What Is a Geohazard? Types, Risks, and Prevention

A geohazard is any geological or ground-related condition that poses a threat to people, property, or infrastructure. The term is a contraction of “geological hazard,” and while it sounds straightforward, experts across disciplines disagree on exactly what it covers. Researchers in engineering geology, geography, and geotechnical engineering each bring different examples to the table, and the resulting diversity of classification schemes reflects the fact that “geohazard” is less a precise scientific category than a practical umbrella for a wide range of Earth processes that can hurt us.

What Counts as a Geohazard

The term emerged somewhat independently in several fields, and that parallel evolution has left it with fuzzy edges. A geographer studying disaster management might emphasize floods and droughts. An engineering geologist might focus on landslides and ground settlement. A geotechnical engineer might zero in on soil liquefaction beneath a building foundation. A 2024 paper in the Quarterly Journal of Engineering Geology and Hydrogeology traced this confusion, noting that while “everyone knows what a geohazard is,” the examples people offer “betray significant differences of opinion.”¹ The most useful way to think about geohazards is in the context of risk: a geological process only becomes a hazard when it has the potential to produce undesirable consequences for people or what they have built.

With that framing, the major categories include earthquakes and related seismic effects, volcanic eruptions, landslides and other forms of mass movement, ground subsidence and sinkholes, and coastal erosion. Some definitions also fold in flooding and soil erosion when geological conditions drive or worsen the event. And critically, human activity can create geohazards that would not otherwise exist, or amplify natural ones.

Earthquakes and Seismic Effects

Earthquakes are the geohazard most people picture first, and for good reason. Between 1900 and 2022, earthquakes accounted for about 41% of all direct economic losses from natural disasters worldwide, more than storms, floods, or any other single category.² But the shaking itself is only part of the story. What the ground beneath you is made of dramatically affects how much damage an earthquake can do.

Soft soil layers can amplify seismic waves as they travel upward from bedrock, a phenomenon called soil amplification. In layered ground profiles, researchers have measured amplification factors ranging from less than 2 up to roughly 5.7, depending on the arrangement and type of soil.³ Clay at the surface tends to produce higher amplification and longer-period shaking, while sandy surfaces generate stronger short-period amplification. This is why two buildings the same distance from an earthquake’s epicenter can experience very different levels of damage: the geology directly underfoot matters as much as proximity to the fault.⁴

Earthquakes can also trigger liquefaction, where saturated, loose soil temporarily behaves like a liquid during shaking. The mechanism involves a rapid loss of shear strength in the soil as water pressure spikes between grains.⁵ Structures built on liquefiable ground can tilt, sink, or collapse even if they were well-engineered for ordinary shaking.

Landslides and Slope Instability

Landslides range from slow, creeping soil movements that take years to damage a house’s foundation to catastrophic debris flows that bury entire communities in minutes. The underlying mechanism almost always involves a loss of strength in the material making up a slope. Two of the most common triggers are heavy rainfall and earthquakes, and both work through a similar pathway: they increase water pressure inside the slope material, which reduces the friction holding the slope in place.⁶

When an earthquake hits a saturated slope, the cyclic loading can generate excess water pressure that persists after the shaking stops. Laboratory testing has shown that this sustained pressure, combined with the inability of water to drain quickly enough through thick soil layers, can keep a landslide moving long after the triggering earthquake is over.⁷ This explains why earthquake-triggered landslides sometimes seem to develop slowly, with slopes continuing to fail hours or days after the main shock.

Human activities compound the problem. Deforestation removes root systems that help anchor soil. Road cuts steepen slopes beyond their natural angle of stability. And poorly designed drainage around construction sites concentrates water in exactly the wrong places.

Volcanic Hazards

Volcanoes produce a remarkably diverse set of geohazards. Ashfall (technically called tephra fall) is the most widespread, capable of blanketing areas hundreds of kilometers from an eruption. It disrupts transportation, contaminates water supplies, and can collapse roofs under its weight. But the truly destructive volcanic hazards are the flow types.

Pyroclastic density currents are fast-moving mixtures of hot volcanic debris and gas that race downhill at speeds up to 300 meters per second with temperatures reaching 1,100 °C. They generate dynamic pressures strong enough to flatten buildings and are hot enough to ignite common construction materials on contact.⁸ Lava flows move more slowly but destroy everything in their path through sheer heat and mass. Lahars, volcanic mudflows formed when erupted material mixes with water, can travel tens of kilometers down river valleys, growing in volume as they pick up sediment. A review of eruptions over the past century found that while ashfall mainly disrupts infrastructure and reduces its function, flow hazards such as pyroclastic currents, lava flows, and lahars cause permanent, catastrophic damage.⁸

Subsidence and Sinkholes

Ground subsidence is a slower geohazard but an expensive one. It happens when the ground surface drops, often because water or oil has been withdrawn from underground, compressing the layers above. In some regions, subsidence occurs at rates of centimeters per year, gradually damaging foundations, cracking roads, and warping buried pipelines.

