Application of LuTan-1 SAR Data in Railway Subsidence Monitoring

InSAR technology is currently a crucial tool for large-scale surface deformation monitoring, particularly excelling in regional subsidence monitoring. In this study, focusing on the Jinan section of the Shandong railway, newly launched LUTAN-1 satellite SAR data was employed. DInSAR and Stacking techniques were applied to analyze the subsidence in response to the long-term operation load of high-speed trains and surrounding human activities. Through the analysis of data from June 2023 to December 2023, regional subsidence was identified in this section, with a subsidence rate reaching 8 cm/year. Comparative analysis between DInSAR technology monitoring results and precise leveling monitoring results showed consistency in the subsidence deformation trend in the region, achieving centimeter-level deformation monitoring in the study area.The LuTan-1 satellite provided robust SAR data support for railway subsidence monitoring, offering substantial reliability and accuracy to deepen the understanding of railway subsidence issues. These research findings hold significant practical implications for adopting timely maintenance and repair measures, ensuring the safety and reliable operation of the railway system.


Background 1.1 Introduction
The L-band Differential Interferometric Synthetic Aperture Radar (DInSAR) satellite, also known as "LuTan-1," is China's first civil L-band SAR satellite primarily designed for terrain mapping and deformation monitoring.The LUTAN-1 constellation supports two different flight formations, as illustrated in Figure 1.The first is a dual-satellite formation used for topographic mapping and corresponding data collection.The second mode is a tandem flight mode, where the satellites follow each other to maintain a phase difference of 180°, providing single-satellite observational data for deformation monitoring.LUTAN-1 exhibits excellent performance in deformation monitoring, owing to its longer wavelength, which imparts robust penetration capability and good penetration through vegetation and cloud cover, making it suitable for areas with dense vegetation.The twin LUTAN-1 satellites operate on the same orbital plane, allowing for a differential interferometric observation with 3m high-resolution image data of the same area within a 4-day interval, ensuring high consistency in the acquired data.Therefore, utilizing InSAR technology enables the monitoring of ground deformation with centimeter to even millimeter precision (High, W., 2020).These two satellites were successfully launched on January 26, 2022, and February 28, 2022, at the Jiuquan Satellite Launch Center.The SAR sensor is equipped with six different imaging modes featuring various spatial resolutions and imaging widths, with the key parameters outlined in Table 1.LuTan-1 has demonstrated its versatility in multiple applications.Starting from April 2023, the LuTan-1 satellite will consistently provide high-quality observational data, offering support for global natural resource monitoring, particularly in disaster monitoring and early warning applications (Gao, W., Gan, J., & Wang, C, 2015).
Figure 1.Satellite constellation art view and related products Railways, as the backbone of the comprehensive transportation system, are a pioneering field for accelerating the construction of a strong transportation nation and an important support for the construction of Chinese-style modernization.By 2025, China National Railway Group Co., Ltd.(hereinafter referred to as: China Railway) will have basically completed the "six modern systems" and achieved the goals of the railway's "14th Five-Year" development plan.As China's railway, especially high-speed railway, continues to increase its operational mileage, improving the safety environment along the railway becomes more prominent in ensuring high-quality railway development and the safety of people's lives and property.On the one hand, the stability of the railway subgrade is affected by environmental factors such as the geological structure along the line and regional surface subsidence; on the other hand, the deformation and damage of the track slab caused by the combined effects of internal factors such as subgrade static load and train dynamic load also pose a great threat to the stability of the subgrade.Therefore, this study conducted a preliminary assessment of SAR data for deformation monitoring along the railway section in Jinan, Shandong.DInSAR (Li, H, 2015).technologywas employed to calculate the deformation magnitude along the railway (Zhang, T., Shi, Y., Wang, J., et al, 2021).The deformation characteristics of the ground surface in the railway and surrounding areas were extracted and analyzed.

DInSAR Method
DInSAR  (1) In the formula, where def respectively represent the deformation phase, flat Earth phase, atmospheric phase, orbit error phase, topographic phase, and decoherence/thermal noise phase.Using precise satellite orbit data and baseline information from interferometric pairs, the orbit error phase and flat Earth phase are removed.External DEM data is used to eliminate the topographic phase.The influence of decoherence/thermal noise phase can be mitigated through filtering.For the sake of simplicity in understanding and calculation, the impact of atmospheric delay phase is ignored here.Therefore, the deformation phase can be expressed as: In accordance with the principles of the InSAR method, it can be derived that: Therefore, the deformation phase at the ground point is: Finally, through transformation, the deformation change along the radar line of sight for the ground object can be determined r ∆ .

