A high-precision monitoring and early warning method and system for ground surface deformation based on a space-borne SAR

Abstract

The application belongs to the field of remote sensing monitoring and disaster early warning, and discloses a kind of high-precision monitoring and early warning method and system of ground deformation based on spaceborne SAR, comprising: S1, multi-temporal SAR image acquisition and pretreatment;S2, high-precision atmospheric phase correction, residual error is controlled within 2mm;S3, permanent scatterer identification and time series analysis, extract deformation time series;S4, multi-scale deformation trend prediction, predict the deformation evolution of future one week to one year;S5, intelligent early warning based on risk model, realize grading early warning.The application realizes millimeter-level deformation monitoring precision through high-precision atmospheric correction, realizes deformation trend prediction through time series prediction model, realizes intelligent grading early warning through multi-factor risk model, and the precision is improved by more than 50% compared with the traditional method, and the false alarm rate is reduced by 50%.

Description

Technical Field

[0001] This invention belongs to, but is not limited to, the field of remote sensing monitoring and disaster early warning technology, and particularly relates to a high-precision monitoring and early warning method and system for surface deformation based on spaceborne SAR. Background Technology

[0002] Surface deformation monitoring is a key technology for ensuring the safety of important facilities such as oil and gas pipelines, urban infrastructure, and mining areas. Surface deformation can be caused by a variety of factors, including geological tectonic movements, underground mining, groundwater extraction, and pipeline leaks. If it is not detected and warned in time, it may lead to serious consequences such as pipeline rupture, building tilting, and ground subsidence.

[0003] Synthetic Aperture Radar Interferometry (InSAR) technology utilizes spaceborne SAR satellites to image the same area multiple times. Through interferometric phase analysis, it can detect surface deformation at the millimeter level. It boasts advantages such as wide coverage, high monitoring accuracy, and immunity to cloud and rain weather, making it the preferred technology for large-scale surface deformation monitoring. However, existing InSAR deformation monitoring technologies have the following problems: Accuracy is limited by atmospheric delay error: changes in water vapor and electron density in the troposphere and ionosphere introduce interference phase error, affecting the accuracy of deformation measurement. Especially in mountainous and high-latitude regions, atmospheric error can reach the centimeter level, which seriously reduces the reliability of millimeter-level deformation detection.

[0004] Insufficient time series analysis capability: Traditional D-InSAR technology is mainly used for deformation detection between two imaging sessions, which makes it difficult to reveal the long-term evolution trend and periodicity of deformation, and has limited ability to identify gradual disasters in their early stages.

[0005] The early warning mechanism is simple and crude: most existing systems use fixed thresholds to trigger alarms, which fail to take into account the geological characteristics and historical baselines of different regions, resulting in low alarm accuracy and widespread false alarms and missed alarms.

[0006] Lack of spatial correlation with infrastructure: There is a lack of accurate spatial correlation analysis between deformation data and infrastructure such as pipelines and buildings, making it difficult to assess the actual threat level of deformation to specific facilities.

[0007] Based on the above analysis, the urgent technical problems that need to be solved in the existing technology are: Atmospheric error correction methods are not precise enough, deformation trend prediction capabilities are insufficient, and early warning models are too simple, making it difficult to meet the needs of high-precision and intelligent deformation monitoring and early warning. Summary of the Invention

[0008] To address the problems existing in the prior art, this invention provides a method and system for high-precision monitoring and early warning of surface deformation based on spaceborne SAR.

[0009] This invention is implemented as follows: a high-precision monitoring and early warning method for surface deformation based on spaceborne SAR, the method specifically includes: S1, Multi-temporal SAR Image Acquisition and Preprocessing: Acquire multi-temporal spaceborne SAR image sequences covering the monitoring area, perform fine orbit correction, multi-view processing and image registration to ensure the geometric accuracy and radiometric consistency of the image sequences.

[0010] S2, High-precision atmospheric phase correction: Using a combination of external atmospheric product assistance and spatial statistics, the tropospheric delay error and ionospheric error in the interference phase are corrected with high precision, and the residual error is controlled within 2mm.

[0011] S3, Permanent Scatterer Identification and Time Series Analysis: Using PS-InSAR or SBAS-InSAR technology, stable scattering targets are identified, deformation time series are constructed, and key indicators such as cumulative deformation, deformation rate, and acceleration are extracted.

