Seismic Forecasting Method, Device, Equipment and Medium Based on Remote Sensing Data

Abstract

The present invention relates to the technical field of satellite image processing, and discloses a seismic prediction method, device, equipment and medium based on remote sensing data. The method includes: acquiring original remote sensing data of a target area within a first time range, and performing seismic anomaly detection on the original remote sensing data to generate initial anomaly data; filtering the initial anomaly data based on a preset seismic anomaly recognition condition to generate target anomaly data; wherein the preset seismic anomaly recognition condition includes a spatial anomaly recognition condition and a temporal anomaly recognition condition, and the target anomaly data is used to represent anomalies caused by seismic activities; determining at least one local area in the target area based on the target anomaly data, and performing seismic prediction on the local area within a second time range. Through the above technical solution, the recognition accuracy of data anomalies of geophysical parameters caused by earthquakes is improved, thereby improving the accuracy of seismic prediction.

Description

Technical Field

[0001] The present invention relates to the field of satellite image processing technology, and in particular to an earthquake prediction method, device, equipment and medium based on remote sensing data. Background Art

[0002] A strong earthquake in a densely populated area will cause immeasurable human and economic losses. Therefore, accurate monitoring and prediction of earthquakes has become very urgent. The basic assumption of current earthquake prediction research is that there is precursor information in the abnormal phenomenon of geophysical parameters, and this information can be used to predict future earthquakes in the medium and short time scales (several months to weeks). Satellite remote sensing data can monitor various physical parameters of the earth without contact and around the clock. Therefore, it is necessary to provide a solution for earthquake monitoring and prediction using satellite remote sensing data, in order to reduce the cost of earthquake prediction and improve the frequency and accuracy of earthquake prediction. Summary of the invention

[0003] In order to solve the above technical problems, the present invention provides an earthquake prediction method, device, equipment and medium based on remote sensing data to improve the accuracy of earthquake prediction.

[0004] In a first aspect, the present invention provides an earthquake prediction method based on remote sensing data, the method comprising:

[0005] Acquiring original remote sensing data of the target area within a first time range, and performing seismic anomaly detection on the original remote sensing data to generate initial anomaly data;

[0006] Based on preset earthquake anomaly identification conditions, the initial anomaly data is anomaly filtered to generate target anomaly data; wherein the preset earthquake anomaly identification conditions include spatial anomaly identification conditions and temporal anomaly identification conditions, and the target anomaly data is data used to characterize anomalies caused by earthquake activity;

[0007] Based on the target abnormal data, at least one local area in the target area is determined, and an earthquake prediction within a second time range is performed for the local area.

[0008] In some embodiments, the spatial anomaly recognition condition is that the average anomaly absolute value of each valid pixel covered within a preset spatial range is greater than or equal to a first anomaly threshold; wherein the valid pixel is a pixel whose anomaly absolute value is greater than or equal to a second anomaly threshold, and the first anomaly threshold is greater than the second anomaly threshold;

[0009] The temporal anomaly recognition condition is that the proportion of abnormal observations that meet the spatial anomaly recognition condition within the first time range is greater than or equal to a preset proportion threshold.

[0010] In some embodiments, the performing anomaly filtering on the initial anomaly data based on the preset seismic anomaly identification condition to generate target anomaly data includes:

[0011] According to the preset moving direction, the preset moving step length and the preset spatial range, the initial abnormal data is subjected to sliding window processing based on the spatial abnormality identification condition, so as to perform primary abnormality filtering on the initial abnormal data and generate intermediate abnormal data;

[0012] Based on the time anomaly identification condition, the intermediate anomaly data is subjected to secondary anomaly filtering to generate the target anomaly data.

[0013] Optionally, performing sliding window processing based on the spatial anomaly identification condition on the initial abnormal data according to the preset moving direction, the preset moving step length and the preset spatial range to perform initial abnormality filtering on the initial abnormal data, and generating intermediate abnormal data includes:

[0014] Determining a current processing pixel from the initial abnormal data;

[0015] Determine whether each pixel value within the preset spatial range centered on the currently processed pixel satisfies the spatial anomaly recognition condition, and determine whether to filter the currently processed pixel based on the determination result;

[0016] According to a preset moving direction, determine the next pixel with a preset step distance from the current processing pixel, and use the next pixel to update the current processing pixel;

[0017] Return to the step of determining whether the values ​​of each pixel within the preset spatial range centered on the current processing pixel meet the spatial anomaly identification condition, and determining whether to filter the current processing pixel based on the determination result, until the initial abnormal data is traversed.

[0018] Optionally, performing secondary anomaly filtering on the intermediate anomaly data based on the time anomaly identification condition to generate the target anomaly data includes:

[0019] For each pixel position, determining the proportion of the number of abnormal values ​​contained in the intermediate abnormal data in the number of pixels, and retaining the pixel position when it is determined that the proportion meets the preset ratio threshold;

[0020] The target anomaly data is generated based on the retained pixel positions.

[0021] In some embodiments, the spatial anomaly identification condition and the temporal anomaly identification condition are determined based on a preset earthquake case data set; wherein the preset earthquake case data set is used to record the earthquake time, earthquake location and earthquake intensity of multiple historical earthquake events.

[0022] In some embodiments, the target area is within the regional range of a preset earthquake monitoring area; wherein the preset earthquake monitoring area is a geographical area where the number of historical earthquakes has reached a preset threshold.

[0023] In some embodiments, the time interval between two adjacent earthquake predictions is the duration corresponding to the second time range.

[0024] In some embodiments, after determining at least one local area in the target area based on the target abnormal data and performing earthquake prediction for the local area within a second time range, the method further includes:

[0025] An evaluation index value of the earthquake prediction is determined based on the target abnormal data; wherein the evaluation index includes at least one of the accuracy rate, false alarm rate, missed alarm rate, normal rate and Matthews correlation coefficient.

