MEASURE. CHECK. INTERPRET. ACT.

Geotechnical Monitoring Knowledge Base

Practical guidance on geotechnical monitoring, from instrument selection and baseline planning to data QA/QC, trigger frameworks, trend interpretation, InSAR and engineer-reviewed decisions.

Monitoring Knowledge Base

Start with the engineering question, not the instrument.

A monitoring system earns its value when it answers a defined question about ground, groundwater, structure or asset behaviour. This knowledge base is organised around that principle: define what needs to be understood, select measurements that can resolve it, establish a usable reference condition, check the data, and interpret change in the context of the works.

QuestionWhat behaviour or failure mechanism matters?
MeasureWhich observation can detect it reliably?
ValidateIs the reading complete, traceable and credible?
InterpretDoes the trend fit the engineering context?
ActWhat review, verification or response follows?
FHWA guidance states that instrumentation should be installed and monitored where it is needed to answer specific critical questions, with timely monitoring, standardised records and geotechnical interpretation of observations. That is a useful starting discipline for any monitoring programme.

Fundamentals

What a monitoring plan should establish before readings begin.

Instrumentation is only one part of the plan. A technically useful monitoring programme also needs an engineering purpose, a baseline, defined responsibilities, an agreed reading frequency, a clear route for abnormal data, and a record of how changes will be reviewed.

01

Purpose

State the question being monitored: deformation, pore pressure, settlement, structural rotation, vibration, convergence, crack movement or another defined response.

02

Expected mechanism

Relate the measurement to the ground model, asset geometry, construction sequence and plausible mechanism. A reading without mechanism is difficult to interpret.

03

Measurement strategy

Define instrument type, location, reference, range, resolution, frequency, redundancy and how manual or independent checks will be made where warranted.

04

Baseline

Collect enough pre-works information to understand normal variation, instrument stability and any thermal, seasonal or operational effects relevant to the asset.

05

Response framework

Set the review and escalation process before a threshold is reached: who checks the data, who verifies it, who is informed and who has authority to change the works.

06

Close-out

Define what evidence is needed before monitoring frequency can be reduced or instruments can be decommissioned. Stable post-work behaviour should be demonstrated, not assumed.

Why is there no universal monitoring layout?
Because geology, groundwater, asset sensitivity, construction method, access, expected deformation mechanism and contractual requirements differ from project to project. This page therefore explains principles and instrument roles; it does not prescribe a universal spacing, frequency or trigger value.
Where should the monitoring scope be recorded?
Project documentation should define performance requirements, layouts, power and communications, baseline period, measurement frequency, reporting and the data exchange route between parties. Crossrail’s published developer guidance is one official example of this type of structured I&M scope.

Instrument Knowledge

Choose the observation that matches the behaviour you need to understand.

Different methods observe different components of movement and operate at different spatial and temporal scales. Selection should begin with the engineering question, then consider access, range, resolution, reference stability, automation needs and independent verification.

Engineering question Common observation methods What to check before use
Lateral ground deformation Manual or in-place inclinometers; geodetic survey where surface or asset movement is relevant Reference depth, casing condition, installation geometry, baseline, repeatability and the expected shear/deformation zone
Settlement or heave Precise levelling points, prisms/automated total stations, hydrostatic levelling systems, settlement markers or other project-specific methods Stable datum, line-of-sight, temperature effects, benchmark integrity, expected magnitude and required frequency
Groundwater or pore pressure Standpipe or vibrating-wire piezometers and other project-specific groundwater instrumentation Response zone, installation seal, datum, barometric/temperature considerations, reading frequency and relation to the hydrogeological model
Structural rotation or convergence Tilt sensors, electro-level systems, prisms, convergence systems and precise survey Sensor axis, fixing method, thermal influence, reference stability and whether displacement or rotation is the governing quantity
Crack or joint movement Crack meters, tell-tales, displacement gauges or optical survey where appropriate Gauge orientation, initial condition, environmental effects, expected direction and whether local change represents wider asset behaviour
Wide-area ground motion Satellite InSAR, combined where necessary with ground instrumentation and survey Line-of-sight geometry, coherence, reference frame, temporal sampling, spatial coverage and the need for ground truth or local interpretation
Crossrail’s published monitoring records show the practical use of rod extensometers, in-place inclinometers, vibrating-wire piezometers, automated total stations and prisms, precise levelling, tilt sensors, crack meters and other methods across different assets and tunnelling interfaces.

