TL;DR
This study uses time-series InSAR to map urban land subsidence in Wuhan, China, revealing significant spatial heterogeneity and identifying key drivers like urban construction and hydrogeological conditions. A novel risk assessment model, integrating Geodetector-FAHP and Entropy Weight Method, effectively pinpoints high-risk areas, offering crucial insights for urban hazard mitigation.
What Problem Does This Paper Address?
Urban land subsidence is a growing geological hazard with significant negative impacts on infrastructure and daily life. While conventional monitoring methods exist, they often have limitations in terms of cost and comprehensive coverage. Time-series InSAR offers a powerful alternative for large-scale, high-resolution monitoring. However, understanding the risk associated with subsidence, which involves both the hazard (likelihood and severity of the event) and vulnerability (potential for damage), requires more than just deformation mapping. Previous studies in Wuhan have focused on mapping subsidence but lacked a holistic risk assessment that integrates multiple contributing factors and provides actionable insights for mitigation. This paper aims to fill this gap by not only mapping subsidence but also developing and applying a refined risk assessment framework for Wuhan.
Methods
The study employs a multi-faceted approach:
- Time-Series InSAR (SBAS-InSAR): This technique was used to process Sentinel-1A SAR images from 2020 to 2022, generating detailed maps of land subsidence rates and cumulative displacement. SBAS-InSAR is favored for its ability to overcome spatial decorrelation issues and provide accurate deformation measurements.
- Geodetector-Fuzzy Analytic Hierarchy Process (GD-FAHP): This novel approach was developed to assess the hazard component of subsidence risk. Geodetector is used to quantify the explanatory power of different factors (e.g., subsidence velocity, soft soil distribution) on subsidence, and its outputs are then used to build a fuzzy judgment matrix for FAHP. This combination aims to reduce the subjectivity inherent in traditional AHP by incorporating data-driven factor importance.
- Entropy Weight Method (EWM): This method was used to determine the weights of indicators for the vulnerability assessment. EWM objectively assigns weights based on the variability of indicators, ensuring that factors with greater information content have a stronger influence on the vulnerability assessment.
- Integrated Risk Assessment Model: The hazard and vulnerability indices were combined using a simple multiplication (Risk = Hazard × Vulnerability) with equal weighting, creating a comprehensive risk map for Wuhan.
Datasets
- Sentinel-1A SAR Images: 36 ascending images acquired over Wuhan from January 2020 to December 2022.
- SRTM DEM: 30m resolution Digital Elevation Model for topographic phase removal.
- Precise Orbit Files: Provided by the European Space Agency (ESA) for orbital correction.
- Wuhan Metro System Data: Sourced from Wuhan Metro Group Co., Ltd.
- Spatial Datasets: Soft soil and carbonate rock distributions from the Hubei Geological Bureau.
- Rainfall Statistics: From the Wuhan Water Resources Bulletin.
- Water Level Data: From the Department of Water Resources of Hubei Province.
- Population Density and Unit GDP Data: From the Wuhan Statistical Yearbook 2022.
- Building and Road Vector Data: Downloaded from OpenStreetMap.
- Leveling Data: 69 points provided by the Wuhan Geomatic Institute for accuracy verification.
Key Insights and Ideas
- Urban Construction as a Primary Driver: The study strongly links urban infrastructure development, particularly deep foundation pit excavations and subway construction, to land subsidence. This highlights the critical need for careful planning and execution of large-scale construction projects.
- Synergistic Effects of Geology and Human Activity: The research suggests that the combination of geological factors (soft soil, carbonate rock belts) and anthropogenic influences (groundwater extraction, construction) significantly exacerbates subsidence. Hydrogeological conditions are identified as a prerequisite for subsidence occurrence.
- Dynamic Nature of Subsidence: Subsidence areas are not static; they are observed to be expanding, with new zones emerging. This dynamic behavior necessitates continuous monitoring and adaptive management strategies.
- Refined Risk Assessment Framework: The integration of Geodetector with FAHP and EWM offers a more objective and data-driven approach to assessing land subsidence risk, moving beyond simple deformation mapping to consider multiple contributing factors and their interactions.
Key Results
- Maximum Subsidence Rate: -49 mm/a, with significant spatial heterogeneity.
- Maximum Cumulative Displacement: -135 mm observed in Hongshan District.
- InSAR Accuracy: Validation with leveling data showed a correlation coefficient of 0.71, indicating the reliability of the SBAS-InSAR results. Average RMSE of deformation rate was 3.06 mm/a.
- High-Risk Zones Identified: The risk assessment model pinpointed high-risk areas concentrated in central urban zones, particularly Hongshan and Wuchang Districts, and also highlighted areas along the Yangtze and Han Rivers.
- Vulnerability Factors: Unit GDP density was found to have the highest weighting in the vulnerability assessment, followed by population density.
Limitations and Future Work
- Temporal Resolution Discrepancies: Minor temporal differences between InSAR and leveling data were noted as a reason for some discrepancies in accuracy validation.
- Equal Weighting in Risk Model: The study used equal weights for hazard and vulnerability in the final risk assessment. Future work could explore more nuanced weighting schemes.
- Data Availability: While the paper lists data sources, the accessibility and resolution of some datasets might influence future studies.
- Future Work Suggestions: The paper emphasizes the need for targeted monitoring and mitigation efforts in high-risk areas. The proposed methodology is presented as applicable to similar urban environments, suggesting its potential for broader use. Further research could delve deeper into the specific mechanisms of subsidence in different geological settings and explore more advanced mitigation strategies.
Summary updated