural networks) to predict future changes in the cryosphere and provide data-driven scientific predictions.

[0103] ■Spatio-temporal Regression Models: Predict the future state of the cryosphere by integrating changes in temporal and spatial data.

[0104] ◆Related algorithms: Multiple linear regression: Uses historical data (such as temperature, snowfall, ocean temperature, etc.) to predict changes in glaciers or sea ice. Geographically Weighted Regression (GWR): Takes into account the geographic location of the data, allowing model parameters to vary spatially, thereby more accurately simulating geographic processes.

[0105] ◆Application steps

[0106] Data collection and preprocessing: Collect relevant climate and geographic data, and perform necessary cleaning and standardization.

[0107] Model building: Select an appropriate regression model and define dependent variables (such as glacier area) and independent variables (such as temperature and precipitation).

[0108] ●Parameter estimation: Use statistical methods (such as least squares) to estimate model parameters.

[0109] ●Verification and adjustment: Test the accuracy of the model through methods such as cross-validation and adjust the model as needed.

[0110] ■Spatio-temporal Neural Networks: Spatio-temporal neural network models can capture the complex dynamic relationship between time series and spatial data.

[0111] ◆Related algorithms: Convolutional Neural Network (CNN): Suitable for feature extraction of spatial data. Recurrent Neural Network (RNN) and its variants such as LSTM (Long Short-Term Memory Network): Able to process sequence data and capture dynamic changes in time.

[0112] ◆Application steps

[0113] Data integration: Formatting spatiotemporal data to fit the input requirements of the neural network model.

[0114] Network design: Design a suitable network structure, including convolutional layers, recurrent layers, fully connected layers, etc.

[0115] Training model: Use a large amount of historical data to train the network and adjust the network parameters.

[0116] Performance evaluation and optimization: Evaluate the predictive ability of the model and make necessary adjustments.

[0117] Step 5: Visualization and decision support

[0118] Data visualization: Display the data distribution in the space-time grid through visualization methods such as heat maps, space-time trajectory maps, and three-dimensional maps.

[0119] distribution and changes, making data understanding and interpretation more intuitive.

[0120] Decision support: Based on the analysis results, scientific decision support is provided, such as glacier retreat prediction, snow cover change trend, and assessment of the impact of climate change on the cryosphere, to help decision makers formulate reasonable response strategies.

[0121] In summary, the present invention aims to effectively integrate and analyze a large amount of heterogeneous data from different sources and types in the cryosphere. This method constructs a space-time network through space-time segmentation and data fusion, realizes efficient exploration and analysis of cryosphere big data, thereby improving the accuracy of data processing and analysis efficiency, and revealing the spatiotemporal changes of the cryosphere.

[0122] The above embodiments are only for illustrating the technical concept and features of the present invention, and their purpose is to enable people familiar with the technology in this field to understand the content of the present invention and implement it, and they cannot be used to limit the protection scope of the present invention. Any equivalent changes or modifications made according to the spirit of the present invention should be included in the protection scope of the present invention.

Claims

1. A heterogeneous multi-source cryosphere big data exploration and analysis method based on spatiotemporal segmentation, characterized in that: The method comprises: collecting and preprocessing cryosphere data, wherein the cryosphere data comprises: meteorological data, remote sensing data and geographic information data; The study area is divided into space-time grid units, each of which contains cryosphere data within a specific time period and spatial range. Fusion of heterogeneous cryosphere data within each spatial and temporal grid unit; Identify patterns and anomalies in spatial and temporal grid cells based on fused cryosphere data, and / or predict future trends in the cryosphere.

2. The method according to claim 1, characterized in that: The preprocessing of cryosphere data comprises: Performing data cleaning on the cryosphere data; Filling missing values ​​in the cryosphere data using a mean-filling algorithm, a median-filling algorithm, an interpolation algorithm, or a machine learning-based prediction model; Standardizing the cryosphere data to convert data of different magnitudes to the same standard; wherein the algorithm for standardizing the cryosphere data includes: a Z-score standardization algorithm, a minimum-maximum standardization algorithm, or a logarithmic transformation algorithm; The cryosphere data are denoised using filters and smoothing techniques.

3. The method according to claim 1, characterized in that The fusion of heterogeneous cryosphere data in each space-time grid unit includes: Align cryosphere data from different sources and types in terms of time, space and resolution to obtain spatially registered cryosphere data; According to the fusion purpose and fusion level, the spatially registered cryosphere data are organically synthesized to obtain a representation or estimation that meets the research objectives; wherein the fusion level includes: pixel-level fusion, feature-level fusion and decision-level fusion.