Baseline & Reference

A good baseline separates project effects from normal behaviour.

The first reading is not automatically a useful baseline. A reference period should be long enough, and stable enough, to show instrument behaviour and the environmental or operational variation that may otherwise be mistaken for project-induced movement.

What baseline monitoring should reveal

  • Normal short-term scatter and repeatability.
  • Seasonal, thermal, groundwater or operational cycles where relevant.
  • Reference-point or benchmark stability.
  • Instrument drift, installation behaviour or early faults.
  • Pre-existing asset movement before the works begin.

What must remain traceable

  • Baseline start/end dates and exclusions.
  • Datum, coordinate system and reference network.
  • Calibration and installation records.
  • Any re-zeroing, replacement or processing changes.
  • Construction events occurring during the baseline period.
Official case lesson — Crossrail Hyde Park / Bayswater Road

Crossrail’s published field-instrumentation research reports more than 12 months of baseline monitoring before tunnelling, used to confirm instrument accuracy and to assess thermal and seasonal effects. The lesson is not that every project needs 12 months; it is that baseline duration should be justified by the behaviour that must be separated from construction effects.

Data QA/QC

Before asking what the movement means, check whether the measurement is credible.

Monitoring anomalies can be real, instrument-related, reference-related or processing-related. QA/QC should preserve the raw evidence, document each transformation, and make it possible to distinguish an engineering change from a data-handling problem.

  • Confirm instrument identity, location and measurement axis.
  • Check timestamps, time zone and sampling interval.
  • Preserve raw values before filtering or correction.
  • Review gaps, duplicate records and sudden resets.
  • Check baseline, datum and reference-point changes.
  • Confirm units, sign convention and coordinate system.
  • Compare neighbouring or independent measurements where available.
  • Record calibration, maintenance, replacement and re-zero events.
  • Look for environmental or operational correlations.
  • Keep an auditable record of accepted and rejected data.
Automated systems

Validation should happen before alarm logic.

Leica GeoMoS describes outlier detection, data validation, filtering and automatic remeasurement as part of an automated monitoring workflow. The practical implication is simple: a threshold check is stronger when data quality is checked first.

Multi-source monitoring

Cross-checking improves context.

Trimble’s official monitoring material describes bringing together total-station, GNSS and geotechnical-sensor measurements to track and report movement. Integration does not remove the need to understand each sensor’s reference, precision and limitations.

Triggers & Action Plans

A trigger value is incomplete without an agreed response.

Trigger levels should be tied to the design basis, asset sensitivity, expected behaviour and a defined action plan. The number alone does not tell a project team whether the reading is valid, whether movement is accelerating, or what must happen next.

Define the basis

Record what the trigger is derived from: predicted response, serviceability criteria, asset requirements, observational method or another documented project basis.

Specify the measured quantity

Displacement, rate of change, pore pressure, rotation, strain, convergence or another quantity must be defined with units, sign convention and reference.

Define verification

State how an exceedance is checked: repeat reading, independent survey, adjacent instruments, site inspection, manual measurement or engineering review.

Assign authority

Name who receives the alert, who interprets it, who can increase monitoring frequency and who has authority to alter or stop the relevant activity.

Record the outcome

Every significant trigger event should leave a traceable record of the data, checks, engineering interpretation, decision and close-out.

No universal trigger values are provided here. Crossrail’s published Stepney Green case used project-specific Green, Amber and Red trigger levels derived from its design calculations and linked them to defined reviews and actions. That is a useful example of structure, not a transferable set of numbers.

Engineering Interpretation

A trend becomes useful when it is connected to mechanism and sequence.

Monitoring interpretation is not a search for the largest number on a dashboard. It is a structured comparison between the observed change, the expected behaviour, the timing of the works, groundwater conditions, adjacent instruments and the physical mechanism that could produce the pattern.