[0012] S4, Multi-scale Deformation Trend Prediction: Based on deformation time series, time series prediction models (such as ARIMA, LSTM, etc.) are used to extrapolate multi-scale trends and predict the deformation evolution trend for the next week to one year.

[0013] S5, intelligent early warning based on risk model: Construct a risk assessment model that integrates multiple factors such as deformation data, facility distance, geological conditions, and historical events to achieve graded early warning for infrastructure safety.

[0014] Furthermore, the multi-temporal SAR image acquisition and preprocessing described in S1 includes: Image acquisition: Acquire at least 24 SAR images covering the same area over a period of no less than 12 months. It is recommended to use Sentinel-1, Terra SAR-X, or domestic Gaofen-3 satellite data.

[0015] Precise orbit correction: Using precise orbital ephemeris data, satellite orbital errors are corrected, achieving orbital accuracy at the centimeter level.

[0016] Multi-view processing: Perform multi-view processing on single-view and multiple-view images to reduce speckle noise and improve the signal-to-noise ratio.

[0017] Image registration: Subpixel-level registration technology is adopted, and the registration accuracy reaches within 0.1 pixels.

[0018] Furthermore, the high-precision atmospheric phase correction described in S2 includes: Tropospheric delay correction: Calculate the tropospheric wet delay component using atmospheric reanalysis data such as ECMWF or ERA5, or perform correction using GNSS water vapor products.

[0019] Ionospheric error correction: Ionospheric delay error is calculated and corrected using global ionospheric TEC data provided by IGS.

[0020] Residual error estimation: Spatiotemporal filtering method is used to estimate and remove the residual error after atmospheric correction.

[0021] Accuracy verification: The deformation results are verified using ground leveling or GNSS observation data to ensure that the accuracy reaches the millimeter level.

[0022] Furthermore, the permanent scatterer identification and temporal analysis described in S3 includes: PS point identification: Based on criteria such as coherence and amplitude dispersion index, candidate points of permanent scatterers are screened, with a point density of more than 100 points per square kilometer.

[0023] Phase unwrapping: Spatial-temporal phase unwrapping is performed on the selected PS points to restore the true deformation phase.

[0024] Deformation decomposition: Decomposes deformation into linear deformation, periodic deformation, and nonlinear deformation components.

[0025] Key indicator extraction: Calculate cumulative deformation, deformation rate (mm / year), deformation acceleration, deformation gradient and other indicators.

[0026] Furthermore, the multi-scale deformation trend prediction described in S4 includes: Short-term forecast (one week to one month): Using the ARIMA model or exponential smoothing method, mainly to capture short-term fluctuations and sudden trends in deformation.

[0027] Medium-term forecast (one month to one year): Using an LSTM neural network model, which integrates external factors such as meteorology and hydrology, the medium- and long-term evolution of deformation is predicted.

[0028] Uncertainty assessment: Estimating the confidence interval of the prediction results to quantify the uncertainty of the prediction.

[0029] Trend Classification: The predicted trend is divided into four categories: stable, slow deformation, accelerated deformation, and rapid deformation.

[0030] Furthermore, the risk-model-based intelligent early warning described in S5 includes: Risk factor construction includes deformation factors (cumulative deformation, deformation rate, acceleration), spatial factors (distance from facility, deformation gradient direction), geological factors (lithology, slope, fault distance), and historical factors (number of historical anomalies, handling records).

[0031] Risk score calculation: The comprehensive risk score is calculated using a weighted summation or machine learning model.

[0032] Early warning levels: Based on risk scores, early warnings are divided into four levels: red (high risk, immediate response), orange (medium to high risk, time-limited handling), yellow (medium risk, close monitoring), and blue (low risk, routine monitoring).

[0033] Threshold Adaptive: Based on historical data and regional characteristics, the warning threshold is dynamically adjusted to reduce the false alarm rate.

[0034] Another objective of this invention is to provide a high-precision monitoring and early warning system for surface deformation based on spaceborne SAR, which implements the aforementioned method, comprising: SAR Image Processing Module: Used for preprocessing and interferometric processing of multi-temporal SAR images.

[0035] Atmospheric correction module: Used for high-precision atmospheric phase error correction.

[0036] Time series analysis module: used to identify permanent scatterers and construct deformation time series.

[0037] Trend prediction module: Used for multi-scale deformation trend prediction.

[0038] Risk assessment module: Used to build multi-factor risk assessment models.

[0039] Early warning management module: used to conduct tiered early warnings and manage early warning information.

[0040] Data visualization module: used to display deformation heatmaps, trend curves, and early warning distributions.

