Geological Hazards

Learning Objectives

  • Trace a hazard from its source through a pathway to exposed people and infrastructure.
  • Distinguish total subsidence, differential movement, sinkholes and liquefaction mechanisms.
  • Compare swelling, wetting collapse and dispersion without prescribing a foundation from a soil name.
  • Interpret volcanic and coastal hazards using drainage, terrain, water-level and agency information.
  • Calculate effective stress, movement rates, simple loads and conditional monetary loss with explicit limits.
  • Select investigation, mitigation and monitoring questions that address uncertainty and residual risk.

Geological hazard

A potentially damaging Earth process or ground condition, including slow movement as well as sudden events.

Hazard, exposure and vulnerability

Hazard describes the potentially damaging process and its likelihood or intensity; exposure identifies the people or assets in its reach; vulnerability describes their susceptibility to harm.

Follow the source, pathway and receptor

A landslide source may lie outside a road footprint while its runout reaches the road. A river blockage may create upstream inundation and a possible downstream outburst. Exposure changes when access roads, utilities or evacuation routes cross these paths. Map uncertainty and secondary hazards belong in the ground model.

Landslides include falls, slides and flows. Rockfall involves detachment and falling, bouncing or rolling; slides move along shear surfaces; earthflows involve deforming, commonly fine-grained material. Rainfall can increase pore pressure, reduce suction and add weight; excavation, erosion and earthquakes can also change stability. A landslide dam requires assessment and monitoring: breaching is possible, not inevitable.

A slide source feeds runout across a road and into a river, creating potential upstream ponding and downstream release.

Original schematic section based on the USGS Landslide Handbook. Dashed release path denotes a possible breach, not an inevitable event or a predicted inundation boundary. No real site or travel time is represented.

A slide source feeds runout across a road and into a river, creating potential upstream ponding and downstream release.

Subsidence and differential movement

Subsidence is downward ground movement. Differential movement is the difference in displacement between locations; uniform subsidence can produce no relative vertical movement between two supports.

What a benchmark can establish

Groundwater withdrawal may lower pore pressure and increase effective stress, compacting compressible aquitards. Compaction may be elastic or inelastic depending on stress history; recovery of head does not guarantee recovery of lost elevation. Mining, dissolution, wetting collapse and organic-soil drainage are other causes.

Measure against a stable datum outside the moving system and consider survey uncertainty. One benchmark provides a local total displacement, not a building's angular distortion or a sewer's changed gradient. Spatial surveys and ground investigations are needed to infer those effects. Constant historical pumping does not guarantee constant future subsidence.

Downward-positive average subsidence

Use a consistent elevation datum and time interval; extrapolation is a conditional scenario, not a forecast.

vˉs=z1−z2t2−t1,z(t2+T)=z2−vˉsT\bar{v}_{\mathrm{s}}=\frac{z_1-z_2}{t_2-t_1},\qquad z(t_2+T)=z_2-\bar{v}_{\mathrm{s}}T

Variables

SymbolDescriptionUnit
z1,z2z_1,z_2Measured elevations at initial and final epochs, m.-
vˉs\bar{v}_{\mathrm{s}}Average downward movement rate, m/year; positive for subsidence.-
TTFuture scenario interval, years; constant rate is an assumption.-

Differential settlement and pipe gradient

For a straight reach with fixed horizontal separation, downstream settlement increases the upstream-to-downstream fall; upstream settlement reduces it.

δ=∣sU−sD∣,g=zU,0−zD,0−sU+sDL\delta=|s_{\mathrm{U}}-s_{\mathrm{D}}|,\qquad g=\frac{z_{\mathrm{U},0}-z_{\mathrm{D},0}-s_{\mathrm{U}}+s_{\mathrm{D}}}{L}

Variables

SymbolDescriptionUnit
sU,sDs_{\mathrm{U}},s_{\mathrm{D}}Downward-positive displacements at upstream and downstream ends, m.-
zU,0,zD,0z_{\mathrm{U},0},z_{\mathrm{D},0}Initial end elevations in the same datum, m.-
L,gL,gHorizontal reach length, m, and dimensionless positive fall in the intended flow direction.-
Two displacement graphs compare equal downward motion of both points with unequal motion at the same times.

Original synthetic monitoring graphs using the USGS subsidence framework. Both panels use identical time and downward-displacement scales; they are graphs, not physically exaggerated building sections. A stable external datum is assumed.

Two displacement graphs compare equal downward motion of both points with unequal motion at the same times.

Karst and sinkholes

Karst is terrain affected by dissolution of soluble rock. A sinkhole is a closed depression caused by dissolution or movement into subsurface openings.

