EVIDENCE. METHODS. ENGINEERING JUDGMENT.

Geotechnical Monitoring Research & Methods

GeoSmar documents practical methods for monitoring data QA/QC, baseline control, trend and rate analysis, trigger review, multi-sensor correlation, InSAR interpretation and engineer-reviewed reporting.

Research & Methods

Methods should be traceable from the field reading to the engineering conclusion.

This page sets out the working methods GeoSmar uses to think about geotechnical monitoring. The emphasis is practical: define the question, understand the ground and construction context, check the data, compare independent evidence, and state clearly what is observed, what is inferred and what remains uncertain.

Research is useful when it improves a real monitoring decision.

GeoSmar does not treat research as a separate academic layer detached from projects. Methods are developed around recurring engineering problems: unreliable baselines, conflicting instruments, changing groundwater, apparent trigger exceedances, wide-area ground motion, large monitoring datasets and the need to communicate findings consistently.

Data QA/QC Baseline control Trend & rate analysis Trigger review Anomaly diagnostics Multi-sensor correlation InSAR interpretation Engineer-reviewed reporting
This is a global methods page, not a site-specific design note. No geology, stratigraphy, groundwater condition, trigger value or instrument layout should be assumed for a real project until the relevant project records have been reviewed.

Research framework

Start with the engineering question, then choose the measurement.

A monitoring method should be selected because it can answer a defined question at the required scale, frequency and reliability. This approach is consistent with Federal Highway Administration guidance, which states that instrumentation should be installed where necessary to answer specific critical questions and that monitoring plans should address data collection, record keeping and timely communication of findings.

1. Define the question What behaviour must be understood, and what decision will the monitoring support?
2. Define the mechanism What ground, groundwater, structural or construction process could produce the observed response?
3. Select evidence Choose instruments and datasets that can test the mechanism at the right spatial and temporal scale.
4. Check reliability Confirm baseline, continuity, calibration or survey control, data completeness and known limitations.
5. Interpret & review Compare trends, rates and independent datasets before reporting significance and follow-up actions.

Project ground context

Monitoring interpretation depends on the ground model and construction sequence.

Before a project-specific method is proposed, the monitoring problem should be read against the available subsurface and design information. For a real engagement, GeoSmar would expect the relevant project-controlled records rather than infer site conditions from a generic regional description.

Ground model

Geology & stratigraphy

Review borehole logs, geological sections, CPT or other investigation records, fill history, soil or rock units, weathering profile and known lateral variability where these records exist.

Hydrogeology

Groundwater

Review measured groundwater levels, piezometric conditions, seasonal variation, pumping or dewatering history and any hydraulic boundaries relevant to the expected mechanism.

Construction context

Sequence & influence zone

Relate monitoring to excavation depth, tunnelling progress, surcharge, foundation works, dewatering, temporary works, adjacent assets and the timing of construction changes.

What project records are useful before interpreting monitoring data?
Typical inputs include the ground investigation report, borehole and CPT records, geological sections, groundwater observations, design drawings, construction sequence, method statements, instrumentation and monitoring plan, baseline report, trigger table, instrument installation records, survey-control information and previous monitoring reports.
Can a regional geological map replace the project ground model?
No. Regional mapping can provide context, but project interpretation should rely on the site investigation and project records available for the actual works. Local stratigraphy, groundwater and construction conditions can differ materially over short distances.

Method selection

Choose instruments by mechanism, not by habit.

Different instruments answer different questions. A sound monitoring design considers what is being measured, where the expected movement or pressure change may occur, the required accuracy and frequency, access constraints, redundancy and how the data will be interpreted.