Magnitude

How large is the change relative to baseline variability, measurement uncertainty, expected response and the applicable project criteria?

Rate

Is movement stable, increasing, decelerating, cyclic or step-like? Rate of change can matter as much as cumulative movement.

Spatial pattern

Do neighbouring instruments, survey points or remote-sensing observations show a coherent deformation pattern?

Construction sequence

Does the timing align with excavation, tunnelling, dewatering, loading, grouting, mining, earthworks or another recorded activity?

Groundwater

Could pore-pressure or water-level changes explain settlement, heave, slope response or effective-stress changes observed elsewhere?

Instrument behaviour

Could a jump arise from reference movement, re-zeroing, cable or logger issues, casing effects, temperature, access disturbance or processing changes?

Useful question:

If one inclinometer, prism or piezometer changes suddenly while nearby measurements do not, do not start by deciding whether the project is safe or unsafe. Start by establishing whether the observation is repeatable, spatially coherent, temporally plausible and consistent with the engineering mechanism.

InSAR & Remote Sensing

Wide-area ground-motion data adds context that point instruments cannot provide alone.

Satellite InSAR can reveal deformation patterns across corridors, urban areas, slopes, mines and other large assets. It observes movement differently from ground instruments, so interpretation should account for line-of-sight geometry, spatial coverage, temporal sampling, coherence and the reference frame used by the product.

Screening

Where is movement occurring?

Use wide-area velocity and time-series information to identify zones that merit closer engineering review or targeted field investigation.

History

Was the movement already present?

Historical satellite time series can help separate long-term ground behaviour from movement that starts or changes during a project period.

Correlation

Does it agree with ground monitoring?

Where spatial and temporal scales allow, compare InSAR with survey, GNSS, settlement or other ground observations rather than treating either source in isolation.

Copernicus’ European Ground Motion Service uses Sentinel-1 InSAR and publishes ground-motion products for applications including dams, bridges, railways, buildings, landslides, subsidence and mining. Its official documentation also distinguishes Basic, Calibrated and Ortho products and explains that some products require more expert interpretation than others.

Data Integration

Integration should preserve engineering meaning, not just move numbers between systems.

A useful monitoring data model keeps source, timestamp, location, instrument identity, observed property, units, baseline, quality status and processing history traceable. That matters when project teams compare different suppliers, manual readings, automated loggers, survey data and satellite-derived observations.

Minimum data context

Instrument ID Location Timestamp Units Observed property Baseline Quality flag Source system Processing history

Open-interface context

The OGC SensorThings API provides a standard way to manage and retrieve observations and metadata from heterogeneous sensor systems. It is not a geotechnical monitoring specification, but its separation of Things, Locations, Sensors, Datastreams, ObservedProperties and Observations is relevant when building traceable multi-source monitoring architectures.

Official Case-Based Lessons

Public engineering cases show why monitoring needs context and verification.

The examples below come from official public sources. They are included for technical learning and are not GeoSmar projects.

Crossrail · London

Long baseline before tunnelling

At Hyde Park and Bayswater Road, a comprehensive scheme used rod extensometers, in-place inclinometers and multi-level vibrating-wire piezometers. More than 12 months of baseline data were used to understand accuracy, thermal effects and seasonal behaviour before tunnel construction.

Crossrail · Lindsey Street Bridge

Automated data plus manual verification

The published case describes automated total stations and 3D prisms providing hourly readings, complemented by manually monitored BRE sockets. The lesson is the value of an independent method when movement is important to construction control.

FHWA · Highway Geotechnics

Monitoring tied to critical questions

FHWA guidance identifies standpipe piezometers, slope inclinometers and surface monuments as standard methods in relevant geotechnical investigations and calls for timely monitoring, standardised records and interpretation by the geotechnical team.

Public cases are useful for learning the reasoning behind a monitoring scheme. They should not be copied as a design template without checking the new project’s ground conditions, geometry, asset sensitivity, construction method, contractual responsibilities and monitoring objectives.

Knowledge Topics

A practical reading map for geotechnical monitoring teams.

This hub is designed to expand into focused technical notes and guides. The priority topics below reflect the questions that repeatedly arise when project teams review monitoring data.