Cover material changes the surface expression

Sandy cover may gradually migrate into openings, forming cover-subsidence sinkholes. Cohesive cover may bridge a void before abrupt cover collapse. Dissolution sinkholes can also develop where soluble rock is exposed or thinly covered. Water-level changes, concentrated infiltration and loading can affect development. A smooth surface or one clear borehole does not exclude nearby cavities.

Use geological mapping, drainage history and appropriate direct and indirect investigation together. Indirect anomalies do not uniquely establish a cavity. Cordon a newly opening depression and follow the site's incident procedures; inspection from an unsupported edge can expose personnel to further collapse.

Sandy sediment migrates progressively into rock openings while cohesive cover bridges a void and can collapse abruptly.

Original conceptual sections following USGS sinkhole descriptions. Bedrock openings are schematic; cover type suggests a mechanism but does not predict the time, size or location of collapse.

Sandy sediment migrates progressively into rock openings while cohesive cover bridges a void and can collapse abruptly.

Liquefaction

Loss of strength or stiffness in susceptible saturated soil during rapid loading, associated with increased pore water pressure and reduced effective stress.

Effective stress explains a mechanism, not a triggering forecast

In susceptible contractive soil, rapid cyclic loading can raise pore pressure because drainage cannot dissipate it fast enough. The grain skeleton then carries less effective stress. Consequences may include lateral spreading, settlement, sand ejection and loss of foundation support. Actual response depends on stratigraphy, density, fines, drainage, stress state and loading.

Zero effective vertical stress is an idealized endpoint. It does not mean every soil has zero residual shear strength, nor does it calculate permanent displacement. Pressure dissipation does not undo a damaged foundation. Triggering and deformation assessments require appropriate field/laboratory data and established methods.

Excess pressure and effective vertical stress

Positive compression in saturated soil. The ratio relation assumes constant total vertical stress and positive initial effective stress.

σv′=σv−u,ru=Δuσv,0′,σv′=σv,0′(1−ru)\sigma'_{\mathrm{v}}=\sigma_{\mathrm{v}}-u,\qquad r_u=\frac{\Delta u}{\sigma'_{\mathrm{v},0}},\qquad \sigma'_{\mathrm{v}}=\sigma'_{\mathrm{v},0}(1-r_u)

Variables

SymbolDescriptionUnit
σv,σv′\sigma_{\mathrm{v}},\sigma'_{\mathrm{v}}Total and effective vertical stress at the same point, kPa.-
u,Δuu,\Delta uTotal gauge pore water pressure and its increase above the initial pressure, kPa.-
rur_uDimensionless excess-pressure ratio; the teaching model explores zero through one.-

Explore a prescribed pressure state

Change the excess-pressure ratio while keeping the initial state fixed. Compare the pressure increase with the effective-stress decrease, then test both endpoints. The simulator supplies no earthquake duration, magnitude or site safety classification.

Excess pore pressure and effective stress

Concept and model scope

Change the prescribed excess-pressure ratio to see how pore water pressure and effective vertical stress share a fixed total vertical stress.

Model and limits: Saturated soil, positive compression and constant total stress while the ratio changes. Δu = rᵤσ′ᵥ₀ and σ′ᵥ = σ′ᵥ₀(1 − rᵤ). Inputs are prescribed states, not an earthquake duration law. USGS Professional Paper 1661-E describes the effective-stress mechanism; this diagram does not compute triggering, residual strength, lateral spreading or settlement.

Controls
100 kPa

Initial effective vertical stress

Initial grain-skeleton stress σ′ᵥ₀ = σᵥ − u₀. Changing it defines a new initial state and changes the total vertical stress. At a fixed ratio, a larger initial value gives proportionally larger excess pressure and remaining effective stress.

Range: 10–200 kPa. Step: 10 kPa.

50 kPa

Initial pore water pressure

Initial gauge pressure u₀ at the soil element. Total stress is σ′ᵥ₀ + u₀. It stays constant as the excess-pressure ratio changes. This is pressure at one point, not a rising water-table elevation.

Range: 0–200 kPa. Step: 10 kPa.

0.40

Excess-pressure ratio

rᵤ = Δu/σ′ᵥ₀ uses excess pressure above u₀, not total pore pressure. Zero leaves the initial effective stress unchanged; one gives zero effective vertical stress in this idealization. The ratio is selected, not predicted from shaking.

Range: 0–1 . Step: 0.05 .

