Date of Award
2024
Degree Type
Dissertation
Degree Name
Doctor of Philosophy in Oceanography
Specialization
Marine Geology and Geophysics
Department
Oceanography
First Advisor
Meng (Matt) Wei
Abstract
Urban subsidence is a critical issue that poses significant risks to cities worldwide. This research leverages Interferometric Synthetic Aperture Radar (InSAR) technology to provide a comprehensive analysis of subsidence patterns and impacts across various urban contexts. The study is presented through three interconnected manuscripts, each addressing different aspects of subsidence.
Globally, subsidence in coastal cities is analyzed using Sentinel-1 data from 2015 to 2020. In Manuscript 1, we constructed a subsidence map for 99 major coastal cities, revealing that many of these cities are subsiding at rates that surpass sea level rise. Human activities, particularly groundwater extraction, are identified as primary drivers of this subsidence. Our findings highlight the urgent need for comprehensive monitoring and policy interventions to mitigate the associated risks of flooding and infrastructure damage.
Focusing on the Taipei Basin, Manuscript 2 examines vertical ground deformation from December 2018 to September 2023. Despite a ban on domestic groundwater extraction since the 1970s, recent years have seen significant subsidence events from December 2018 to June 2021, exacerbated by the 2020 drought. This subsidence was followed by rapid uplift as groundwater levels recovered, illustrating the dynamic interplay between climatic events and human activities. The study underscores the importance of monitoring both natural and anthropogenic factors to manage urban subsidence effectively. This pattern of subsidence and uplift parallels phenomena observed in drought-affected California, where groundwater level changes significantly impact surface deformation.
In New York City, as detailed in Manuscript 3, our analysis identifies post-glacial rebound as the primary driver of current subsidence, given that most buildings are constructed on bedrock. Using InSAR and GNSS data, we quantified subsidence rates and explored the contributions of building loads and other anthropogenic factors. However, the cumulative effects of urban infrastructure, particularly the heavy load of buildings and altered hydrogeological conditions due to urbanization, introduce additional subsidence challenges. This study uniquely maps the expected and observed subsidence, distinguishing between geological and anthropogenic factors. The findings underscore the need for integrated urban planning and policy adjustments as New York City, like many coastal cities globally, faces increased flood risks exacerbated by both natural subsidence and intensified urban development.
Manuscript 4 introduces a novel application of the U-Net deep learning architecture to enhance the accuracy of InSAR time series analysis, particularly in denoising data affected by atmospheric noise. By employing a convolutional neural network that maintains spatial integrity through an encoder-decoder structure with skip connections, our model significantly improves the detection of subtle subsidence signals obscured by traditional noise reduction methods. Tested using both synthetic and real-world data across various coastal cities, this method demonstrates superior noise reduction capabilities and an enhanced signal-to-noise ratio. The U-Net model's adaptability through transfer learning promises wide applicability in diverse environments. This approach improves the reliability of subsidence detection and has broad implications for urban planning and risk management in vulnerable coastal regions worldwide.
Collectively, these studies illuminate the multifaceted nature of urban subsidence, demonstrating the significant impact of both human activities and environmental factors. The insights gained are critical for developing effective strategies to manage and mitigate subsidence risks in urban areas globally. Enhanced monitoring, policy interventions, and advanced modeling techniques are essential for mitigating the adverse effects of subsidence and ensuring the resilience of urban infrastructures.
Recommended Citation
Wu, Pei-Chin, "USING INSAR TECHNOLOGY TO ASSESS THE IMPACT OF COASTAL SUBSIDENCE IN MAJOR CITIES GLOBALLY" (2024). Open Access Dissertations. Paper 1662.
https://digitalcommons.uri.edu/oa_diss/1662