Stacking InSAR Method
Railway subsidence along the ground is a gradual process that typically requires long-term monitoring to fully understand surface changes.Time series InSAR technology, with its advantages of high spatiotemporal resolution, long-term monitoring, real-time monitoring, and sensitivity to periodic changes, makes it an ideal choice for railway subsidence monitoring.This method can provide railway managers with timely, accurate, and comprehensive surface deformation information, helping to prevent and promptly address railway subsidence issues and improve the safety and reliability of the railway system.
Commonly used methods include Stacking technology, Wavelet time series InSAR, Small Baseline Subset (SBAS) method, Permanent Scatterers (PS), and Distributed Scatterers (DS) algorithms.These methods provide high spatiotemporal resolution surface deformation information by integrating SAR observations from multiple times, effectively suppressing the effects of temporal and spatial decorrelation, DEM residuals, and atmospheric disturbances.They greatly compensate for the shortcomings of traditional differential interferometry and can obtain high-precision (from millimeter to sub-millimeter level) surface deformation rates and sequences more accurately.Time series SAR analysis offers long-term dynamic monitoring of surface subsidence along railways, identifying seasonal, periodic, and gradual subsidence characteristics, and provides high-precision subsidence rates and cumulative amounts.Time series SAR analysis helps to understand the cumulative effects and trends of subsidence, providing support for regional situation analysis.is the standard deviation of the average deformation rate.

Processing Strategy
In the handling of railway and surrounding ground subsidence, DInSAR can provide monitoring with high sensitivity and high spatiotemporal resolution.Organize and analyze the elevation reference data in the experimental area, filter SAR images covering the railway along the monitoring requirements, and conduct DInSAR processing on the experimental images.In the treatment of railway subsidence, DInSAR can provide high sensitivity and high spatiotemporal resolution monitoring.The key processing steps of DInSAR include image registration and resampling, interferogram generation, removal of flat Earth and topographic phase, interferogram filtering and quality assessment (Wang, Z., Xu, S., Wang, N., et al, 2018), and phase unwrapping, among others (Ji, Z., Bo, H., Wang, D., et al, 2019).The final result is the differential interferometric phase dominated by surface deformation, which is used to monitor the settlement conditions along railway lines and provides highprecision information on trends and cumulative settlement amounts.It is more sensitive to deformation in the vertical direction and is capable of long-term, large-scale deformation field monitoring at the centimeter level, or even millimeter-level deformation.The basic data processing flow of DInSAR is illustrated in the diagram below Figure 3:

Study Area
The research area is located within Jinan City in Shandong Province, China.The Jinan to Zaozhuang high-speed railway mainline has a total length of 265.57km, with a newly constructed mainline of 260.32 km and utilizing 5.25 km of existing infrastructure.The designed operating speed is 350 km per hour.The total length of newly constructed mainline bridges and tunnels is 214.222km, accounting for 81.3% of the total railway length.Among these, the bridge length is 178.027 km, representing 67.6% of the total length, and the tunnel length is 36.195km, constituting 13.7% of the total length(Ge, K., Wang, Y., Yu, C., et al, 2017).Additionally, there is a newly constructed Jinan East-Jiaozhou Yard connecting line of 6.812 single-track kilometers and a Jilai-Jibin connecting line of 6.097 single-track kilometers.The monitoring area is shown in Figure 4.

Dataset
This article primarily utilizes SAR data processing software developed independently by the National Satellite Remote Sensing Application Center of the Ministry of Natural Resources (Ma, T., Tan, H., Li, T., et al. 2021), known as LandSAR.For the surface deformation measurement experiment along the railway in Jinan, Shandong, a total of 15 scenes of Lutan-1 SAR images covering the area were selected as the data source, and on-site measurements were conducted for validation (Zhang, T., Shi, Y., Wang, J., et al, 2021).Among these, 15 scenes of follow-mode data were used for deformation measurement (Cheng, P., Wen, H., Liu, H., et al, 2019), and the configuration parameters for the data are provided in Table 3, including a strip imaging mode, a resolution of 3 m, and a swath width of 50 km × 50 km.

Results Analysis
The surface deformation field product along the railway (Chen, D., Lu, Y., Jia, D, 2018)line in the Jinan to Zaozhuang region occurred from June 2023 to December 2023 .Among them, the descending data has a maximum vertical baseline of 206 m and a maximum temporal baseline of 80 days.This combination meets the requirements for high-precision observations.In the first phase of deformation product production, 15 scenes of L-SAR data were used, processing 28 interferometric pairs, obtaining 48 deformation field products, and completing the deformation field product for this region.The results are shown in the Figure 5.