Method Typical engineering question What should be checked before interpretation
Inclinometer / IPI Where is lateral ground or structural movement occurring with depth? Baseline, casing orientation, depth reference, profile continuity, installation history and whether movement is localised or distributed.
Piezometer / groundwater monitoring How are pore pressure or groundwater conditions changing? Instrument elevation, datum, response time, barometric or temperature effects where relevant, dewatering sequence and nearby groundwater observations.
Levelling / settlement point Is a surface, structure or foundation settling or heaving? Survey control, benchmark stability, closure, repeatability, baseline epoch and whether the measured point represents the feature of interest.
Total station / GNSS How are points moving in plan and elevation over time? Reference stability, coordinate system, atmospheric or sighting effects, prism condition, epoch consistency and network geometry.
Tilts, cracks & displacement gauges Is a structure rotating, opening, closing or moving locally? Temperature sensitivity, fixing stability, local versus global movement, zero setting and adjacent structural observations.
Vibration monitoring What vibration level is associated with construction or operational activity? Sensor fixing, sampling configuration, event timing, source activity, applicable project criteria and whether the measured location represents the receptor.
InSAR Where is wider-area line-of-sight ground or asset motion occurring over time? Coherence, viewing geometry, temporal coverage, reference area, atmospheric effects, land cover and comparison with ground observations where available.
Instrument lists are not designs. Final selection, location, spacing, frequency, accuracy and redundancy are project-specific and should follow the ground model, risk assessment, design requirements and monitoring objectives.

Data QA/QC

A clean chart does not make uncertain data reliable.

Data screening should happen before interpretation. The purpose is not to remove inconvenient readings, but to understand whether the dataset is complete, internally consistent and technically usable.

  • Confirm instrument ID, location, elevation and units
  • Check baseline date and accepted baseline values
  • Check missing, duplicate or out-of-sequence records
  • Review abrupt jumps and step changes
  • Check survey or reference-point stability
  • Compare nearby or related instruments where available
  • Record maintenance, recalibration or replacement events
  • Keep raw data separate from corrected or processed data
  • Preserve calculation and revision history
  • State known limitations with the interpretation
Should an apparent outlier simply be deleted?
Not without a documented technical reason. A suspicious reading may be an error, but it may also be the first indication of real movement. The appropriate approach is to flag it, compare related evidence, review instrument or survey records and retain traceability to the original value.
Why separate raw and processed data?
Corrections, filtering, coordinate transformations and derived calculations can change the appearance of a dataset. Keeping the source data intact allows later reviewers to reproduce the analysis and understand exactly how a reported result was produced.

Trend & anomaly analysis

Magnitude matters. Rate, direction and persistence often matter just as much.

GeoSmar treats monitoring interpretation as a time-dependent engineering problem. A reading is reviewed in relation to its baseline, recent rate of change, construction stage, nearby measurements and the expected mechanism.

Magnitude

How far has the value moved from the accepted baseline or another defined reference?

Rate

Is the movement stable, slowing, accelerating or changing after a construction event?

Spatial pattern

Does the change appear in one point, a group of instruments, a profile or a wider deformation zone?

Correlation

Does the timing align with excavation, tunnelling, groundwater change, loading, rainfall or another documented project event?

An anomaly is a prompt for review, not a diagnosis. The engineering mechanism still has to be tested against independent evidence and project context.

Triggers & response

Trigger values need a response framework, not just a coloured line.

Monitoring thresholds are project-specific. Their technical meaning depends on the design basis, asset sensitivity, construction stage, measurement uncertainty and the actions assigned to each level. A monitoring method should therefore consider both the trigger condition and what happens after it is reached.

Detection

What triggered?

Confirm the instrument, value, timestamp, baseline, threshold logic and whether the event is a single reading or a persistent change.

Verification

Is it credible?

Check sensor status, nearby data, survey control, construction activity and other evidence before assuming the alert represents real movement.

Response

Who decides what happens next?

The monitoring plan or contract should define notification routes, review responsibilities, escalation, verification and authority for any operational or construction action.

InSAR & remote sensing

Use satellite ground-motion data for the questions it can answer well.

