EVIDENCE. CONTEXT. ENGINEERING LESSONS.
Geotechnical Monitoring Case Studies
Verified case studies and public-source engineering analyses show how monitoring, survey and InSAR data are used to understand movement, validate evidence and support decisions across infrastructure and critical assets.
Case studies
Monitoring lessons are most useful when the evidence is traceable.
This page is a case-study library for geotechnical monitoring and ground-movement engineering. It brings together GeoSmar material when it is available and clearly labelled public-source analyses of major infrastructure and geohazard projects. The aim is not to borrow another organisation’s project history. It is to examine what the published monitoring evidence shows, what the local ground conditions meant for interpretation, and what a project team can learn from the monitoring strategy.
Only where GeoSmar was actually engaged
A project is labelled a GeoSmar Engagement only when the scope, role and publishable evidence can be verified. Client-confidential details are not inferred or reconstructed.
Official project information, independently discussed
Government, project-owner, research-agency and other official sources are used to discuss monitoring methods and engineering lessons. These are not presented as GeoSmar projects.
Methods without invented project claims
Where a workflow is demonstrated using sample or synthetic data, it is identified as a technical demonstration rather than a live project case.
Case library
Five verified examples, five different monitoring questions.
The cases below were selected because the official sources disclose enough engineering detail to support a useful discussion of ground conditions, monitoring design, data interpretation or spatial ground-motion assessment.
Crossrail — Finsbury Circus, London
Linked automated monitoring inside and outside buildings during SCL tunnelling and compensation grouting.
Crossrail C510 — Whitechapel
Long-term settlement after SCL excavation, including London Clay consolidation and monitoring uncertainty.
Central Artery/Tunnel — Boston
Pile-driving-induced heave in thick soft marine clay, monitored with deformation points, piezometers, a heave gauge and inclinometer.
Cleveland Corral — California
Long-duration landslide monitoring using pore-water pressure, surface displacement, soil moisture and precipitation data.
Thames Tideway Tunnel — London
Large-area InSAR monitoring used alongside conventional monitoring over a major urban tunnel corridor.
Crossrail · Finsbury Circus · London
When the whole building is inside the zone of influence, the reference system becomes part of the engineering problem.
Crossrail’s Learning Legacy describes linked monitoring used around Finsbury Circus during SCL tunnelling and compensation grouting for Liverpool Street Station. The case is valuable because the problem was not simply “measure settlement”. The buildings, access constraints, wide zone of influence and rapid ground response required several monitoring systems to be linked onto a common movement framework.
What made the monitoring difficult?
What monitoring approach was used?
What is the transferable engineering lesson?
What would GeoSmar review on a comparable project?
Crossrail C510 · Whitechapel · London
A trigger can close, but the ground may still be moving.
Crossrail’s published study of long-term settlement following SCL tunnel excavation at Whitechapel and Liverpool Street found that ground movement did not stabilise soon after the tunnel face left the zone of influence. The case is a strong reminder that construction-stage monitoring and long-term behaviour are not always the same problem.
London Clay consolidation
The official analysis attributes the long-term settlement primarily to consolidation and drainage effects around the permeable tunnel lining. At Kempton Court the platform tunnel was excavated entirely within London Clay, while the top of the Lambeth Group was reported about 6 m below the invert.
Short datasets can mislead
Crossrail compared predictions made from six months, twelve months and a longer dataset. The published paper concludes that long monitoring periods and frequent readings were important for reliable long-term settlement prediction.
Uncertainty accumulates
The paper identifies benchmark stability, nearby excavations, depressurisation, wider ground movement, local instability and seasonal effects as sources that can influence long-term rate calculations and trend interpretation.
Closing a construction activity does not automatically close the monitoring question. On projects where consolidation, groundwater change or time-dependent deformation is plausible, the monitoring termination criteria should be tied to observed behaviour and the engineering mechanism rather than a calendar date alone.
Central Artery/Tunnel · Boston · United States
Soft marine clay turned pile installation into a ground-movement problem.
The U.S. Federal Highway Administration documented pile-driving lessons from Boston’s Central Artery/Tunnel project. At one airport-area site, the published subsurface sequence included fill and organic silt and sand over a very thick soft marine clay deposit, underlain by glacial soils and bedrock. Pile driving produced heave that affected the ground and an adjacent building.
Instrumentation evolved with the problem
Deformation monitoring points were installed along the adjacent building. During the second phase, nested vibrating-wire piezometers were installed close to deformation points; a multipoint heave gauge and an inclinometer were also added to measure vertical and lateral ground response.
Mitigation had to be checked against data
FHWA records the use of wick drains and pile preaugering as mitigation measures. The monitoring evidence showed that heave continued, and the piezometer response indicated that the wick drains were not rapidly dissipating the excess pore pressure generated by pile driving.
Cleveland Corral landslide · California · United States
For a landslide, rainfall alone is not the movement record.
The U.S. Geological Survey has published near-real-time monitoring data for the episodically active Cleveland Corral landslide complex near U.S. Highway 50 in El Dorado County, California. The released dataset covers 1997–2018 and combines hydrologic and deformation observations.
Piezometers
Subsurface pore-water pressure was recorded at different depths and locations within the slide mass.