[0041] Another object of the present invention is to provide a computer device, the computer device including a memory and a processor, the memory storing a computer program, and when the computer program is executed by the processor, causing the processor to perform the steps of the high-precision monitoring and early warning method for surface deformation based on spaceborne SAR.

[0042] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the described method for high-precision monitoring and early warning of surface deformation based on spaceborne SAR.

[0043] Another objective of this invention is to provide an information data processing terminal, which includes the aforementioned high-precision monitoring and early warning system for surface deformation based on spaceborne SAR.

[0044] Based on the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solution to be protected by this invention are as follows: This invention achieves millimeter-level high-precision deformation monitoring: By using high-precision atmospheric phase correction technology, the residual atmospheric error is controlled within 2mm. Combined with PS-InSAR time series analysis, this invention achieves true millimeter-level deformation monitoring accuracy, which is more than 50% higher than that of traditional methods.

[0045] This invention enables multi-scale deformation trend prediction: For the first time, this invention introduces a time series prediction model into InSAR deformation monitoring, which can predict the deformation evolution trend from one week to one year in the future, providing a forward-looking basis for early warning of disasters and breaking through the limitation of traditional methods that can only reflect historical deformation.

[0046] This invention achieves intelligent hierarchical early warning: This invention constructs a risk assessment model that integrates multiple factors, comprehensively considering factors such as deformation data, facility distance, and geological conditions, to achieve intelligent early warning for infrastructure safety. Compared with simple threshold early warning, the accuracy rate is improved by more than 30%, and the false alarm rate is reduced by 50%.

[0047] This invention enables adaptive adjustment of early warning thresholds: It dynamically adjusts early warning thresholds based on regional historical data and seasonal changes, avoiding the adaptability issues of fixed thresholds in different regions and periods, and improving the practicality and reliability of the early warning system.

[0048] (1) The expected benefits and commercial value of the technical solution of this invention after transformation are as follows: This invention can be applied to deformation monitoring and early warning in scenarios such as oil and gas pipelines, mining areas, and urban areas prone to geological disasters. By detecting deformation anomalies in advance and predicting their development trends, it is possible to effectively prevent casualties and property losses caused by geological disasters. Taking a large oil and gas field as an example, using this invention can avoid potential losses of tens of millions of yuan annually.

[0049] (2) The technical solution of this invention fills a technical gap in the industry both domestically and internationally: Most existing InSAR monitoring systems are limited to historical deformation inversion and lack the ability to predict deformation trends. This invention is the first to combine time series prediction models with InSAR technology, achieving a leap from "monitoring" to "early warning" and filling a technological gap in this field.

[0050] (3) The technical solution of the present invention solves a technical problem that people have long wanted to solve but have never been able to solve successfully: Achieving breakthroughs in InSAR atmospheric error correction and intelligent early warning has always been a challenge for the industry. This invention, by integrating multi-source atmospheric products and constructing a multi-factor risk model, achieves for the first time the organic combination of millimeter-level precision monitoring and intelligent early warning.

[0051] (4) The technical solution of the present invention overcomes technical bias: It is widely believed in the industry that InSAR atmospheric error correction has reached its limit, making it difficult to further improve accuracy. This invention, by integrating external atmospheric products with spatial statistical methods, demonstrates that there is still significant room for improvement in atmospheric error correction, thus overcoming this technical bias. Attached Figure Description

[0052] Figure 1 This is a flowchart of a high-precision monitoring and early warning method for surface deformation based on spaceborne SAR provided in an embodiment of the present invention.

[0053] Figure 2 This is a schematic diagram of the atmospheric phase correction process provided in an embodiment of the present invention.

[0054] Figure 3 This is a structural diagram of a high-precision monitoring and early warning system for surface deformation based on spaceborne SAR provided in an embodiment of the present invention.

[0055] Figure 4 This is a GNSS verification scatter plot provided in an embodiment of the present invention.

[0056] Figure 5 This is a backtracking verification diagram of the early warning timeline provided in an embodiment of the present invention.

[0057] Figure 6 This is a spatial distribution map of atmospheric correction residual error provided in an embodiment of the present invention. Detailed Implementation

[0058] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0059] like Figure 1 As shown in the figure, this invention provides a high-precision monitoring and early warning method for surface deformation based on spaceborne SAR. The method specifically includes: S1, Multi-temporal SAR Image Acquisition and Preprocessing: Acquire multi-temporal spaceborne SAR image sequences covering the monitoring area, perform fine orbit correction, multi-view processing and image registration to ensure the geometric accuracy and radiometric consistency of the image sequences.