[0026] In a second aspect, the present invention provides an earthquake prediction device based on remote sensing data, the device comprising:

[0027] An initial abnormal data generation module is used to obtain the original remote sensing data of the target area within a first time range, and perform seismic anomaly detection on the original remote sensing data to generate initial abnormal data;

[0028] A target abnormal data generation module, used to perform abnormal filtering on the initial abnormal data based on preset earthquake abnormality identification conditions to generate target abnormal data; wherein the preset earthquake abnormality identification conditions include spatial abnormality identification conditions and temporal abnormality identification conditions, and the target abnormal data is data used to characterize abnormalities caused by earthquake activities;

[0029] The earthquake prediction module is used to determine at least one local area in the target area based on the target abnormal data, and to predict an earthquake in the local area within a second time range.

[0030] In some embodiments, the spatial anomaly recognition condition is that the average anomaly absolute value of each valid pixel covered within a preset spatial range is greater than or equal to a first anomaly threshold; wherein the valid pixel is a pixel whose anomaly absolute value is greater than or equal to a second anomaly threshold, and the first anomaly threshold is greater than the second anomaly threshold;

[0031] The temporal anomaly recognition condition is that the proportion of abnormal observations that meet the spatial anomaly recognition condition within the first time range is greater than or equal to a preset proportion threshold.

[0032] In some embodiments, the target abnormal data generation module includes:

[0033] A spatial filtering submodule, configured to perform sliding window processing based on the spatial anomaly identification condition on the initial abnormal data according to a preset moving direction, a preset moving step length and the preset spatial range, so as to perform a primary abnormality filtering on the initial abnormal data and generate intermediate abnormal data;

[0034] The time filtering submodule is used to perform secondary anomaly filtering on the intermediate anomaly data based on the time anomaly identification condition to generate the target anomaly data.

[0035] Optionally, the spatial filtering submodule is specifically used for:

[0036] Determining a current processing pixel from the initial abnormal data;

[0037] Determine whether each pixel value within the preset spatial range centered on the currently processed pixel satisfies the spatial anomaly recognition condition, and determine whether to filter the currently processed pixel based on the determination result;

[0038] According to a preset moving direction, determine the next pixel with a preset step distance from the current processing pixel, and use the next pixel to update the current processing pixel;

[0039] Return to the step of determining whether the values ​​of each pixel within the preset spatial range centered on the current processing pixel meet the spatial anomaly identification condition, and determining whether to filter the current processing pixel based on the determination result, until the initial abnormal data is traversed.

[0040] Optionally, the time filtering submodule is specifically used for:

[0041] For each pixel position, determining the proportion of the number of abnormal values ​​contained in the intermediate abnormal data in the number of pixels, and retaining the pixel position when it is determined that the proportion meets the preset ratio threshold;

[0042] The target anomaly data is generated based on the retained pixel positions.

[0043] In some embodiments, the spatial anomaly identification condition and the temporal anomaly identification condition are determined based on a preset earthquake case data set; wherein the preset earthquake case data set is used to record the earthquake time, earthquake location and earthquake intensity of multiple historical earthquake events.

[0044] In some embodiments, the target area is within the regional range of a preset earthquake monitoring area; wherein the preset earthquake monitoring area is a geographical area where the number of historical earthquakes has reached a preset threshold.

[0045] In some embodiments, the time interval between two adjacent earthquake predictions is the duration corresponding to the second time range.

[0046] In some embodiments, the earthquake prediction device based on remote sensing data further includes an evaluation index value determination module, which is used to:

[0047] After determining at least one local area in the target area based on the target abnormal data and predicting an earthquake for the local area within a second time range, an evaluation index value of the earthquake prediction is determined based on the target abnormal data; wherein the evaluation index includes at least one of the accuracy rate, false alarm rate, omission rate, normal rate and Matthews correlation coefficient.

[0048] In a third aspect, the present invention provides an electronic device, the electronic device comprising:

[0049] Processor and memory;

[0050] The processor is used to execute the steps of the method described in any embodiment of the present invention by calling the program or instruction stored in the memory.

[0051] In a fourth aspect, the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a program or instructions, wherein the program or the instructions enable a computer to execute the steps of the method described in any embodiment of the present invention.

[0052] The earthquake prediction method, device, equipment and medium based on remote sensing data provided by the embodiments of the present invention can obtain the original remote sensing data of the target area within a first time range, and perform earthquake anomaly detection on the original remote sensing data to generate initial anomaly data; based on preset earthquake anomaly identification conditions, perform anomaly filtering on the initial anomaly data to generate target anomaly data; wherein the preset earthquake anomaly identification conditions include spatial anomaly identification conditions and temporal anomaly identification conditions, and the target anomaly data is data used to characterize anomalies caused by earthquake activity; based on the target anomaly data, determine at least one local area in the target area, and perform earthquake prediction on the local area within a second time range. By providing preset earthquake anomaly identification conditions in spatial and temporal dimensions, the accuracy of identifying data anomalies of geophysical parameters caused by earthquakes is improved, thereby improving the accuracy of earthquake prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.

[0055] Figure 1 is a flow chart of an earthquake prediction method based on remote sensing data provided by an embodiment of the present invention;

[0056] Figure 2 is a schematic diagram of the interval time of continuous earthquake prediction provided by an embodiment of the present invention;

[0057] Figure 3 is a schematic structural diagram of an earthquake prediction device based on remote sensing data provided by an embodiment of the present invention;

[0058] Figure 4 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0059] In order to more clearly understand the above-mentioned objectives, features and advantages of the present invention, the scheme of the present invention will be further described in detail below. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.

[0060] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present invention, rather than all of the embodiments.

[0061] The earthquake prediction method based on remote sensing data provided by the embodiment of the present invention is mainly applicable to the scenarios of monitoring and predicting possible earthquake activities and evaluating earthquake prediction capabilities. The earthquake prediction method based on remote sensing data provided by the embodiment of the present invention can be performed by an earthquake prediction device based on remote sensing data, which can be implemented by software and / or hardware, and the device can be integrated in an electronic device, such as a laptop computer, a desktop computer or a server.