Effective vertical stress versus prescribed excess-pressure ratioThe straight line runs from 100 kPa at ratio zero to zero kPa at ratio one. The selected ratio 0.40 gives 60.0 kPa. Total vertical stress remains 150 kPa along this line. Axes are a graph, not soil geometry or elapsed time.Effective vertical stress (kPa)0501001502000.000.250.500.751.00Excess-pressure ratio rᵤ (dimensionless)
A prescribed state on a constant-total-stress line. The point remains visible at both endpoints.
Total vertical stress
150.0 kPa
Excess pore pressure
40.0 kPa
Total pore water pressure
90.0 kPa
Effective vertical stress
60.0 kPa

σv′=σv−u\sigma'_{\mathrm{v}} = \sigma_{\mathrm{v}} - u: 150.0 − 90.0 = 60.0 kPa.

Effective stress remains positive. A positive value alone does not establish that a site is safe from cyclic deformation or liquefaction.

Problematic soil response

An unfavorable ground response to loading, wetting, drying or seepage that must be characterized for the proposed works.

Three mechanisms require different evidence

  • Expansive soil: moisture changes can produce swelling or shrinkage, particularly in soils with swelling clay minerals such as smectite. Magnitude depends on mineralogy, fabric, suction, stress and water access.
  • Collapsible soil: wetting can destabilize an open fabric under existing load. Loss of suction or bonding can reduce volume; not all loess collapses and collapse is not restricted to loess.
  • Dispersive clay: particle separation in water can promote internal erosion under susceptible chemical and hydraulic conditions. Muddy seepage signals particle transport but does not uniquely diagnose dispersion.

Characterization may include mineralogical/index evidence, swell or collapse testing under representative stress, dispersion tests and seepage/filter assessment. Foundation, moisture-control, replacement or ground-treatment options depend on depth, movement demand, loading and constructability. A material name alone cannot establish a universal foundation prohibition or solution.

Wetting makes one specimen swell, another collapse under unchanged load, and a third release particles into seepage water.

Original conceptual specimen diagrams informed by FHWA geotechnical site characterization. Dashed outlines are initial states, not a deformation scale. No numerical heave, collapse strain or dispersivity classification is implied.

Wetting makes one specimen swell, another collapse under unchanged load, and a third release particles into seepage water.

Volcanic pathways and engineering consequences

PHIVOLCS distinguishes lava, ashfall, pyroclastic density currents and lahars. Ash is fragmented volcanic material; accumulated deposits can load roofs and disrupt equipment. Lahars mix water with volcanic sediment and can remobilize deposits during rain after an eruption. Hot pyroclastic density currents are extremely dangerous; ordinary structural protection is not a substitute for official exclusion and evacuation instructions.

Distance from a vent alone does not establish safety. Check drainage connectivity, topography, eruption/deposit history, wind-dependent ash dispersion, current hazard maps and agency bulletins. A bridge study must consider debris impact, blockage, aggradation, scour, freeboard, access and residual risk. No single bridge type guarantees survival.

A synthetic plan shows a downwind ash sector distinct from a channel-connected lahar path to a crossing.

Original teaching plan informed by PHIVOLCS. The sector and channel bands are conceptual, not hazard boundaries; unshaded land is not certified safe. The crossing is an investigation location, not an approved bridge design.

A synthetic plan shows a downwind ash sector distinct from a channel-connected lahar path to a crossing.

Uniform ash deposit load

Use the deposit's appropriate bulk unit weight and thickness; wetting and nonuniform accumulation can change the load.

qash=γasht,Wash=qashAq_{\mathrm{ash}}=\gamma_{\mathrm{ash}}t,\qquad W_{\mathrm{ash}}=q_{\mathrm{ash}}A

Variables

SymbolDescriptionUnit
γash\gamma_{\mathrm{ash}}Assumed or measured ash bulk unit weight, kN/m³.-
t,At,AUniform deposit thickness, m, and horizontal projected loaded area, m².-
qash,Washq_{\mathrm{ash}},W_{\mathrm{ash}}Vertical load intensity, kPa, and total vertical load, kN; these are demand, not roof capacity.-

Coastal processes have different origins

Coastal erosion changes the shoreline through sediment transport and episodic events; rates vary in space and time. A constant-rate retreat calculation is a screening scenario, not a setback standard or a collapse date. Armoring may change sediment supply, reflection and local scour; evaluate effects beyond the protected frontage.

Storm surge is a storm-related abnormal sea-level rise, influenced by wind, pressure, bathymetry and coast shape. Tide, wave setup and individual waves also affect coastal flooding. A tsunami is a wave series generated by rapid water displacement, often by an earthquake or landslide; it can arrive without a local storm. Use PAGASA surge information and PHIVOLCS tsunami information with local evacuation plans.

For an explicitly assumed constant bed-lowering rate, scour depth equals rate times elapsed time. If the pile itself does not move, remaining embedment below the new bed is initial embedment minus that scour depth. Real scour is generally time-dependent; this bookkeeping does not determine foundation capacity or failure.

A common-datum coastal section separates astronomical tide, surge increment and an oscillating wave crest above still water.

Original schematic based on PAGASA storm-surge guidance. Wave setup, runup and site bathymetry are not modeled; the illustrated sum is not a design flood elevation. Tsunami origin is explained separately in the lesson.

A common-datum coastal section separates astronomical tide, surge increment and an oscillating wave crest above still water.

Expected annual monetary loss

The probability-weighted monetary loss over one year for a specified model; it is not a prediction of the loss in a particular year or a measure of life safety.

One-event annual loss model

For a binary annual scenario with event probability p and conditional mean damage fraction v applied to exposed asset value E.

Levent=vE,EAL=pLevent=pvEL_{\mathrm{event}}=vE,\qquad \mathrm{EAL}=pL_{\mathrm{event}}=pvE

Variables

SymbolDescriptionUnit
ppAnnual event probability from zero to one for the specified scenario, not a magnitude or a return period in years.-
vvConditional mean fraction of the specified asset value lost if the event occurs, from zero to one.-
E,LeventE,L_{\mathrm{event}}Exposed monetary asset value and conditional event loss in the same currency. If event loss is already supplied, do not multiply it by v again.-

Keep the loss model within its scope

Broader disaster risk includes life, social, environmental and service consequences. The simple monetary formula assumes a stated event and loss definition; multiple intensities or repeat events require a richer model. Avoid double counting overlapping scenarios. A building damage fraction does not establish a casualty fraction. Compare mitigation through residual loss, life safety, service continuity, maintenance and feasibility; a short undiscounted payback alone is not an approval.

Simple mitigation payback

A screening time from expected annual loss reduction; this excludes discounting, maintenance and uncertainty in actual loss timing.

Tsimple=C/(EAL0−EAL1)T_{\mathrm{simple}}=C/(\mathrm{EAL}_0-\mathrm{EAL}_1)

Variables

SymbolDescriptionUnit
C,TsimpleC,T_{\mathrm{simple}}Initial mitigation cost and simple payback time when annual expected loss reduction is positive; excludes discounting and ongoing costs.-
EAL0,EAL1\mathrm{EAL}_0,\mathrm{EAL}_1Before/after expected annual losses using compatible scenario boundaries, currency and assumptions.-

Independent annual occurrence scenario

For n independent years with unchanged annual probability p; this model does not describe clustering or changing hazards.

P(at least one event in n years)=1−(1−p)nP(\text{at least one event in }n\text{ years})=1-(1-p)^n

Variables

SymbolDescriptionUnit
ppUnchanged annual event probability from zero to one.-
nnNonnegative whole number of years; events need not occur at regular intervals.-
For independent years with one-percent annual probability, the chance of at least one event reaches about 39.5 percent in fifty years and 63.4 percent in one hundred years.

Original quantitative graph of one minus 0.99 to the power n. Constant probability and independent years are explicit teaching assumptions, not a calibrated regional hazard estimate.

For independent years with one-percent annual probability, the chance of at least one event reaches about 39.5 percent in fifty years and 63.4 percent in one hundred years.

From screening to a defensible mitigation decision

Begin with current MGB, PHIVOLCS and PAGASA information, historical evidence and a site reconnaissance. Document map scale, date, uncertainty and mechanisms; a susceptibility class is not automatically an annual probability. Build a site investigation that tests the controlling ground and water hypotheses.

Compare avoidance or relocation, modified geometry/drainage, ground treatment, structural measures and operational controls for the actual mechanism. Local slope treatment can alter a local failure likelihood even though it cannot prevent tectonic earthquakes. Establish measurable monitoring thresholds, accountable actions and contingency routes. Early warning reduces consequences only if detection, communication and timely action work together; it does not remove exposure permanently.

Key Takeaways
  • Trace source, pathway and exposed assets, including connected and secondary hazards.
  • Use multiple locations and a stable datum to distinguish total from differential movement.
  • Increased pore pressure reduces effective stress under the stated stress assumptions; a pressure ratio is not a complete liquefaction assessment.
  • Soil names, distance from a volcano and a hazard-map category do not uniquely select a foundation or bridge.
  • State units, scenario probabilities, damage definitions and uncertainty before calculating monetary loss or future movement.
  • Retain residual-risk, life-safety and monitoring decisions after choosing mitigation.

References