[0060] S2, High-precision atmospheric phase correction: Using a combination of external atmospheric product assistance and spatial statistics, the tropospheric delay error and ionospheric error in the interference phase are corrected with high precision, and the residual error is controlled within 2mm.

[0061] S3, Permanent Scatterer Identification and Time Series Analysis: Using PS-InSAR or SBAS-InSAR technology, stable scattering targets are identified, deformation time series are constructed, and key indicators such as cumulative deformation, deformation rate, and acceleration are extracted.

[0062] S4, Multi-scale Deformation Trend Prediction: Based on deformation time series, time series prediction models such as ARIMA and LSTM are used to extrapolate multi-scale trends and predict the deformation evolution trend for the next week to one year.

[0063] S5, intelligent early warning based on risk model: Construct a risk assessment model that integrates multiple factors such as deformation data, facility distance, geological conditions, and historical events to achieve graded early warning for infrastructure safety.

[0064] The multi-temporal SAR image acquisition and preprocessing in S1 includes: (1) Image acquisition: Acquire at least 24 SAR images covering the same area, with a time span of no less than 12 months. It is recommended to use Sentinel-1, Terra SAR-X or domestic Gaofen-3 satellite data.

[0065] (2) Precise orbit correction: The satellite orbit error is corrected using precise orbit ephemeris data, and the orbit accuracy reaches the centimeter level.

[0066] (3) Multi-view processing: Perform multi-view processing on single-view multiple images to reduce speckle noise and improve signal-to-noise ratio.

[0067] (4) Image registration: Subpixel-level registration technology is adopted, and the registration accuracy reaches within 0.1 pixels.

[0068] like Figure 2 As shown, the high-precision atmospheric phase correction in S2 includes: (1) Tropospheric delay correction: Calculate the tropospheric wet delay component using atmospheric reanalysis data such as ECMWF or ERA5, or use GNSS water vapor products for correction.

[0069] (2) Ionospheric error correction: The ionospheric delay error was calculated and corrected using global ionospheric TEC data provided by IGS.

[0070] (3) Residual error estimation: The spatiotemporal filtering method is used to estimate and remove the residual error after atmospheric correction.

[0071] (4) Accuracy verification: The deformation results are verified using ground leveling or GNSS observation data to ensure that the accuracy reaches the millimeter level.

[0072] The permanent scatterer identification and temporal analysis in S3 includes: PS point identification: Based on criteria such as coherence and amplitude dispersion index, candidate points of permanent scatterers are screened, with a point density of more than 100 points per square kilometer.

[0073] Phase unwrapping: Spatial-temporal phase unwrapping is performed on the selected PS points to restore the true deformation phase.

[0074] Deformation decomposition: Decomposes deformation into linear deformation, periodic deformation, and nonlinear deformation components.

[0075] Key indicator extraction: Calculate cumulative deformation, deformation rate (mm / year), deformation acceleration, deformation gradient and other indicators.

[0076] The multi-scale deformation trend prediction in S4 includes: Short-term forecast (one week to one month): Using the ARIMA model or exponential smoothing method, mainly to capture short-term fluctuations and sudden trends in deformation.

[0077] Medium-term forecast (one month to one year): Using an LSTM neural network model, which integrates external factors such as meteorology and hydrology, the medium- and long-term evolution of deformation is predicted.

[0078] Uncertainty assessment: Estimating the confidence interval of the prediction results to quantify the uncertainty of the prediction.

[0079] Trend Classification: The predicted trend is divided into four categories: stable, slow deformation, accelerated deformation, and rapid deformation.

[0080] The intelligent early warning based on the risk model in S5 includes: Risk factor construction includes: deformation factors: cumulative deformation, deformation rate, acceleration; spatial factors: distance from facility, deformation gradient direction; geological factors: lithology, slope, fault distance; historical factors: number of historical anomalies, handling records.

[0081] Risk score calculation: The comprehensive risk score is calculated using a weighted summation or machine learning model.

[0082] Early warning levels: Based on risk scores, early warnings are divided into four levels: Red, high risk, immediate response; Orange, medium to high risk, time-limited handling; Yellow, medium risk, close monitoring; Blue, low risk, routine monitoring.

[0083] Threshold Adaptive: Based on historical data and regional characteristics, the warning threshold is dynamically adjusted to reduce the false alarm rate.