InSAR compares radar observations acquired at different times to measure changes in the satellite line of sight. ESA states that radar interferometry can detect slight ground movement across wide areas, while the Copernicus European Ground Motion Service uses Sentinel-1 InSAR data to provide large-scale ground-motion information for infrastructure and natural-hazard applications.

Useful for

Screening, history and spatial context

InSAR can help identify wider deformation patterns, review historical movement where suitable data exist, compare corridors or large assets and support decisions about where additional ground investigation or instrumentation may be justified.

Needs care

Geometry, coherence and interpretation

InSAR is not a direct replacement for every ground instrument. Viewing geometry, land cover, coherence, atmospheric effects, reference selection, acquisition frequency and the direction of motion all affect interpretation.

Why compare InSAR with ground instrumentation?
The methods observe movement differently. InSAR can provide wide spatial coverage and historical context, while ground instruments can provide local measurements at specific depths, points or structural components. Agreement or disagreement between the datasets can itself be technically useful when the measurement geometry and limitations are understood.
Can InSAR alone explain the geotechnical mechanism?
No. A deformation pattern shows that movement may be occurring; it does not by itself establish the cause. Geology, groundwater, construction activity, loading, asset behaviour and other project information are needed before a geotechnical mechanism can be assigned.

Validation & reproducibility

A method should be reviewable by another engineer.

For recurring monitoring intelligence, reproducibility is as important as speed. A later reviewer should be able to identify the source dataset, repeat the calculation, understand any correction or exclusion and see which version of the analysis was used in the report.

  • Define the source-of-truth dataset
  • Keep raw data immutable where practicable
  • Record units, datums and coordinate systems
  • Version calculation logic and scripts
  • Document filters, corrections and exclusions
  • Record instrument replacement or baseline reset
  • Retain figure and report revision history
  • Separate automated flags from engineer conclusions
  • Use independent checks for critical calculations
  • State uncertainty and method limitations
The same principle applies to automated reporting: automation can organise, calculate and screen data, but technical conclusions should remain traceable to the project evidence and engineer review.

Contract & governance interfaces

Many monitoring disputes begin with an undefined interface rather than a bad instrument.

Technical methods work best when responsibilities are clear. At tender or mobilisation stage, the monitoring scope should define who owns the data, who verifies the instruments, who controls baselines and triggers, how late or missing data are handled and who has authority to issue engineering instructions.

Data ownership & access

Define the source dataset, file format, access route, reporting cut-off, retention period and whether raw data are available to independent reviewers.

Baseline & trigger control

Define who accepts the baseline, who may revise it, who owns trigger values and how changes are approved and recorded.

Instrument responsibility

Separate responsibility for installation, calibration, maintenance, field verification, data transmission and engineering interpretation.

Reporting & sign-off

Define reporting frequency, review period, recipients, engineer sign-off, revision control and the treatment of late or corrected data.

Alarm escalation

Define notification hierarchy, acknowledgement, verification, emergency contact routes and which party has authority to change construction or operations.

Change management

Record instrument relocation, replacement, altered construction sequence, design revisions and changes in monitoring frequency or scope.

GeoSmar can review monitoring governance and technical interfaces, but project-specific contractual interpretation should be checked against the actual contract documents and, where necessary, the client’s legal and contractual advisers.

Official guidance & case evidence

Public methods are most useful when their source is clear.

The examples below are official third-party references used to illustrate current monitoring practice. They are not GeoSmar projects and do not imply partnership, endorsement or project involvement.

FHWA — instrumentation linked to engineering questions

FHWA guidance states that instrumentation should answer specific critical project questions and that monitoring plans should address timely collection, standardised records and communication of findings with geotechnical interpretation.

Official FHWA source ↗

USACE — instrumentation, monitoring and performance assessment

USACE Engineer Manual EM 1110-2-1908 provides guidance for personnel responsible for instrumentation, monitoring and assessment of embankment dams and levees, reinforcing the link between measurement and performance evaluation.