Inclinometer data

Baseline profiles, cumulative and incremental displacement, reference stability, sudden shifts, casing behaviour and cross-checks with nearby observations.

Piezometer data

Groundwater versus pore pressure, response zones, barometric/environmental effects, dewatering response and correlation with ground movement.

Settlement monitoring

Datum control, cumulative settlement, differential movement, rate of change, construction sequence and comparison between manual and automated observations.

Monitoring baselines

How long to monitor, what variability to expect, how to record reference changes and how baseline quality affects later trigger review.

Trigger frameworks

How thresholds connect to predictions, serviceability criteria, alert validation, escalation routes and engineering authority.

Data anomalies

How to investigate step changes, isolated exceedances, drift, missing data, conflicting instruments and apparent movement that may not be physical.

Automated monitoring

Sampling frequency, communications, system availability, data validation, alarm logic, manual fallback and responsibility for unattended periods.

InSAR interpretation

Line-of-sight movement, ascending/descending geometry, coherence, velocity maps, time series, wide-area screening and ground correlation.

Independent monitoring review

How to review a monitoring plan, contractor report, trigger event, data quality issue or unexplained trend without duplicating the field contractor’s role.

GeoSmar Approach

The knowledge base supports a practical engineering review layer.

GeoSmar is positioned between measurement and decision. We can work with data generated by a client’s existing monitoring contractor, survey team, logger, platform or satellite-data source, then focus on data quality, trend interpretation, trigger review and the questions that need further technical attention.

Independent review

Review monitoring plans, baseline logic, data quality, trigger frameworks, anomalies and the evidence supporting a contractor or consultant’s interpretation.

Monitoring diagnostics

Investigate unexplained readings, conflicting datasets, sudden changes, unusual rates or apparent exceedances before conclusions are drawn.

Monitoring intelligence

Provide recurring engineering review of monitoring data, focusing on what changed, whether it is credible, why it may matter and what should be checked next.

FAQs

Common questions about geotechnical monitoring interpretation.

How long should baseline monitoring last?
There is no universal duration. The period should be long enough to establish normal variability, reference stability and any relevant seasonal, thermal, groundwater or operational effects. The required duration also depends on the project programme and the behaviour being monitored.
Should a trigger exceedance automatically stop the works?
Not necessarily. The project action plan should define what each level means, how the reading is verified, who reviews it and what actions follow. Some trigger levels prompt increased frequency or engineering review; others may require operational restrictions or work changes. The response must come from the project’s agreed framework.
Can one instrument confirm a geotechnical problem?
Sometimes a single observation is important, but stronger interpretation usually comes from repeatability, nearby measurements, independent methods and consistency with the expected mechanism. An isolated change should be checked before it is treated as a complete engineering conclusion.
What is the difference between data QA/QC and engineering interpretation?
QA/QC asks whether the data are complete, traceable and credible. Engineering interpretation asks what a credible change means in the context of geology, groundwater, asset response and construction sequence. Both are required; good data can still be misinterpreted, and poor data can produce convincing-looking charts.
Can InSAR replace ground instrumentation?
Not as a general rule. InSAR and ground instruments measure movement in different ways and at different scales. InSAR can add wide-area historical and ongoing context, while ground instruments may provide direct subsurface, pore-pressure, structural or high-frequency measurements that satellite observations do not provide.
Can GeoSmar review monitoring data from another supplier?
Potentially, yes. A useful review depends on access to sufficient underlying data, metadata, baseline information, calibration or installation records where relevant, project drawings, trigger criteria and construction context. GeoSmar’s intended role is vendor-neutral interpretation rather than replacement of every existing field system.

Technical Discussion

Have a monitoring result that does not make sense?

Send a sample dataset, monitoring report, trigger table or short project brief. GeoSmar can help define whether the issue is best approached through data QA/QC, independent review, monitoring diagnostics, InSAR interpretation or a broader monitoring-intelligence workflow.

Official References

Sources used for this knowledge-base page.

External technical statements on this page are based on the official public sources below. References to other organisations describe the wider monitoring field and do not imply partnership, endorsement or a GeoSmar project relationship.

Scroll to Top