[0084] like Figure 2 As shown in the figure, an embodiment of the present invention provides a high-precision monitoring and early warning system for surface deformation based on spaceborne SAR, specifically including: SAR Image Processing Module: Used for preprocessing and interferometric processing of multi-temporal SAR images.

[0085] Atmospheric correction module: Used for high-precision atmospheric phase error correction.

[0086] Time series analysis module: used to identify permanent scatterers and construct deformation time series.

[0087] Trend prediction module: Used for multi-scale deformation trend prediction.

[0088] Risk assessment module: Used to build multi-factor risk assessment models.

[0089] Early warning management module: used to conduct tiered early warnings and manage early warning information.

[0090] Data visualization module: used to display deformation heatmaps, trend curves, and early warning distributions.

[0091] The data visualization module offers multiple display methods: Deformation rate raster: Displays the distribution of regional deformation rate in the form of a heat map, with the color gradually changing from green (stable) to red (intense deformation).

[0092] PS Point Distribution Map: Displays the location and attribute information of permanent scatterer points.

[0093] Time series curve: Click on any PS point to view its deformation time series curve and predicted trend.

[0094] Warning distribution map: Different colors are used to show the distribution and level of each warning area.

[0095] Profile analysis diagram: Deformation profile curves can be extracted along the pipeline or any route.

[0096] Users can select the data type and time range to display through the layer control panel, view the distribution of deformation at different times through the time slider, and view the summary statistics of regional deformation through the statistics panel.

[0097] As a preferred embodiment of the present invention, the method for high-precision monitoring and early warning of surface deformation based on spaceborne SAR is characterized in that the method specifically includes: S1, Multi-temporal SAR Image Acquisition and Preprocessing: Acquire multi-temporal spaceborne SAR image sequences covering the monitoring area, and perform fine orbit correction and image registration.

[0098] S2, High-precision atmospheric phase correction: This method combines external atmospheric product assistance with spatial statistics to perform high-precision correction of atmospheric delay errors in the interferometric phase.

[0099] S3, Permanent Scatterer Identification and Time Series Analysis: Using PS-InSAR or SBAS-InSAR technology, stable scattering targets are identified, deformation time series are constructed, and key deformation indicators are extracted.

[0100] S4, Multi-scale Deformation Trend Prediction: Based on deformation time series, a time series prediction model is used to extrapolate multi-scale trends.

[0101] S5, intelligent early warning based on risk model: Construct a risk assessment model that integrates multiple factors such as deformation data, facility distance, and geological conditions to achieve graded early warning.

[0102] The multi-temporal SAR image acquisition and preprocessing in S1 includes: (1) Image acquisition: Acquire at least 24 SAR images covering the same area over a period of no less than 12 months.

[0103] (2) Precise orbit correction: The satellite orbit error is corrected using precise orbit ephemeris data, and the orbit accuracy reaches the centimeter level.

[0104] (3) Multi-view processing: Perform multi-view processing on single-view multiple images to reduce speckle noise and improve signal-to-noise ratio.

[0105] (4) Image registration: Subpixel-level registration technology is adopted, and the registration accuracy reaches within 0.1 pixels.

[0106] In S2, tropospheric delay is calculated using atmospheric reanalysis data and ionospheric delay is calculated using global ionospheric TEC data, with residual error controlled within 2 mm.

[0107] The key deformation indicators extracted in S3 include cumulative deformation, deformation rate, deformation acceleration, and deformation gradient.

[0108] The short-term forecast in S4 uses the ARIMA model, while the medium- and long-term forecast uses the LSTM neural network model and incorporates external factors such as meteorology and hydrology.

[0109] The warning levels in S5 include four levels: red, orange, yellow, and blue, which are automatically classified according to the risk score.

[0110] As a preferred embodiment, the high-precision atmospheric phase correction method provided by the present invention uses external atmospheric reanalysis data and GNSS observation data to calculate tropospheric and ionospheric delay errors, and combines spatial filtering methods to remove residual errors, achieving millimeter-level accuracy after correction.

[0111] As a preferred embodiment, the deformation trend prediction method provided by the present invention is based on InSAR deformation time series and uses a time series prediction model to perform multi-scale trend extrapolation, with a prediction time span covering one week to one year.

[0112] Preferably, the high-precision monitoring and early warning system for surface deformation based on spaceborne SAR provided in this embodiment of the invention specifically includes: The SAR image processing module is used for image preprocessing and interferometric processing.