Accuracy verification
To verify the accuracy of the DInSAR results during the study period and assess the reliability of ground deformation monitoring using LuTan-1 data, this experiment selected GPS data from 30 points in the Jinan area of Shandong for accuracy assessment, the distribution of some collected points is shown in the Figure 8.The surface deformation monitored by the DInSAR method is in the direction of the SAR satellite line of sight, whereas the deformation (Shi, G., Chen, Q., Liu, X., et al, 2022) observed by GPS includes both vertical and horizontal components.In order to accurately assess the precision of using LuTan-1 data with the DInSAR method, it is necessary to convert the horizontal movement and vertical deformation of the ground observation station into the SAR satellite line of sight.The transformation formula is as follows: cos sin cos sin sin In the formula,

Conlusions
The results indicate that the DInSAR results obtained using LuTan-1 SAR data are highly consistent with on-site measurement results, providing accurate monitoring outcomes for the research area.Compared to leveling measurements, LuTan-1 DInSAR deformation monitoring demonstrates a superior accuracy of 5 cm (RMSE).Stacking result can be observed that Jinan-Zaozhuang railway line range from -8 cm/year to 44 cm/year, indicating a region with relatively significant subsidence.The quantitative results validate the effectiveness of LuTan-1 SAR data in railway deformation monitoring.Two comparisons show that InSAR monitoring technology has great potential for promotion in high-speed railway monitoring, making the use of InSAR monitoring methods a new choice for monitoring future railway settlement.
is an extension of InSAR technology.It involves the repeated observation of the Earth's surface undergoing deformation using spaceborne SAR systems.By differencing the interferometric phase and topographic phase of SAR images before and after deformation, DInSAR eliminates terraininduced phase and flat Earth phase, as well as mitigates atmospheric delay and noise phases, A simple schematic diagram of the radar differential interference process is shown in Figure2.This process yields the surface deformation phase in the radar line of sight (LOS), providing information about ground deformation.DInSAR technology is particularly sensitive to vertical deformation and can achieve long-term monitoring of centimeter-level or even millimeter-level deformations over large areas.Its high resolution and continuous spatial coverage distinguish it from other geodetic measurement methods such as Global Positioning System (GPS), Satellite Laser Ranging (SLR), and Very Long Baseline Interferometry (VLBI).

Figure 2 A
Figure 2 A simple schematic diagram of the radar differential interferometry process.In this diagram, R1 represents the slant range corresponding to the satellite's first pass, R2 represents the slant range corresponding to the satellite's second pass, and the difference between the two slant ranges (ΔR) represents the deformation parameter that InSAR aims to obtain.The interferometric phase of SAR images before and after deformation in the same area can be represented by the following formula: of Sight (LOS) average deformation rate, def φ is the deformation phase, i d is the deformation quantity for the i-th differential interferogram, i T ∆ is the time interval for the i-th interferogram, N denotes the number of differential interferograms involved in the solution, and ( ) V σ 

Figure 4 .
Figure 4. Monitoring Area and Image Coverage

Figure 5 .
Figure 5. Deformation Monitoring ResultsThe monitoring results of the research area from July 5 to September 15, 2023, are shown in Figure6.The railway section is located in mountainous terrain with various human activities, such as urban terraced fields.It can be observed that there are four suspected deformation hazards along the Jinan-Zaozhuang section, mainly related to slope stability.Due to the strong revisit capability and longer wavelength of the LuTan-1 satellite, with a 350 m baseline, significant subsidence along the railway line was detected using LuTan-1 SAR data.From July 5 to September 15, 2023, the maximum deformation amount (LOS direction) along the Jinan-Zaozhuang railway line reached 34 cm within a 48-day time interval.

Figure 6 .
Figure 6.Deformation Monitoring Results from July to SeptemberAfter calculating the deformation rate in Chengdu urban area using time-series InSAR technology, a buffer zone with a range

Figure 7 .
Figure 7. Stacking Strain Rate Products the north-south direction; θ represents the radar incidence angle; β represents the azimuth angle of the satellite.

Figure 8
Figure 8 Distribution of GPS measurement pointsStatistical analysis of the GPS and DInSAR results for corresponding points is shown in Figure9.Based on the measured GPS data, it was found that the deformation trends of the GPS and DInSAR results for corresponding points were generally consistent(Yao, W., Xu, K., Zhu, X., et al. 2021).However, there were some discrepancies in individual point values, which could be attributed to factors such as personnel and equipment, leading to relatively lower measurement accuracy.Further efforts will be made to improve the measurement precision through multiple or high-precision field measurements to obtain more accurate detection results.The statistical analysis indicates that the average absolute error of the DInSAR deformation results is 4.62 mm, and the root mean square error is 5.83 mm, meeting the design accuracy requirements.

Figure 9 .
Figure 9.Comparison of Deformation Values between DInSAR and GPS Corresponding Points

Table 2 .
Comparison of InSAR Technology with Traditional Geodetic Surveying Techniques has obvious advantages as the underlying data support for the informatization of railway external environment safety.The development and application of InSAR technology has provided an important technical method for high-speed railway deformation monitoring.Based on L-band differential interferometric SAR satellites, macroscopic inspection of geological disaster risks along the railway can be carried out, paving the way for the normalized application of satellite remote sensing technology in railway external environment safety monitoring.Its high resolution and feature of continuous spatial coverage are not present in geodetic methods such as Global Positioning System (GNSS), Satellite Laser Ranging (SLR), and Very Long Baseline Interferometry (VLBI).The advantages of InSAR technology over traditional surveying techniques are shown in Table2.