Official USACE manuals ↗

ESA / Copernicus — wide-area ground-motion methods

ESA describes radar interferometry as a method for detecting slight ground movement across wide areas. Copernicus EGMS applies Sentinel-1 InSAR data for ground-motion information relevant to infrastructure, subsidence and natural-hazard assessment.

Official Copernicus source ↗

Senceive — São Paulo Metro Line 6

Senceive’s official case study reports that the monitoring programme incorporated 279 automated sensors and millions of automated readings alongside a much larger manual dataset, illustrating how higher-frequency monitoring changes the volume and resolution of evidence available to engineering teams.

Official Senceive case ↗

Worldsensing — Eppenberg Tunnel

Worldsensing’s official Eppenberg Tunnel case describes wireless integration with geotechnical instruments and automated supervision of sensor measurements, reducing the need for manual site readings while improving monitoring efficiency.

Official Worldsensing case ↗

Sixense — InSAR integrated with ground monitoring

Sixense describes Atlas InSAR as a ground and infrastructure displacement monitoring service that can be combined with ground instrumentation and adapted to construction, operation and maintenance stages with different reporting frequencies.

Official Sixense source ↗

GeoSmar research direction

Research should move toward repeatable engineering workflows.

GeoSmar’s research direction is centred on methods that can be reused across projects without removing engineering judgment. The aim is to make monitoring review more consistent, auditable and scalable while preserving a clear boundary between automated processing and professional interpretation.

Data quality diagnostics

Methods for identifying missing data, step changes, baseline problems, instrument inconsistency and suspicious measurements before interpretation.

Trend & anomaly assessment

Repeatable ways to review magnitude, rate, acceleration and spatial consistency across large monitoring datasets.

Ground + satellite correlation

Methods for comparing local ground instrumentation with wider-area InSAR-derived movement while respecting the geometry and limitations of both.

Alert intelligence

Structured review of threshold exceedances, data quality and engineering context before an automated alarm is treated as a technical conclusion.

Automated reporting

Automating repetitive charting, checking and document preparation while retaining engineer review for interpretation and conclusions.

Applied validation

Testing methods against documented project datasets, controlled technical exercises and openly attributed public cases where appropriate.

Potential collaboration

GeoSmar is open to technical discussions with asset owners, consultants, monitoring contractors, universities, data providers and technology companies where a defined engineering problem can be studied with clear data provenance, confidentiality boundaries and attribution.

Frequently asked questions

Questions about monitoring methods.

Does GeoSmar publish project-specific trigger values?
No generic trigger value should be presented as suitable for every project. Trigger criteria depend on design, asset sensitivity, monitoring uncertainty, construction stage, contractual requirements and the response actions agreed for that project.
Can the same monitoring method be used for every tunnel, slope or building?
No. The engineering mechanism, ground conditions, geometry, access, asset sensitivity and construction sequence determine what needs to be measured and how frequently. The same instrument can be useful in two projects for different reasons.
How does GeoSmar treat AI or automated analysis?
Automation can assist with data screening, plotting, calculations, comparison and draft reporting. GeoSmar does not treat an automated output as a final engineering conclusion without project context, stated limitations and professional review.
Can GeoSmar review a monitoring method developed by another contractor or consultant?
Yes, subject to scope and available information. An independent review can consider the monitoring objectives, ground and construction context, instrument selection, baseline, data flow, trigger framework, reporting process and whether the available evidence supports the stated interpretation.
Can public case studies be used as direct design precedents?
They can provide technical context, but they do not replace project-specific design. Published cases may differ in geology, structure, risk, monitoring frequency, contractual framework and available technology. Their source and limitations should remain clear.

Technical discussion

Have a monitoring method, dataset or anomaly you want to review?

Send the relevant monitoring plan, sample data, report or project brief. GeoSmar can help define the technical question, identify the project information needed for a defensible review and discuss whether the problem is best approached through monitoring intelligence, independent review, data diagnostics, InSAR interpretation or method development.

Scroll to Top