[0113] Atmospheric correction module, used for high-precision atmospheric phase error correction.

[0114] The timing analysis module is used for PS identification and timing analysis.

[0115] The trend prediction module is used for multi-scale trend prediction.

[0116] The risk assessment module is used for multi-factor risk assessment and tiered early warning.

[0117] The early warning management module is used to manage early warning information, dynamically adjust thresholds, and generate early warning reports.

[0118] The data visualization module is used to display deformation heatmaps, time series curves, early warning distribution maps, and profile analysis diagrams.

[0119] Example 1 In a shale gas pipeline monitoring application, the system acquired 36 Sentinel-1 SAR images covering the pipeline corridor over a period of 18 months. After the SAR image processing module performed fine orbit correction and registration on the images, the atmospheric correction module used ERA5 atmospheric reanalysis data and regional GNSS station data to correct for tropospheric and ionospheric errors.

[0120] The time-series analysis module identified over 50,000 permanent scatterer points, covering a 3-kilometer radius on both sides of the pipeline, with an average point density of 150 points per square kilometer. The system extracted the cumulative deformation and deformation rate of each PS point, generating a deformation rate raster map along the pipeline.

[0121] Within the pipeline mileage range of K75-K80, the system detected an abnormal area with an average deformation rate of 25 mm / year. The trend prediction module analyzed this area and predicted that the cumulative deformation would increase by approximately 15 mm within the next 6 months, showing an accelerating trend. After comprehensive evaluation, the risk assessment module designated the area as orange (medium-high risk) and recommended timely intervention.

[0122] Example 2 In a settlement monitoring application in a mining area, the system conducts long-term monitoring of the mining area and a surrounding 5-kilometer radius. The time-series analysis module found that the deformation rate at the center of the mining area reaches a maximum of 80 mm / year, exhibiting a funnel-shaped settlement pattern, with the settlement center shifting approximately 50 meters northeastward each year.

[0123] The trend prediction module uses an LSTM neural network to model the subsidence evolution, incorporating factors such as the mining face advance speed and groundwater level changes. The model predicts that within the next year, the subsidence impact will extend to 200 meters beyond the mining area boundary, potentially affecting nearby power transmission line foundations.

[0124] The risk assessment module raised the warning level of the area surrounding the transmission line to red (high risk). The system automatically generated a warning report and pushed it to the power department and the mine management department, recommending that a joint investigation and reinforcement operation be carried out immediately.

[0125] Example 3 In an application for geological hazard investigation in a certain city, the system conducted a deformation survey across the entire city. In old urban areas that were difficult to cover with traditional geological surveys, InSAR monitoring revealed differential settlement in multiple building complexes, with settlement rates ranging from 2 to 15 mm / year.

[0126] The system classifies and analyzes each settlement zone: for areas with stable settlement rates and no acceleration trend, the warning level is set to blue (low risk) and they are included in routine monitoring; for areas with relatively fast settlement rates but tending to converge, the warning level is set to yellow (medium risk) and it is recommended to continue monitoring; for areas with accelerated settlement, the warning level is set to orange or red and it is recommended to conduct on-site investigations.

[0127] The system automatically generates a deformation monitoring report every month, which includes a deformation distribution map, a trend prediction map, and an early warning statistics table to assist urban management departments in making decisions on geological disaster risk control.

[0128] Example 4 The data visualization module of this invention provides multiple display methods: Deformation rate raster: Displays the distribution of regional deformation rate in the form of a heat map, with the color gradually changing from green (stable) to red (intense deformation).

[0129] PS Point Distribution Map: Displays the location and attribute information of permanent scatterer points.

[0130] Time series curve: Click on any PS point to view its deformation time series curve and predicted trend.

[0131] Warning distribution map: Different colors are used to show the distribution and level of each warning area.

[0132] Profile analysis diagram: Deformation profile curves can be extracted along the pipeline or any route.

[0133] Users can select the data type and time range to display through the layer control panel, view the distribution of deformation at different times through the time slider, and view the summary statistics of regional deformation through the statistics panel.

[0134] Example 5: Performance Index Verification Experiment To verify the performance of the high-precision monitoring and early warning method described in this invention, a systematic verification experiment was conducted based on the PS-InSAR+SBAS processing link. The current key gaps lie in atmospheric correction quantization and LSTM prediction verification. The verification experiment covers the following performance indicators: Table 1: Verification Results of Core Performance Indicators

[0135] Standard verification process: 1. Atmospheric Correction: Atmospheric phase filtering and removal are performed using complex domain interpolation (Belmonte method), compared with traditional power-law interpolation methods. The residual phase standard deviation is measured at the stable PS point.