Sinkholes are the dramatic cousin of subsidence, especially common in karst terrain where soluble rock like limestone underlies the surface. When groundwater levels drop in these areas, a negative-pressure zone forms underground, causing tensile failure in the surrounding soil and creating an underground void. If the overlying soil is thin, the surface collapses at the same time. If it is thick, internal collapse happens first, progressing layer by layer upward until the surface suddenly gives way and a sinkhole appears.⁹ This staged process is why sinkholes can seem to appear without warning: the collapse has been progressing underground for weeks or months before the surface fails.

Coastal Erosion and Cliff Retreat

Along coastlines, wave energy, rising sea levels, and rainfall combine to eat away at cliffs and shorelines. Research along the Santa Barbara coast in California found that cliff base retreat is strongly linked to wave energy flux, with a threshold effect: retreat rates spiked when average water levels sat more than 0.8 meters above the cliff toe.¹⁰ The base and middle of the cliff erode during high-energy swell events, while the top retreats more with seasonal rainfall. These are distinct processes working on the same cliff face, and they rarely happen at the same time, which makes prediction tricky.

Sea-level rise adds another layer of complexity. Modeling of future cliff-front wave conditions suggests that rising seas will shift the wave regime at many coastal cliff sites, increasing the proportion of unbroken waves hitting the cliff rather than waves that have already broken and dissipated energy.¹¹ At some sites, this shift could paradoxically reduce cliff retreat rates because unbroken waves deliver energy differently than broken surf. But at other sites, deeper water at the cliff base could mean more sustained wave attack. The outcome depends heavily on local platform geometry, which is why blanket predictions about coastal erosion and climate change are unreliable.

Human Activity as a Trigger

Not all geohazards originate purely from natural processes. It has been understood for decades that human activities including reservoir impoundment, mining, fluid withdrawal, and underground injection can trigger earthquakes.¹² Wastewater injection from oil and gas operations, in particular, has been linked to sharp increases in seismicity in regions that previously experienced very few earthquakes.

Satellite radar imagery over West Texas has documented ground surface uplift in areas where fluids are being injected underground. The injected formation experiences rising pore pressure and decreasing effective stress, which promotes upward movement of the ground surface.¹³ This uplift can damage infrastructure directly and may also contribute to fault reactivation in the surrounding area. The broader point is that geohazard assessment increasingly needs to account for industrial activity alongside natural geological processes.

Why Geohazards Cascade

One of the most dangerous characteristics of geohazards is that they rarely occur in isolation. An earthquake can trigger landslides, which dam rivers, which then produce catastrophic floods when the dam breaches. A wildfire can strip vegetation from a hillside, which then fails as a debris flow during the next heavy rain. Researchers describe these chains as cascading hazards, where surface processes like coseismic landslides and post-fire debris flows form complex sequences that make subsequent hazards more likely.¹⁴

Glacial environments offer a vivid example. At Ranzerio Lake, a glacier tongue collapse with an estimated volume of 3.8 million cubic meters was identified as the primary trigger of a moraine dam breach, which in turn generated a glacial lake outburst flood.¹⁵ The glacier collapse, the dam failure, and the downstream flood are three hazards linked in a single chain. Managing any one of them in isolation would miss the bigger picture.

Permafrost Thaw and Climate Feedbacks

Climate change is reshaping the geohazard landscape in high-latitude and high-elevation regions. Rapid permafrost thaw increases hillslope susceptibility to landsliding by reducing cohesion in the ground and increasing the flow of water through previously frozen material.¹⁶ This is expected to increase both the frequency and size of landslides in coming decades across vast areas of Alaska, Siberia, Scandinavia, and mountainous regions worldwide.