4. The method according to claim 3, characterized in that According to the fusion purpose and fusion level, the spatially registered cryosphere data are organically synthesized to obtain a representation or estimation that meets the research objectives, including: The meteorological data and remote sensing data are fused and processed based on the data fusion algorithm to make up for the fact that meteorological data is difficult to depict the impact on fine scales; wherein the data fusion algorithm includes: principal component analysis, Bruettmann model, band ratio method, discrete wavelet transform and deep learning method; and / or, Based on the data fusion algorithm, geographic information data is fused with meteorological data / remote sensing data to provide geographic element information for meteorological data and remote sensing data.

5. The method according to claim 1, characterized in that The identifying patterns and anomalies in the space-time grid cells based on the fused cryosphere data includes: Select key factors from the fused cryosphere data; where the research target is glacier change or sea ice change, the key factors include: temperature, precipitation, sea ice thickness and glacier coverage; Select a clustering algorithm based on the cryosphere data size and research objectives; Based on the key factor clustering algorithm, the fused cryosphere data are clustered into multiple clusters, and the patterns and anomalies in the space-time grid cells are determined by analyzing the characteristics of each cluster.

6. The method according to claim 1, characterized in that The identifying patterns and anomalies in the space-time grid cells based on the fused cryosphere data includes: Based on the fused cryosphere data, the spatiotemporal sequence data of the research target is generated, and the key spatiotemporal features of the spatiotemporal sequence data are extracted; where the research target is the relationship between temperature rise and glacier retreat, the spatiotemporal sequence data includes: glacier area and sea ice concentration; Selecting an association rule algorithm according to the research objective and the data scale of the fused cryosphere data, wherein the association rule algorithm includes: an Apriori algorithm or an FP-growth algorithm; Applying the association rule algorithm based on the key spatiotemporal features to obtain patterns in the spatiotemporal grid cells whose occurrence times are greater than a set threshold; The key spatiotemporal patterns are screened for the patterns in the spatiotemporal grid cells whose occurrence times are greater than a set threshold.

7. The method according to claim 1, characterized in that The prediction of future changes in the cryosphere based on the fused cryosphere data includes: Generate training data set and test data set; Selecting a spatiotemporal regression model based on the research objective; wherein the spatiotemporal regression model is constructed using a multivariate linear regression algorithm or a geographically weighted regression algorithm; Defining dependent variables and independent variables of the spatiotemporal regression model; wherein the dependent variable includes: glacier area, and the independent variables include: temperature and precipitation; estimating parameters of a spatiotemporal regression model using statistical methods on the training data set; Validate and tune the parameters of the spatiotemporal regression model on the test dataset; The fused cryosphere data are input into the adjusted spatiotemporal regression model to obtain the future change trend of the cryosphere.

8. The method according to claim 1, characterized in that: The prediction of future changes in the cryosphere based on the fused cryosphere data includes: Generate training data set and test data set; Selecting a spatiotemporal neural network based on the research objective; wherein the spatiotemporal regression model is constructed using a convolutional neural network, a recurrent neural network algorithm, or a long short-term memory network; Training the spatiotemporal neural network on the training data set; Verify and tune the parameters of the spatiotemporal neural network on the test dataset; The fused cryosphere data are input into the adjusted spatiotemporal neural network to obtain the future change trend of the cryosphere.

9. The method according to any one of claims 1 to 8, characterized in that: After identifying patterns and anomalies in the space-time grid cells based on the fused cryosphere data, and / or predicting future trends in the cryosphere, the method further includes: Display the distribution and changes of cryosphere data in space-time grid cells based on heat maps, space-time trajectory maps and three-dimensional maps; and / or, Based on the patterns and anomaly identification results in the space-time grid units, as well as the future changes in the cryosphere, scientific decision support is provided; among which, the scientific decisions include: glacier retreat prediction, snow cover change trend, and assessment of the impact of climate change on the cryosphere.

10. A heterogeneous multi-source cryosphere big data exploration and analysis system based on time and space segmentation, characterized in that: The system comprises: A preprocessing module, used for collecting and preprocessing cryosphere data, wherein the cryosphere data includes: meteorological data, remote sensing data and geographic information data; The space-time partitioning module is used to divide the study area into space-time grid cells; each space-time grid cell contains cryosphere data within a specific time period and spatial range; Heterogeneous data fusion module, used to fuse heterogeneous cryosphere data in each space-time grid unit; The spatiotemporal analysis and modeling module is used to identify patterns and anomalies in spatiotemporal grid cells based on the fused cryosphere data and / or predict future changes in the cryosphere.

Images