[0062] Figure 1 is a flow chart of an earthquake prediction method based on remote sensing data provided by an embodiment of the present invention. Figure 1 , the earthquake prediction method based on remote sensing data specifically includes:

[0063] S110, acquiring original remote sensing data of the target area within a first time range, and performing seismic anomaly detection on the original remote sensing data to generate initial anomaly data.

[0064] Wherein, the target area refers to the area where seismic activity is to be monitored. The first time range is the historical time period before the current date, for example, it can be the historical date range immediately adjacent to the current date. Exemplarily, the first time range is the historical date range of 3 months immediately before the current date. The reason for this setting is that although a longer abnormal monitoring time window can capture more counts of abnormal values ​​with shorter duration, unrelated abnormalities far away from the time of earthquake occurrence will lead to higher uncertainty in earthquake-related signals. Therefore, setting the duration of the first time range to 3 months can improve the correlation between the detected abnormalities and seismic activity to a certain extent, thereby improving the accuracy of subsequent earthquake predictions. Original remote sensing data refers to remote sensing product data that has not been processed by subsequent data, which is related to the selected geophysical parameters. Initial abnormal data refers to the abnormal values ​​of geophysical parameters obtained by abnormal detection of remote sensing data, which are abnormal data that have not been processed by subsequent abnormal value filtering.

[0065] Specifically, based on the basic assumption that there is precursor information in geophysical parameter anomalies and that the precursor information can be used to predict future short- to medium-scale earthquakes (several months to several weeks), before conducting earthquake monitoring and prediction, the embodiments of the present invention first select appropriate geophysical parameters to be monitored, such as surface temperature, latent heat flux, air temperature, humidity, outgoing long-wave radiation, and brightness temperature measured by thermal infrared, based on multiple factors such as the physical nature of earthquakes, sensitivity to earthquake occurrence processes, inversion accuracy of satellite observations, accessibility of data sources, and spatiotemporal resolution of remote sensing data. In addition, the original remote sensing data corresponding to the selected geophysical parameters within the first time range is obtained from the remote sensing product website or database.

[0066] Then, the original remote sensing data is subjected to certain preprocessing operations, such as geographic positioning, projection transformation, physical quantity conversion, mosaicking, synthesis and quality control, to ensure the consistency and accuracy of the preprocessed original remote sensing data. At the same time, considering that different anomaly analysis methods / anomaly detection methods have different requirements for input data, the above-mentioned preprocessed original remote sensing data can be further subjected to preprocessing operations specific to the anomaly detection method, such as normalization of neighborhood pixels. The preprocessed remote sensing data obtained in this way will serve as the basis for subsequent data processing.

[0067] Afterwards, an anomaly detection method is used to perform outlier detection on the preprocessed raw remote sensing data, that is, to identify the abnormal signals of the above-selected geophysical parameters in the preprocessed raw remote sensing data, and the obtained results are the initial abnormal data.

[0068] The above-mentioned anomaly detection method may be a method capable of identifying anomalies in remote sensing data, such as the offset index ZS method or the robust satellite technique (RST) method, etc. Both methods aim to identify significant changes relative to their "normal" conditions in the space and / or time domain.

[0069] The above ZS method calculates the abnormal disturbance through the normalized Z score, which is calculated as follows:

[0070]

[0071] Where v(x,y,t) is the current pixel value at position (x,y) and time t; μ(x,y) is the average value calculated from the reference pixel values ​​(also called background field) over many years at position (x,y) and time t (or a time close to time t); δ(x,y) is the standard deviation of the above background fields.

[0072] In the ZS method, the upper and lower boundaries of the envelope are set to μ±nδ. If the ZS value calculated from the pixel value in the preprocessed raw remote sensing data falls outside the relevant upper and lower boundaries of the envelope, an abnormal signal will be detected.

[0073] The above RST is a multi-temporal statistical method used to analyze long-term satellite records with similar observation conditions (e.g., the same month, time of day, and sensor data). RST combines the neighborhood difference in the eddy method with the normalization of the background field in the ZS method, and clearly defines the mathematical expression of the pre-seismic anomaly in statistics. The RST formula is:

[0074]

[0075]

[0076] in, is the spatial average of a homogeneous neighborhood; Δv(x,y,t) represents the difference between the pixel value at position (x,y) and the surrounding uniform pixels; μ Δv (x,y) and δ Δv (x,y) are the mean and standard deviation of the background field calculated from Δv(x,y,t), respectively.

[0077] In some embodiments, the target area is within the area range of a preset earthquake monitoring area.

[0078] The preset earthquake monitoring area is a geographical area where the number of historical earthquakes reaches a preset number threshold. The preset number threshold can be set empirically to determine earthquake-active areas. For example, the preset earthquake monitoring area is an earthquake-active area in the global scope / national scope / regional scope where earthquakes occur frequently.

[0079] Specifically, in order to further improve the accuracy of earthquake prediction, a preset earthquake monitoring area can be predetermined. Then, based on the assumption that future earthquakes will only occur in the preset earthquake monitoring area, and that geophysical parameter anomalies in areas outside the preset earthquake monitoring area are irrelevant to the occurrence of earthquakes, a target area is determined from the preset earthquake monitoring area for earthquake monitoring and prediction.

[0080] The above-mentioned preset earthquake monitoring area can be obtained from historical earthquake data, and can be used to indicate the possible area of ​​the upcoming earthquake in the predictive analysis, so as to ensure the earthquake prediction capability higher than the natural probability as much as possible. Since earthquakes are not randomly distributed in space, historical earthquake cases can reflect seismic activities in the global / national / regional scope. For example, in an embodiment of the present invention, a total of 4719 earthquake catalog data with a magnitude ≥6 and a focal depth ≤70km from 1980 to 2020 provided by the United States Geological Survey USGS are used to deduce the preset earthquake monitoring area in the world. For the grid whose number of earthquakes in the earthquake catalog data reaches the preset number threshold, a 5°×5° grid is marked around it to obtain the preset earthquake monitoring area in the world.