Extensometers
Downslope displacement of the ground surface was recorded at different locations.
Water-content sensors
Soil moisture was measured at the toe of the landslide.
Rain gauge
Precipitation, including rainfall and snowmelt, was recorded as an environmental driver.
Rainfall becomes more useful when reviewed alongside pore pressure, water content and observed displacement rather than interpreted as a stand-alone warning variable.
A long record helps distinguish seasonal response, episodic movement and persistent changes that may not be visible in a short construction-style monitoring window.
The cited USGS data-release page describes the monitoring system but is not used here as a substitute for a project geotechnical ground model. Any stratigraphic interpretation for a live project would require the relevant official investigation records.
Thames Tideway Tunnel · London · United Kingdom
Satellite monitoring can extend the field of view beyond individual instruments.
Sixense’s official project page describes Atlas InSAR monitoring for the Thames Tideway Tunnel East section. The published scope covered more than 5.5 km of alignment with a corridor extending 500 m either side of the tunnel. The project page states that 83 high-resolution TerraSAR-X images were used for the baseline study, providing up to 147,126 measurement points.
Wide spatial coverage
InSAR provided a dense ground-motion layer across a large urban corridor rather than only at instrumented points.
Historical baseline
The project used satellite imagery to establish a retrospective movement baseline before the main monitoring period.
Integration, not replacement
The official page describes delivery through a web platform that could integrate InSAR information with conventional monitoring methods.
Engineering lessons
Different projects, the same recurring monitoring questions.
The cases above involve different assets, ground conditions and monitoring technologies. Their transferable value is in the decisions around the data.
What is the movement mechanism?
Settlement, heave, consolidation, groundwater response and landslide displacement require different evidence. Instrument choice should follow the mechanism being tested.
Is the reference stable?
Reference points, benchmarks and baseline periods are part of the measurement system. Their stability has to be reviewed before small changes are treated as real movement.
Does the monitoring period match the behaviour?
Construction activity can finish before consolidation or hydrologic response stabilises. Monitoring duration should reflect the expected engineering process.
Are datasets being cross-checked?
Piezometers, inclinometers, optical survey, levelling, environmental observations and InSAR each measure a different part of the problem. Agreement and disagreement are both useful.
Can mitigation be verified?
If grouting, drainage, excavation sequencing or another measure is intended to alter behaviour, the monitoring system should be able to test whether that change occurred.
Who decides what an exceedance means?
The contract should separate data acquisition, data validation, engineering review, alert communication and authority to change or stop the works.
Evidence & attribution
A case study should make ownership of the evidence obvious.
GeoSmar uses a simple publishing rule so that project history, public research and technical demonstrations are not mixed together.
Where GeoSmar can contribute
The useful role often begins after the instruments have already produced data.
GeoSmar is positioned as a monitoring-intelligence and independent-review layer. On a project comparable to the cases above, the first task would be to define the engineering question and identify the evidence already available before proposing another sensor or platform.
Review the monitoring architecture
Check instrument purpose, reference system, baseline, spatial coverage, reading frequency, trigger logic, reporting chain and known blind spots.
Investigate contradictory or abnormal data
Compare neighbouring instruments, rates, groundwater response, survey controls, construction sequence and potential measurement artefacts.
Add wider ground-motion context
Where satellite geometry, coherence and project conditions are suitable, review satellite-derived deformation alongside ground instrumentation and survey data.
Define what must be measured
Develop the monitoring philosophy around the expected mechanism, asset sensitivity, construction stage, hydrogeology and decision framework.
Connect thresholds to response
Separate automated exceedance detection from data validation, engineering review, escalation and decision authority.
Keep the engineering context current
Use periodic or event-driven review where movement, groundwater, construction sequence or asset condition changes through time.
Who owns the raw data? Who establishes and approves the baseline? Who validates an exceedance? Who maintains the instruments and survey control? What response time is required? Which party has authority to change the works? How are missing data handled? When can monitoring be reduced or closed? These questions should be explicit before a monitoring dataset is used as a contractual decision tool.
Official source register
Primary and official project sources used on this page.
The case summaries above are deliberately limited to information that can be traced to the organisations below. No third-party blog or unattributed project summary is used as the basis for a project fact.
Crossrail, FHWA, USGS, Sixense and Copernicus are cited as source organisations. Their inclusion does not imply partnership with, endorsement of, or project participation by GeoSmar.
Frequently asked questions
How GeoSmar uses case evidence.
Are all projects on this page GeoSmar projects?
Why include geology and stratigraphy in a monitoring case study?
Can a monitoring case study be useful without publishing client data?
Can GeoSmar review a monitoring problem from an existing project?
What should be sent for an initial discussion?
Discuss a case
Have a monitoring result that does not fit the expected behaviour?
Send GeoSmar a project brief, sample monitoring report or selected dataset. The first discussion can focus on the evidence already available: what changed, when it changed, whether the change is credible, what the ground or construction context suggests, and which additional checks would add the most value.
GeoSmar is the market-facing brand of Rauz Caucasus LLC. Project scope, response obligations, professional responsibility and any requirement for local statutory involvement are defined on a project-specific basis.