Urban landscapes appear solid, but beneath the concrete lies a highly dynamic geotechnical system. As cities grow vertically and horizontally, they exert massive mechanical and environmental pressures on the subsurface. Ground deformation—manifesting as gradual regional subsidence or sudden, localized sinkholes—is rarely driven by a single isolated factor. Instead, it emerges from a complex interplay between anthropogenic activities (such as underground transit construction, deep excavations, utility deterioration, and groundwater pumping) and inherent geological vulnerabilities (like soft clay strata, coastal silt, and carbonate karst belts).
To effectively mitigate these hazards, we must understand the underlying physical mechanisms that govern how human engineering alters subsurface stress and fluid dynamics.
Stress Redistribution from Underground Excavations
When we dig into the earth to build subways or deep foundations, we disrupt the existing stress equilibrium of the soil matrix. This process triggers a phenomenon known as stratum loss: the removal of subsurface volume causes the surrounding soil to shift inward to fill the void, creating a displacement field that propagates upward to the surface.
As investigated by Heng Zhang et al. in their study, this excavation-induced movement creates distinct deformation zones—ranging from localized settlement to structural uplift—around existing infrastructure like metro tunnels. The spatial propagation of this displacement typically follows a parabolic decay pattern relative to the distance from the excavation site, meaning its mechanical impact is highly concentrated but fades rapidly beyond a critical threshold, as demonstrated by Weixi Zhang et al. in their paper.
The risk is amplified in dense urban environments where external surface loads from high-rise buildings exert additional downward stress on the destabilized soil, a factor quantified by Yao Bai et al. in their research. When new shield tunnels are driven close to existing foundations in sensitive clay strata, the resulting soil displacement can cause adjacent buildings to tilt or crack, as shown by Liping Cai et al. in their analysis and Chun - lai Chen et al. in their simulation. To counter this, modern engineering relies on advanced soil stabilization methods, such as digital minor-disturbance deep mixing, which actively controls lateral soil displacement to within millimeters to protect nearby sensitive structures, as detailed by Ce Du et al. in their work.
Subsurface Erosion and Void Formation from Aging Infrastructure
While massive excavation projects cause acute, visible ground shifts, a more insidious driver of urban subsidence operates silently within municipal utility networks. Aging water and sewer infrastructure frequently develops undetected leaks, which act as catalysts for internal soil erosion.
When a pipe breaches, groundwater and surrounding soil particles are drawn into the conduit under seepage pressure. This fluid-mechanical process washes away the soil matrix, gradually carving out hidden subsurface cavities. Pilljae Kwak in their paper and Jinyoung Kim et al. in their study explain that these leaks cause soil to wash into pipes, creating hidden subsurface cavities that eventually lead to dangerous road collapses. Laboratory modeling by Jinyoung Kim et al. in subsequent work demonstrates that the height of the groundwater table is a primary accelerator of this phenomenon; elevated water levels drastically increase seepage stress, speeding up the collapse of sandy ground into these hidden voids.
Interestingly, the physical characteristics of the infrastructure itself dictate the severity of the hazard. Jin‐Young Kim et al. in their research established a positive correlation between the diameter of underground conduits and subsidence frequency, because larger failed pipes facilitate more extensive soil erosion and accommodate larger subsurface voids. Conversely, the geotechnical composition of the soil can act as a natural brake: if the soil's fine content (such as clay particles) exceeds 15%, it significantly stabilizes the matrix and slows down cavity propagation, as noted by Jinyoung Kim et al. in their 2017 paper. When these man-made infrastructure failures are left unmonitored, they become the primary predictors of sudden urban surface collapses, often superseding natural geotechnical properties in risk models, as shown by Jun Hwan Park et al. in their study.
Fluid Extraction, Effective Stress, and Consolidation
On a regional scale, the most widespread driver of urban ground deformation is the extraction of subsurface fluids, primarily groundwater. The underlying mechanism is governed by Terzaghi's principle of effective stress:
$$\sigma' = \sigma - u$$
where $\sigma'$ is the effective stress supported by the soil skeleton, $\sigma$ is the total overburden stress, and $u$ is the pore water pressure. When groundwater is heavily pumped for municipal expansion or industrial activities like coastal aquaculture, pore water pressure ($u$) drops, transferring the load directly to the soil grains ($\sigma'$ increases). This increase in effective stress causes the soil matrix to compact, leading to large-scale land subsidence, as documented by Ziwei Wang in their study and Wei Tang 0008 et al. in their paper.
The rate and severity of this compaction depend heavily on the local hydrogeological setting. Cities built on soft mucky soils, fluvial plains, or reclaimed coastal land experience accelerated consolidation under the combined weight of new infrastructure and fluid depletion, as shown by Yizhan Zhao et al. in their analysis, Chaolei Wang et al. in their findings, and Huanghao Yu et al. in their study.
Furthermore, as shown by Yueting Li et al. in their research, the geometry of the underlying bedrock—such as the steepness of buried ridges—can induce non-uniform subsidence patterns, which in turn trigger the formation of destructive ground fissures, a relationship explored by Yuting Wu et al. in their paper. Conversely, when cities implement large-scale water diversion projects to reduce groundwater reliance, the pore pressure can recover, leading to measurable ground rebound and a stabilization of the urban surface, as observed by Guan Chen et al. in their analysis.
Disentangling Multi-Causal Subsidence via Remote Sensing
Because urban ground deformation is a multi-variable problem where construction loads, tunneling, fluid dynamics, and natural geology overlap, isolating the exact cause of subsidence at any given point is a major challenge. Modern research addresses this by pairing high-resolution satellite remote sensing with advanced risk modeling.
Time-series Interferometric Synthetic Aperture Radar (InSAR), including Persistent Scatterer (PS-InSAR) and Small Baseline Subset (SBAS-InSAR) techniques, allows researchers to map millimeter-level surface displacements over entire metropolitan areas. By analyzing cross-heading orbits and integrating radar data with spatial datasets of building density and geological layers, frameworks proposed by Bahaa Mohamadi et al. in their paper can quantitatively decouple singular, dual, and complex subsidence causes. This reveals, for instance, whether a localized sinking funnel is driven primarily by historical land reclamation dewatering or the physical weight of high-rise structures.
[Satellite Radar (InSAR)] ---> Maps Millimeter-Level Surface Displacement
|
v
[Cross-Heading Orbit Analysis] + [Geological/Urban Spatial Data]
|
v
[Multi-Parameter Inversion Models] ---> Decouples Singular vs. Complex Causes
(e.g., Dewatering vs. Construction Load)
When combined with numerical simulations (such as FLAC3D) and real-time sensor networks, these integrated models capture the dynamic interactions between fluctuating water tables and subway operations. As demonstrated by Lv Zhou et al. in their risk modeling and the authors of this integrated monitoring study, these tools provide city planners with a predictive framework to mitigate structural degradation before it manifests as catastrophic failure.
Key Foundational Work
Applying Building Information Modelling in Environmental Impact Assessment for Urban Deep Excavation Projects (2012) — contributed 1 concept elements to this direction.