[0081] S120: Based on preset earthquake anomaly recognition conditions, perform anomaly filtering on the initial anomaly data to generate target anomaly data.

[0082] Among them, the preset seismic anomaly identification condition is a pre-set condition for identifying whether the abnormal value of the geophysical parameter is caused by seismic activity. The preset seismic anomaly identification condition includes a spatial anomaly identification condition and a temporal anomaly identification condition. That is, the abnormal values ​​caused by seismic activity are identified from the spatial dimension and the temporal dimension to increase the accuracy of anomaly identification. The target anomaly data is the data used to characterize the anomaly caused by seismic activity.

[0083] In some embodiments, the spatial anomaly identification condition is that the average abnormal absolute value of each valid pixel covered within the preset spatial range is greater than or equal to the first abnormal threshold. Wherein, the preset spatial range is a pre-set area with a certain spatial size, such as a spatial range of 5°*5° around the central pixel. A valid pixel is a pixel whose abnormal absolute value is greater than or equal to the second abnormal threshold. In the embodiment of the present invention, two critical values ​​of abnormal values ​​with different numerical values ​​are set, namely, a second abnormal threshold for identifying whether the pixel value in the initial abnormal data is valid and a first abnormal threshold for identifying whether the abnormal value is caused by seismic activity, and the first abnormal threshold is slightly greater than the second abnormal threshold. For example, if the second abnormal threshold is determined to be 2 within the value range of 1.5 to 3.5 of the second abnormal threshold, then the first abnormal threshold can be taken as a value greater than 0.5 of the second abnormal threshold, that is, 2.5. The spatial anomaly identification condition is that there are valid pixels in the preset spatial range whose absolute value of the pixel value in the initial abnormal data (i.e., the abnormal absolute value) is greater than or equal to the second abnormal threshold, and the mean of the absolute values ​​of these valid pixels (i.e., the average abnormal absolute value) is greater than or equal to the first abnormal threshold.

[0084] In some embodiments, the temporal anomaly recognition condition is that the proportion of abnormal observations that meet the spatial anomaly recognition condition within the first time range is greater than or equal to a preset proportion threshold. The abnormal observation proportion refers to the proportion of abnormal observations that meet the spatial anomaly recognition condition in all observation times. The preset proportion threshold is a critical value of the preset abnormal observation proportion. For example, the temporal resolution of the original remote sensing data is one day, and the first time range is 3 months, a total of 120 days, so there are 120 observations within the first time range. If a certain pixel meets the spatial anomaly recognition condition for no less than 10 days within 120 days, the abnormal observation proportion is greater than or equal to 1 / 12. If the preset proportion threshold is 1 / 12, then the abnormal observation proportion in this example is greater than or equal to the preset proportion threshold, that is, the pixel meets the temporal anomaly recognition condition and the spatial anomaly recognition condition.

[0085] The setting of the above-mentioned spatial anomaly identification conditions and temporal anomaly identification conditions can ensure that the identified anomaly values ​​are as closely related to seismic activity as possible, thereby ensuring the ability to predict earthquakes that is higher than the natural probability.

[0086] In some embodiments, the spatial anomaly identification condition and the temporal anomaly identification condition are determined based on a preset earthquake case data set. The preset earthquake case data set is used to record the earthquake time, earthquake location and earthquake intensity (such as earthquake magnitude) of multiple historical earthquake events. In order to improve the lateral comparability of various earthquake prediction methods, a preset earthquake case data set of global / national / regional scope for many years can be pre-constructed in an embodiment of the present invention as a benchmark data for statistical analysis of various geophysical parameters and anomaly detection methods. For example, the global earthquake catalog from the United States Geological Survey (USGS) can be used to select 1825 earthquake events with a magnitude ≥6 and a focal depth ≤70km between 2006 and 2020 to construct a global typical earthquake case data set. Then, a part of the earthquake case data in the preset earthquake case data set is used to perform earthquake retrospective analysis, that is, to analyze the relationship between each earthquake case and the abnormal data in the first time range before it, so as to determine the parameters in the spatial anomaly identification condition and the temporal anomaly identification condition. Another part of the earthquake case data in the preset earthquake case data set is used to perform predictive analysis on the parameters in the determined spatial anomaly identification conditions and temporal anomaly identification conditions, that is, earthquake prediction is performed using the parameters and anomaly data in the spatial anomaly identification conditions and temporal anomaly identification conditions, and the earthquake prediction results are compared with the above-mentioned other part of the earthquake case data for prediction correctness verification analysis.

[0087] Specifically, according to the above description, the initial abnormal data is the abnormal value of the geophysical parameter calculated according to the mathematical formula and the observation data, which cannot characterize the correlation between the abnormal value and the seismic activity. Therefore, in the embodiment of the present invention, after obtaining the initial abnormal data, the initial abnormal data is filtered for abnormal values ​​using the preset seismic anomaly identification conditions to identify the abnormal values ​​in the initial abnormal data that are correlated with the seismic activity, and the obtained result is the target abnormal data.

[0088] S130. Determine at least one local area in the target area based on the target abnormal data, and perform earthquake prediction for the local area within a second time range.

[0089] Among them, the second time range is the time scope of earthquake prediction, which is a date range corresponding to a certain time length immediately after the current date, for example, it can be 10 days, 20 days or 30 days after the current date.

[0090] Specifically, according to the above description, the pixel value contained in the target abnormal data is an abnormal value related to seismic activity, so earthquake prediction can be performed based on the abnormal value. For example, the regional dimension of earthquake prediction is determined according to the location of the pixel value in the target abnormal data, that is, the local area (such as the area of ​​5°*5° around the location of the pixel value). Then, an earthquake prediction of a certain magnitude in the second time range is performed in each local area. For example, an earthquake of a certain magnitude and a certain focal depth will occur within the local area and in the future second time range. The magnitude and focal depth of the above prediction can be determined according to the magnitude and focal depth of the earthquake in the typical earthquake case collected in the above preset earthquake case data set. For example, the magnitude of the prediction is greater than or equal to the magnitude of the earthquake in the typical earthquake case, and the focal depth of the prediction is less than or equal to the focal depth in the typical earthquake case. This can also improve the accuracy of earthquake prediction to a certain extent.

[0091] In some embodiments, the time interval between two adjacent earthquake predictions is the time length corresponding to the second time range. That is, there are no repeated dates between the second time ranges corresponding to two adjacent earthquake predictions. This can avoid repeated data calculations and multiple repeated alarms for the same earthquake, and can also ensure that the earthquake prediction is more objective.

[0092] See also Figure 2 , using the original remote sensing data of the first time range (horizontal line filling) before the current date (slash filling) of the first earthquake prediction for earthquake monitoring, and making earthquake predictions in the second time range (vertical line filling). After the earthquake prediction is completed, the current date is changed to the date immediately after the second time range, that is, Figure 2 The current date ' is filled in with a slash in the example. Then, the second earthquake prediction uses the original remote sensing data of the first time range ' before the current date ' to perform earthquake monitoring and perform earthquake prediction in the second time range '.

[0093] The above-mentioned earthquake prediction method based on remote sensing data provided by the embodiment of the present invention can obtain the original remote sensing data of the target area within the first time range, and perform earthquake anomaly detection on the original remote sensing data to generate initial anomaly data; based on the preset earthquake anomaly identification conditions, perform anomaly filtering on the initial anomaly data to generate target anomaly data; wherein the preset earthquake anomaly identification conditions include spatial anomaly identification conditions and temporal anomaly identification conditions, and the target anomaly data is data used to characterize anomalies caused by earthquake activities; based on the target anomaly data, determine at least one local area in the target area, and perform earthquake prediction on the local area within the second time range. It is achieved that by providing preset earthquake anomaly identification conditions in the spatial dimension and the temporal dimension, the accuracy of identifying data anomalies of geophysical parameters caused by earthquakes is improved, thereby improving the accuracy of earthquake prediction.

[0094] In some embodiments, the filtering process of outliers related to seismic activity in S120 may be implemented as the following steps A and B.

[0095] Step A: According to a preset moving direction, a preset moving step length and a preset spatial range, the initial abnormal data is subjected to sliding window processing based on a spatial abnormality recognition condition to perform a primary abnormality filtering on the initial abnormal data and generate intermediate abnormal data.

[0096] Specifically, in the embodiment of the present invention, each image (pixel value is an abnormal value in the initial abnormal data) contained in the initial abnormal data is subjected to sliding window processing. The processing process in each sliding window is to determine whether each abnormal value in the window meets the spatial abnormality recognition condition, and determine whether to retain the main pixel corresponding to the window according to the judgment result. After each sliding window processing, the window will be moved according to the preset moving direction and the preset moving step length, and the processing in the next sliding window will be performed. After all the images contained in the initial abnormal data are processed, the result is the intermediate abnormal data of the initial abnormal data after spatial filtering.

[0097] Exemplarily, step A may be specifically implemented as follows:

[0098] Step A: determine the current processing pixel from the initial abnormal data.

[0099] Specifically, for any image in the initial abnormal data, a central pixel is determined according to a certain rule (such as the first pixel starting from the upper left corner of the image) as the current processing pixel in the image.

[0100] Step A2: determine whether the values ​​of each pixel within a preset spatial range centered on the currently processed pixel meet the spatial anomaly recognition condition, and determine whether to filter the currently processed pixel based on the determination result.

[0101] Specifically, if it is determined that the values ​​of each pixel within the preset spatial range centered on the current processing pixel meet the spatial anomaly recognition condition, the current processing pixel is retained; if it is determined that the values ​​of each pixel within the preset spatial range centered on the current processing pixel do not meet the spatial anomaly recognition condition, the current processing pixel is filtered out.

[0102] That is, according to the area size of the preset spatial range, with the current processing pixel as the center, the processing area of ​​this filtering process can be determined. Then, the value of each pixel contained in the processing area is compared with the second abnormal threshold to determine the valid pixels contained in the processing area. After that, the average abnormal absolute value of these valid pixels is calculated and compared with the first abnormal threshold. If the comparison result is that the average abnormal absolute value is greater than or equal to the first abnormal threshold, the current processing pixel is retained; if the comparison result is that the average abnormal absolute value is less than the first abnormal threshold, the current processing pixel is eliminated, such as replacing the abnormal value of the current processing pixel with a preset invalid value.

[0103] Step A3: according to the preset moving direction, determine the next pixel with a preset step distance from the current processing pixel, and use the next pixel to update the current processing pixel.

[0104] Specifically, in the above image, according to the preset moving direction, a preset step length is moved from the current processing pixel to determine a new pixel. The new pixel is the next pixel corresponding to the current processing pixel in the sliding window processing. Then, the next pixel is used as the new current processing pixel, that is, the current processing pixel is updated.

[0105] Step A4: Return to step A2 and execute until all the initial abnormal data are traversed.

[0106] Specifically, the sliding window processing is performed cyclically according to the above steps A2 and A3 until the above image processing is completed.

[0107] Afterwards, according to the process of step A1 to step A4, each image in the initial abnormal data is subjected to an abnormal value filtering process in a spatial dimension, so as to obtain intermediate abnormal data.

[0108] Step B: Based on the time anomaly recognition condition, perform secondary anomaly filtering on the intermediate anomaly data to generate target anomaly data.

[0109] Specifically, for each pixel position, the proportion of the number of abnormal values ​​contained in the intermediate abnormal data in the number of pixels is determined, and when it is determined that the proportion meets a preset ratio threshold, the pixel position is retained; and target abnormal data is generated based on the retained pixel position.

[0110] In specific implementation, the intermediate abnormal data is a time series image set, so an abnormal value time series can be obtained at each pixel position. Some values ​​in the abnormal value time series are abnormal values, and some values ​​are invalid values. Then, for any pixel position, the ratio of the number of abnormal values ​​contained in the abnormal value time series and the total number of values ​​contained in the entire abnormal value time series (i.e., the number of pixels) is calculated to obtain the abnormal observation ratio at the pixel position. Then, compare the abnormal observation ratio with the preset ratio threshold. If the abnormal observation ratio is greater than or equal to the preset ratio threshold, then retain the pixel position; if the abnormal observation ratio is less than the preset ratio threshold, then eliminate the pixel position.

[0111] According to the above process, the outliers in the time dimension can be filtered for each pixel position, and finally an image with valid pixel positions can be obtained, namely the target anomaly data. Each valid pixel position contained in the target anomaly data can be considered as a location where earthquake activity will occur. Subsequently, a local area for earthquake prediction can be determined based on each valid pixel position.

[0112] In some embodiments, after S130, the earthquake prediction method based on remote sensing data further includes: determining an evaluation index value for earthquake prediction based on the target abnormal data.

[0113] The evaluation index includes at least one of the accuracy rate, false alarm rate, missed alarm rate, normal rate and Matthews correlation coefficient.

[0114] Specifically, in order to improve the horizontal comparability between various earthquake prediction methods, the embodiment of the present invention provides an evaluation index for earthquake prediction results, that is, at least one of the accuracy rate, false alarm rate, missed alarm rate, normal rate and Matthews correlation coefficient. In this way, the same evaluation index can be used to evaluate the earthquake prediction ability.

[0115] The meanings of the above-mentioned accuracy rate, false alarm rate, missed alarm rate and normal rate can be found in Table 1, which are four independent evaluation indicators.

[0116] Since many earthquake prediction methods only provide part of the above four independent evaluation indicators, such as the prediction accuracy, and ignore other evaluation indicators, thus resulting in an incomplete display of earthquake prediction capabilities, a comprehensive evaluation indicator, namely the improved Matthews correlation coefficient, is determined in the embodiment of the present invention to fully display the comprehensive results of the above four independent evaluation indicators.

[0117] The improved Matthews correlation coefficient (MCC) is the application of the Matthews correlation coefficient in machine learning to the field of earthquake prediction. It is used as the accuracy of earthquake prediction. Its calculation formula is as follows:

[0118]

[0119] Among them, Acc is the accurate number of reports, Nor is the normal number, Fal is the false alarm number, and Mis is the missed alarm number.

[0120] Under the constraint of the improved Matthews correlation coefficient MCC′, the accuracy can be used as the confidence level of earthquake prediction. This not only provides an evaluation indicator suitable for the field of machine learning, improves the understanding of researchers in various fields on the accuracy of earthquake prediction, and further improves the horizontal comparability of various earthquake prediction methods, but also provides a sufficient confidence level for the reliability of earthquake prediction, thereby further improving the effectiveness of earthquake prediction.

[0121] Table 1 Evaluation indicators of earthquake prediction ability

[0122]

[0123] Based on the above method flow of the embodiment of the present invention, the earthquake prediction capabilities of the above ZS method and RST method are compared horizontally using earthquake data from 2006 to 2020 to demonstrate the ability of the embodiment of the present invention to achieve unified and quantitative horizontal comparison of forecasting effectiveness. Among them, the spatial distribution patterns of the accuracy of the ZS method and the RST method are similar, and the highest accuracy is achieved in earthquake-active areas. However, the false alarm rates of these two methods are also very high in earthquake-active areas. Moreover, the normal rate is closely related to the false alarm rate. The MCC′ range of the ZS method and the RST method is -0.48 to 0.21, but the ZS method generates relatively high MCC′ values ​​in most grids. Judging from the spatial distribution of the MCC′ values, the MCC′ values ​​in areas with high accuracy are not high, indicating that the comprehensive index can reflect the comprehensive evaluation capability.

[0124] Figure 3 Schematic diagram of the structure of an earthquake prediction device based on remote sensing data provided by an embodiment of the present invention. Figure 3 The earthquake prediction device 300 based on remote sensing data specifically includes:

[0125] The initial abnormal data generating module 310 is used to obtain the original remote sensing data of the target area within the first time range, and perform seismic anomaly detection on the original remote sensing data to generate initial abnormal data;

[0126] The target abnormal data generation module 320 is used to perform abnormal filtering on the initial abnormal data based on the preset earthquake abnormality identification conditions to generate target abnormal data; wherein the preset earthquake abnormality identification conditions include spatial abnormality identification conditions and temporal abnormality identification conditions, and the target abnormal data is data used to characterize abnormalities caused by earthquake activities;

[0127] The earthquake prediction module 330 is used to determine at least one local area in the target area based on the target abnormal data, and to predict an earthquake in the local area within a second time range.

[0128] The earthquake prediction device based on remote sensing data provided by the embodiment of the present invention can obtain the original remote sensing data of the target area within the first time range, and perform earthquake anomaly detection on the original remote sensing data to generate initial anomaly data; based on the preset earthquake anomaly identification conditions, perform anomaly filtering on the initial anomaly data to generate target anomaly data; wherein the preset earthquake anomaly identification conditions include spatial anomaly identification conditions and temporal anomaly identification conditions, and the target anomaly data is data used to characterize anomalies caused by earthquake activities; based on the target anomaly data, determine at least one local area in the target area, and perform earthquake prediction on the local area within the second time range. By providing preset earthquake anomaly identification conditions in the spatial dimension and the temporal dimension, the accuracy of identifying data anomalies of geophysical parameters caused by earthquakes is improved, thereby improving the accuracy of earthquake prediction.

[0129] In some embodiments, the spatial anomaly recognition condition is that the average anomaly absolute value of each valid pixel covered within a preset spatial range is greater than or equal to a first anomaly threshold; wherein a valid pixel is a pixel whose anomaly absolute value is greater than or equal to a second anomaly threshold, and the first anomaly threshold is greater than the second anomaly threshold;

[0130] The temporal anomaly recognition condition is that the proportion of abnormal observations that meet the spatial anomaly recognition condition within the first time range is greater than or equal to a preset proportion threshold.

[0131] In some embodiments, the target abnormal data generation module 320 includes:

[0132] The spatial filtering submodule is used to perform sliding window processing on the initial abnormal data based on the spatial abnormality recognition condition according to the preset moving direction, the preset moving step length and the preset spatial range, so as to perform the initial abnormality filtering on the initial abnormal data and generate intermediate abnormal data;

[0133] The time filtering submodule is used to perform secondary anomaly filtering on the intermediate anomaly data based on the time anomaly identification conditions to generate target anomaly data.

[0134] Optionally, the spatial filtering submodule is specifically used for:

[0135] Determine the current processing pixel from the initial abnormal data;

[0136] Determine whether the values ​​of each pixel within a preset spatial range centered on the current processing pixel meet the spatial anomaly recognition conditions, and determine whether to filter the current processing pixel based on the judgment result;

[0137] According to the preset moving direction, determine the next pixel with a preset step distance from the current processing pixel, and use the next pixel to update the current processing pixel;

[0138] Return to the step of judging whether the values ​​of each pixel within a preset spatial range centered on the current processing pixel meet the spatial anomaly recognition condition, and determining whether to filter the current processing pixel based on the judgment result, until the initial anomaly data is traversed.

[0139] Optionally, the time filtering submodule is specifically used for:

[0140] For each pixel position, determine the proportion of the number of abnormal values ​​contained in the intermediate abnormal data to the number of pixels, and retain the pixel position when it is determined that the proportion meets a preset ratio threshold;

[0141] Generate target anomaly data based on the retained cell locations.

[0142] In some embodiments, the spatial anomaly identification conditions and the temporal anomaly identification conditions are determined based on a preset earthquake case data set; wherein the preset earthquake case data set is used to record the earthquake time, earthquake location and earthquake intensity of multiple historical earthquake events.

[0143] In some embodiments, the target area is within the regional scope of a preset earthquake monitoring area; wherein the preset earthquake monitoring area is a geographical area where the number of historical earthquakes has reached a preset threshold.

[0144] In some embodiments, the time interval between two adjacent earthquake predictions is the duration corresponding to the second time range.

[0145] In some embodiments, the earthquake prediction device 300 based on remote sensing data further includes an evaluation index value determination module for:

[0146] After determining at least one local area in the target area based on the target abnormal data and predicting an earthquake for the local area within a second time range, an evaluation index value of the earthquake prediction is determined based on the target abnormal data; wherein the evaluation index includes at least one of the accuracy rate, false alarm rate, missed alarm rate, normal rate and Matthews correlation coefficient.

[0147] The earthquake prediction device based on remote sensing data provided in the embodiment of the present invention can execute the earthquake prediction method based on remote sensing data provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0148] It is worth noting that in the above-mentioned embodiment of the earthquake prediction device based on remote sensing data, the various modules and sub-modules included are only divided according to functional logic, but are not limited to the above-mentioned division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional modules / sub-modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0149] Figure 4 It is a structural schematic diagram of an electronic device provided by an embodiment of the present invention.

[0150] See also Figure 4 The electronic device 400 provided in an embodiment of the present invention includes: a processor 420 and a memory 410; the processor 420 calls the program or instructions stored in the memory 410 to execute the steps of the earthquake prediction method based on remote sensing data provided in any embodiment of the present invention.

[0151] like Figure 4 As shown, the electronic device 400 is in the form of a general-purpose computing device. The components of the electronic device 400 may include, but are not limited to: one or more processors 420, a memory 410, and a bus 450 connecting different system components (including the memory 410 and the processor 420).

[0152] Bus 450 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor or a local bus using any of a variety of bus architectures. By way of example, these architectures include, but are not limited to, an Industry Standard Architecture (ISA) bus, a Micro Channel Architecture (MAC) bus, an Enhanced ISA bus, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnect (PCI) bus.

[0153] The electronic device 400 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by the electronic device 400, including volatile and non-volatile media, removable and non-removable media.

[0154] The memory 410 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 411 and / or cache memory 412. The electronic device 400 may further include other removable / non-removable, volatile / non-volatile computer system storage media. By way of example only, the storage system 413 may be used to read and write non-removable, non-volatile magnetic media ( Figure 4 not shown, usually called a "hard drive"). Although Figure 4Not shown in the figure, a disk drive for reading and writing to a removable non-volatile disk (e.g., a "floppy disk"), and an optical disk drive for reading and writing to a removable non-volatile optical disk (e.g., a CD-ROM, a DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to the bus 450 via one or more data medium interfaces. The memory 410 may include at least one program product having a set (e.g., at least one) of program modules that are configured to perform the functions of the embodiments of the present invention.

[0155] A program / utility 414 having a set (at least one) of program modules 415 may be stored, for example, in the memory 410, such program modules 415 including but not limited to an operating system, one or more application programs, other program modules, and program data, each of which or some combination may include an implementation of a network environment. The program modules 415 generally perform the functions and / or methods of any embodiment described herein.

[0156] The electronic device 400 may also communicate with one or more external devices 460 (e.g., keyboard, pointing device, display 470, etc.), one or more devices that enable a user to interact with the electronic device 400, and / or any device that enables the electronic device 400 to communicate with one or more other computing devices (e.g., network card, modem, etc.). Such communication may be performed through an input / output interface (I / O interface) 430. Furthermore, the electronic device 400 may also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN) and / or public network, such as the Internet) through a network adapter 440. Figure 4 As shown, the network adapter 440 communicates with other modules of the electronic device 400 via the bus 450. It should be understood that, although not shown in the figure, other hardware and / or software modules can be used in conjunction with the electronic device 400, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.

[0157] It should be noted that Figure 4 The electronic device 400 shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present invention.

[0158] An embodiment of the present invention further provides a computer-readable storage medium, which stores a program or instruction, and the program or instruction enables a computer to execute the steps of the earthquake prediction method based on remote sensing data provided by any embodiment of the present invention.

[0159] The computer storage medium of the embodiment of the present invention can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In this document, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it.

[0160] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer-readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer-readable signal media may also be any computer-readable medium other than a computer-readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0161] Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0162] Computer program code for performing the operations of the present invention may be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0163] It should be noted that the terms used in the present invention are only for describing specific embodiments, rather than limiting the scope of the present invention. As shown in the present specification and claims, unless the context clearly indicates an exception, the words "one", "a", "a kind of" and / or "the" do not specifically refer to the singular, but may also include the plural. The term "and / or" includes any and all combinations of one or more related listed items. Relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any such actual relationship or order between these entities or operations. The term "include", "comprise" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method or device. In the absence of more restrictions, the elements defined by the sentence "including one..." do not exclude the existence of other identical elements in the process, method or device including the elements.

[0164] The above description is only a specific embodiment of the present invention, so that those skilled in the art can understand or implement the present invention. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments described herein, but should conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A seismic prediction method based on remote sensing data, characterized in that, it includes: Obtain the original remote sensing data of the target area within the first time range, and perform seismic anomaly detection on the original remote sensing data to generate initial anomaly data; Based on the preset seismic anomaly recognition conditions, filter the initial anomaly data to generate target anomaly data; wherein, the preset seismic anomaly recognition conditions include spatial anomaly recognition conditions and temporal anomaly recognition conditions, and the target anomaly data is data used to characterize the anomalies caused by seismic activities; the spatial anomaly recognition condition is that the average absolute value of anomalies covered by each valid pixel within the preset spatial range is greater than or equal to the first anomaly threshold; the valid pixel is a pixel whose absolute value of anomaly is greater than or equal to the second anomaly threshold; the first anomaly threshold is greater than the second anomaly threshold; the temporal anomaly recognition condition is that the anomaly observation ratio satisfying the spatial anomaly recognition condition within the first time range is greater than or equal to the preset ratio threshold; Based on the target anomaly data, determine at least one local area in the target area, and perform seismic prediction on the local area within the second time range.

2. The method according to claim 1, characterized in that, the filtering the initial anomaly data based on the preset seismic anomaly recognition conditions to generate target anomaly data includes: Performing a sliding window process on the initial anomaly data based on the spatial anomaly recognition condition according to the preset moving direction, preset moving step and the preset spatial range to perform primary anomaly filtering on the initial anomaly data and generate intermediate anomaly data; Based on the temporal anomaly recognition condition, perform secondary anomaly filtering on the intermediate anomaly data to generate the target anomaly data.

3. The method according to claim 2, characterized in that, the performing a sliding window process on the initial anomaly data based on the spatial anomaly recognition condition according to the preset moving direction, preset moving step and the preset spatial range to perform primary anomaly filtering on the initial anomaly data and generate intermediate anomaly data includes: Determine the current processing pixel from the initial anomaly data; Judge whether the pixel values of each pixel within the preset spatial range centered on the current processing pixel satisfy the spatial anomaly recognition condition, and determine whether to filter the current processing pixel based on the judgment result; According to the preset moving direction, determine the next pixel at a preset step distance from the current processing pixel, and update the current processing pixel with the next pixel; Return to execute the step of judging whether the pixel values of each pixel within the preset spatial range centered on the current processing pixel satisfy the spatial anomaly recognition condition, and determine whether to filter the current processing pixel based on the judgment result until the initial anomaly data is traversed.

4. The method according to claim 1, characterized in that, the spatial anomaly recognition condition and the temporal anomaly recognition condition are determined based on a preset seismic case dataset; wherein, the preset seismic case dataset is used to record the seismic time, seismic location and seismic intensity of multiple historical earthquake events.

5. The method according to claim 1, wherein, the target area is within the regional scope of a preset earthquake monitoring area; wherein, the preset earthquake monitoring area is a geographical area where the number of historical earthquake occurrences reaches a preset number threshold.

6. The method according to claim 1, wherein, the time interval between two adjacent earthquake forecasts is the duration corresponding to the second time range.

7. An earthquake forecasting device based on remote sensing data, wherein, comprising: an initial anomaly data generation module, configured to obtain the original remote sensing data of the target area within a first time range, and perform earthquake anomaly detection on the original remote sensing data to generate initial anomaly data; a target anomaly data generation module, configured to perform anomaly filtering on the initial anomaly data based on preset earthquake anomaly recognition conditions to generate target anomaly data; wherein, the preset earthquake anomaly recognition conditions include spatial anomaly recognition conditions and temporal anomaly recognition conditions, and the target anomaly data is data used to characterize anomalies caused by seismic activities; the spatial anomaly recognition condition is that the average absolute anomaly of each valid pixel covered within a preset spatial range is greater than or equal to a first anomaly threshold; the valid pixel is a pixel whose absolute anomaly is greater than or equal to a second anomaly threshold; the first anomaly threshold is greater than the second anomaly threshold; the temporal anomaly recognition condition is that the anomaly observation ratio satisfying the spatial anomaly recognition condition within the first time range is greater than or equal to a preset ratio threshold; an earthquake forecasting module, configured to determine at least one local area in the target area based on the target anomaly data, and perform earthquake forecasting on the local area within a second time range.

8. An electronic device, wherein, the electronic device includes: a processor and a memory; the processor is configured to execute the steps of the earthquake forecasting method based on remote sensing data according to any one of claims 1 to 6 by calling a program or instruction stored in the memory.

9. A computer-readable storage medium, wherein, the computer-readable storage medium stores a program or instruction, and the program or instruction causes a computer to execute the steps of the earthquake forecasting method based on remote sensing data according to any one of claims 1 to 6.

Images