Field observations in Alaska have documented a troubling feedback loop. When permafrost thaws and initiates a landslide, the ground disturbance within the slide area increases heat transfer to the subsurface, which causes permafrost to thaw even faster inside the landslide than in surrounding undisturbed terrain. Two recently initiated landslides in discontinuous permafrost were still moving years later, with maximum ground surface drops of around 0.8 to 1.0 meters over the study period.¹⁷ In mountain environments, permafrost degradation may also lead to thaw settlement and reduced slope stability, threatening infrastructure like roads, pipelines, and ski resorts built in areas once considered stable.¹⁸

Monitoring From Space

One of the most significant advances in geohazard management over the past two decades has been satellite-based monitoring. Interferometric Synthetic Aperture Radar, or InSAR, uses radar signals bounced off the ground from orbiting satellites to measure tiny changes in surface elevation over time. The technique can detect ground deformation of just a few millimeters, making it possible to spot early signs of subsidence, landslide creep, or volcanic inflation long before they become emergencies.¹⁹

The European Copernicus program’s Sentinel-1 satellites have been particularly transformative. Their regular revisit schedule produces enormous stacks of radar images that can be processed to track deformation across entire regions. Researchers have processed hundreds of Sentinel-1 scenes over coastal southern Italy to monitor both land subsidence and landslide movement simultaneously, demonstrating the potential for continuous, wide-area surveillance.²⁰ The challenge now is less about data collection and more about processing speed and interpretation, turning terabytes of radar imagery into actionable warnings before the ground moves.

Susceptibility Mapping and Prediction

Knowing where geohazards are most likely to strike is essential for land-use planning, building codes, and emergency preparedness. Modern susceptibility mapping combines geological data, terrain characteristics, land cover, rainfall patterns, and increasingly, machine learning algorithms. In China’s Greater Bay Area, one of the most densely urbanized regions on Earth, researchers evaluated several machine-learning approaches for predicting multiple types of geohazards simultaneously and found that gradient-boosting models performed best, achieving high predictive accuracy.²¹ That study also investigated the interactions between different geohazard types, recognizing that in urban settings, subsidence, landslides, and flooding rarely occur as independent events.

Paleoseismology offers a much longer view. By studying geological evidence of ancient earthquakes, including displaced layers in trenches dug across faults, disturbed tree rings, and sediment records from nearby lakes, scientists can estimate how often a fault produces large earthquakes and how intense the shaking tends to be.²² On New Zealand’s Alpine Fault, paleoseismic work has progressed from simply confirming that large surface-rupturing earthquakes occur to estimating recurrence behavior, shaking intensities, and even the likelihood of the next event.²³ This kind of deep-time evidence fills a gap that instrumental records, which only go back about a century, cannot.

Prevention and Engineering Solutions

Preventing geohazards entirely is often impossible, but reducing the risk they pose is not. Engineering approaches target the specific mechanism involved. For landslide-prone slopes, retaining walls are a common solution, but their design matters enormously. Numerical modeling has shown that introducing additional structural elements like buttresses and struts into retaining wall designs can optimize material use while maintaining slope stability, essentially achieving better safety with less concrete and steel.²⁴

Nature-based solutions are gaining traction alongside traditional engineering. Green infrastructure such as vegetated corridors and buffer strips around transportation networks can reduce soil erosion and help maintain slope stability. A systematic review of green infrastructure for seismic resilience found that strategically placed vegetation and engineered wetlands can protect critical lifeline systems including water supply, transportation, and communication networks during and after earthquakes.²⁵ These approaches work best not as replacements for hard engineering but as complements, adding layers of redundancy and ecological benefits that purely structural solutions cannot provide.

Indigenous Knowledge and Community-Based Warning

Long before satellite radar and machine learning, communities living in geologically active regions developed their own approaches to recognizing and surviving hazards. On the Mentawai Islands west of Sumatra, ancestors observed the behavior of animals like squirrels and chickens as indicators of impending earthquakes and tsunamis. This knowledge was traditionally passed down through storytelling and songs about correct behavior during disasters.²⁶ These methods function without any technical infrastructure, making them valuable in remote areas where modern warning systems are absent or unreliable.

The challenge is that this knowledge erodes with each generation that does not experience a disaster firsthand. Researchers studying indigenous early warning indicators have argued that integrating these community-based observations with modern meteorological forecasts can strengthen disaster risk management, particularly for rural populations.²⁷ The idea is not to treat traditional knowledge as a quaint alternative to science, but to recognize that centuries of lived experience in a specific landscape encode real observational data about how that landscape behaves before, during, and after geological events.

The Economic Scale of Geohazards

Between 1900 and 2022, natural disasters caused roughly $5.2 trillion in direct economic losses globally. Earthquakes alone accounted for about $2.1 trillion of that total, followed by storms at $1.2 trillion and floods at $1.1 trillion. Landslides and volcanic activity contributed smaller but still significant shares.² These figures capture only direct losses: destroyed buildings, damaged infrastructure, lost agricultural production. They do not include the longer-term economic drag of displaced populations, disrupted supply chains, or the cost of rebuilding to higher standards.

The distribution of these losses is deeply unequal. Wealthy nations absorb higher absolute dollar losses because they have more expensive infrastructure in harm’s way, but they also recover faster. