RAINFALL · NATURAL TERRAIN · LANDSLIDES

Extreme Rainfall & Landslide Monitoring in Hong Kong

Understand how rainfall, groundwater response, slope movement, debris-flow pathways and downstream exposure shape natural-terrain monitoring and engineering review.

HONG KONG / NATURAL TERRAIN

Rainfall Does Not Tell the Whole Landslide Story

In Hong Kong, intense or prolonged rainfall can trigger natural hillside failures, but individual outcomes also depend on geology, slope geometry, drainage pathways, groundwater response and exposed facilities. A rain gauge or inclinometer alone cannot establish where a debris flow will start or how far it may travel.

CEDD Report No. 1/2026 describes approximately 3,300 identified natural hillside catchments with a history of failure and proximity to buildings or important transport corridors. This is a prioritisation context, not evidence that every catchment is unstable or in immediate danger. See the official CEDD report.

Source zones Rainfall intensity Hydrological response Runout pathways Road and building exposure Engineering review

NATURAL-TERRAIN HAZARDS

Study Source, Pathway and Exposure as One System

An effective natural-terrain risk assessment does not stop at detecting movement at a selected sensor. It considers how a failure could develop, move and affect people or infrastructure.

SOURCE

Potential initiation area

Geology, weathering profile, geomorphology, groundwater and historic failures.

TRIGGER

Rainfall and subsurface water

Rainfall intensity, duration, antecedent conditions and pressure response.

PATHWAY

Travel and channel behaviour

Runout, channelisation, entrainment, obstructions and drainage interaction.

EXPOSURE

Downstream receptors

Buildings, highways, railways, utilities, access routes and potential consequences.

Sensor selection follows the hazard mechanism. Local instrumentation alone cannot replace catchment-scale geological and topographic assessment.

OFFICIAL HONG KONG CONTEXT

What Previous Rainstorms Demonstrate

The Hong Kong Observatory documented a record hourly rainfall of 158.1 mm at its headquarters during 7 September 2023. The event caused extensive flooding and landslides. GEO also publishes detailed engineering studies showing why individual slope mechanisms must be investigated separately.

HKO

September 2023 extreme rainfall

The HKO recorded record-breaking short-duration rainfall at its headquarters, illustrating how quickly regional hazard conditions can develop.

GEO / YIU HING ROAD

Detailed landslide study

GEO Report No. 377 documents a major 8 September 2023 incident where different failure components and road consequences required engineering investigation.

GEO / CATCHMENTS

Risk-based natural-terrain work

CEDD Report No. 1/2026 describes prioritisation and mitigation for vulnerable hillside catchments.

These are independent public references. They are not GEOOE projects, pilot installations or evidence of GEOOE involvement.

MONITORING STRATEGY

Match the Observation to the Failure Mechanism

Monitoring can supply useful supporting evidence, but it is only one part of a hazard-management system.

ObservationPotential valueKey limitation
Rainfall recordsTrigger context and spatial intensityDo not predict a precise individual failure location
Groundwater / piezometric pressureHydrological response beneath monitored groundResponse varies with instrument depth and hydrogeology
Inclinometer / extensometerLocal deformation or discontinuity movementMay miss failures outside its installed location
GNSS / survey / remote sensingSelected surface movement and wider spatial observationsCoverage, precision, revisit interval and line-of-sight differ
Terrain mapping and channel inspectionSource, pathway and receptor understandingNeeds specialist geological interpretation and periodic updating

WARNING & RESPONSE

Separate Long-Term Observations from Immediate Warnings

Hong Kong already operates a government Landslip Warning System based on real-time rainfall information, forecast and a rainfall–landslide relationship. Project-specific monitoring can complement this broader system; it should not imply that one private sensor can guarantee advance warning.

  • Define who monitors official weather and landslip advisories
  • Identify which assets need dedicated continuous observations
  • Specify response time, communication redundancy and responsible authority
  • Establish inspection and data-validation steps before an event
  • Document site access restrictions during severe weather
  • Agree actions, communication routes and review responsibilities in advance
Intermittent DAX-style retrieval is not a substitute for real-time warning or emergency communications where continuous monitoring and immediate action are required.

GEOOE / ENGINEERING TECHNOLOGY

Research the Monitoring Gaps that Matter

GEOOE’s technology directions may be relevant to selected slope and natural-terrain problems, particularly where data access is difficult and measurements need to be connected to engineering interpretation.

SENSING

Infrastructure Sensing

Define instruments and observation locations around a specific geological mechanism.

ACCESS

Distributed Access research

Study periodic retrieval for suitable non-critical or lower-frequency distributed assets.

INTELLIGENCE

Engineering data integration

Place rainfall, groundwater, movement, inspection and terrain observations into reviewable context.

For a closer technical discussion, see Slope & Landslide Monitoring and Distributed Access & Field Connectivity.

TECHNICAL LIMITS

Where Instrumentation Is Not Enough

The physical and operational limits of measurement should be explicit in any project proposal.

SPATIAL

Local samples are not complete coverage

A sensor cannot detect failures everywhere in a complex catchment.

TEMPORAL

Revisit intervals matter

InSAR or periodic visits may miss rapidly developing failure processes.

HYDROLOGY

Water response is not uniform

Pore pressure, perched groundwater and surface flow can vary locally.

MANAGEMENT

Alerts need people and procedures

A warning has value only when the project has communications, access controls and response plans.

PUBLIC AUTHORITIES

Official Rainfall and Landslide References

These primary sources establish the stated Hong Kong facts and provide technical background.

FREQUENTLY ASKED QUESTIONS

Rainfall and Landslide Monitoring — FAQs

Site-specific evaluation remains fundamental.

Can one inclinometer predict a debris flow?
No. It measures movement at an installed location. Natural-terrain hazard also depends on possible initiation areas, material, flow routes and exposed assets.
What does the figure of about 3,300 catchments mean?
CEDD has identified approximately 3,300 vulnerable natural hillside catchments with relevant failure history and proximity to buildings or key transport corridors. The figure does not imply all are unstable or in immediate danger.
Is rainfall monitoring enough?
Rainfall is important for broad landslip risk information but does not by itself locate every prospective failure or determine the path and consequences of debris movement.
Can DAX serve as an early-warning system?
Not by itself where immediate warning is required. Selective or periodic data collection must be separated from safety-critical real-time telemetry and emergency response.
What information supports a technical discussion?
Provide terrain mapping, hazard history, potential receptors, existing sensors, rainfall data sources, field access limitations and required response times.

CONTACT GEOOE

Discuss a Specific Natural-Terrain Monitoring Gap

We welcome conversations about slope instrumentation, field data access and ways to make rainfall and ground-response evidence more useful to engineering teams.

Email [email protected] with your site context, hazard mechanism and monitoring objective.

Scroll to Top