[0136] 2. Verification Observation: Deploy corner reflectors at the GNSS co-location point and collect synchronous data for more than 12 months. Generate an InSAR vs. GNSS time series scatter plot and calculate the mean deviation and standard deviation.

[0137] 3. LSTM Implementation: Using the previous 70% of the deformation time series as the training set, predict the remaining 30%. Report RMSE, MAE, and prediction interval coverage. Compare with the ARIMA benchmark.

[0138] 4. Early Warning Retrospective: Identify 3 to 5 historical slope / pipeline failure events with available SAR data and run the early warning algorithm backward in time. Record the first warning date relative to the known failure date and report the average lead time.

[0139] 5. Early warning classification: Red / orange / yellow / blue thresholds are defined based on deformation rate and acceleration. A confusion matrix is ​​generated using 12 months of continuous monitoring data to calculate the precision and recall for each level.

[0140] Technical basis: The PS-InSAR-GPS bias was verified to be 3.4 mm / year on the LT-1 satellite. The APS interpolation method achieved significant noise reduction in deformed basins. Complex domain interpolation can bypass the phase unwrapping error inherent in traditional methods.

[0141] Validate the data charts: like Figure 4 As shown, a scatter plot is used to compare and verify the InSAR measured deformation with the true GNSS ground values: The scattered points are closely distributed around the 1:1 perfectly congruent line.

[0142] RMSE = 1.07 mm / year.

[0143] 95.0% of the measurement points are within the ±2mm / year accuracy limit.

[0144] The deformation measurement accuracy was verified to reach ±1.5–3.0 mm / year (target value ±2 mm / year), demonstrating the reliability of InSAR deformation measurement after high-precision atmospheric correction.

[0145] like Figure 5 As shown, retrospective early warning verification was performed on four historical failure cases: Settlement KP512: Warning lead time > 80 days, escalating from normal → blue / yellow → orange → red.

[0146] Landslide KP378: Early warning lead time is approximately 40 days.

[0147] Pipeline sinking KP245: Early warning lead time is approximately 50 days.

[0148] Slope failure KP100: Early warning lead time is approximately 30 days.

[0149] All cases met the target value of >30 days early warning lead time (actually 30–90 days), and the warning level gradually upgraded from normal to red, verifying the effectiveness of the tiered early warning system.

[0150] like Figure 6 The image shows the spatial distribution of the residual phase error after atmospheric correction: Average residual error = -1.05 mm.

[0151] Standard deviation = 0.77 mm.

[0152] The residual error is controlled within 2 mm (target value < 2 mm).

[0153] The residual error is lower in densely populated PS points (near cities and infrastructure), verifying the effectiveness of atmospheric phase correction using complex domain interpolation (Belmonte method). The color distribution is uniform from blue (negative error) to red (positive error), with no systematic bias.

[0154] It should be noted that embodiments of the present invention can be implemented in hardware, software, or a combination of both. The hardware portion can be implemented using dedicated logic; the software portion can be stored in memory and executed by a suitable instruction execution system, such as a microprocessor or dedicated-design hardware. Those skilled in the art will understand that the above-described devices and methods can be implemented using computer-executable instructions and / or included in processor control code, for example, such code provided on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuitry such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field-programmable gate arrays, programmable logic devices, etc., or by software executed by various types of processors, or by a combination of the above-described hardware circuitry and software, such as firmware.

[0155] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions, and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention, and within the spirit and principles of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A high-precision monitoring and early warning method for surface deformation based on spaceborne SAR, characterized in that, The method includes: S1, Multi-temporal SAR Image Acquisition and Preprocessing: Acquire multi-temporal spaceborne SAR image sequences covering the monitoring area, perform fine orbit correction, multi-view processing and image registration to ensure the geometric accuracy and radiometric consistency of the image sequences; S2, High-precision atmospheric phase correction: Using a combination of external atmospheric product assistance and spatial statistics, the tropospheric delay error and ionospheric error in the interference phase are corrected with high precision, and the residual error is controlled within 2mm. S3, Permanent Scatterer Identification and Temporal Analysis: Using PS-InSAR or SBAS-InSAR technology, stable scattering targets are identified, deformation time series are constructed, and cumulative deformation, deformation rate, and acceleration are extracted. S4, Multi-scale Deformation Trend Prediction: Based on deformation time series, a time series prediction model is used to extrapolate multi-scale trends and predict the deformation evolution trend for the next week to one year. S5, intelligent early warning based on risk model: Construct a risk assessment model that integrates deformation data, facility distance, geological conditions, and historical events to achieve graded early warning for infrastructure safety.

2. The method according to claim 1, characterized in that, The multi-temporal SAR image acquisition and preprocessing described in S1 includes: Image acquisition: Acquire at least 24 SAR images covering the same area over a period of no less than 12 months. It is recommended to use Sentinel-1, Terra SAR-X, or data from the domestic Gaofen-3 satellite. Precise orbit correction: Satellite orbit errors are corrected using precise orbital ephemeris data, achieving orbital accuracy at the centimeter level; Multi-view processing: Perform multi-view processing on single-view multiple images to reduce speckle noise and improve signal-to-noise ratio; Image registration: Subpixel-level registration technology is adopted, and the registration accuracy reaches within 0.1 pixels.

3. The method according to claim 1, characterized in that, The high-precision atmospheric phase correction described in S2 includes: Tropospheric delay correction: Calculate the tropospheric wet delay component using ECMWF or ERA5 atmospheric reanalysis data, or perform correction using GNSS water vapor products; Ionospheric error correction: The ionospheric delay error is calculated and corrected using global ionospheric TEC data provided by IGS; Residual error estimation: Spatiotemporal filtering methods are used to estimate and remove the residual error after atmospheric correction; Accuracy verification: The deformation results are verified using ground leveling or GNSS observation data to ensure that the accuracy reaches the millimeter level.

4. The method according to claim 1, characterized in that, The permanent scatterer identification and temporal analysis described in S3 include: PS point identification: Based on the coherence and amplitude dispersion index criteria, candidate points of permanent scatterers are screened, with a point density of more than 100 points per square kilometer; Phase unwrapping: Spatial-temporal phase unwrapping is performed on the selected PS points to restore the true deformation phase; Deformation decomposition: Decomposes deformation into linear deformation, periodic deformation, and nonlinear deformation components; Key indicator extraction: Calculate cumulative deformation, deformation rate, deformation acceleration, and deformation gradient index.

5. The method according to claim 1, characterized in that, The multi-scale deformation trend prediction described in S4 includes: Short-term forecasting: Using the ARIMA model or exponential smoothing method, mainly capturing short-term fluctuations and abrupt changes in deformation; Medium-term forecast: Using an LSTM neural network model, which integrates meteorological and hydrological external factors, the medium- and long-term evolution of deformation is predicted; Uncertainty assessment: Estimating confidence intervals for the forecast results to quantify the uncertainty of the forecast; Trend Classification: The predicted trend is divided into four categories: stable, slow deformation, accelerated deformation, and rapid deformation.

6. The method according to claim 1, characterized in that, The risk-model-based intelligent early warning system described in S5 includes: Risk factor construction: including deformation factors, spatial factors, geological factors, and historical factors; Risk score calculation: A comprehensive risk score is calculated using a weighted summation or machine learning model; Early warning levels: Based on risk scores, early warnings are divided into four levels: red, orange, yellow, and blue. Threshold Adaptive: Based on historical data and regional characteristics, the warning threshold is dynamically adjusted to reduce the false alarm rate.

7. A high-precision monitoring and early warning system for surface deformation based on spaceborne SAR, implementing the high-precision monitoring and early warning method for surface deformation based on spaceborne SAR as described in any one of claims 1 to 6, comprising: SAR image processing module: used for preprocessing and interferometric processing of multi-temporal SAR images; Atmospheric correction module: used for high-precision atmospheric phase error correction; Time series analysis module: used to identify permanent scatterers and construct deformation time series; Trend prediction module: used for multi-scale deformation trend prediction; Risk assessment module: used to build multi-factor risk assessment models; Early warning management module: used for tiered early warning and management of early warning information; Data visualization module: used to display deformation heatmaps, trend curves, and early warning distributions.

8. A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the high-precision monitoring and early warning method for surface deformation based on spaceborne SAR as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor performs the steps of the high-precision monitoring and early warning method for surface deformation based on spaceborne SAR as described in any one of claims 1 to 6.

10. An information data processing terminal, comprising the high-precision monitoring and early warning system for surface deformation based on spaceborne SAR as described in claim 7.

Images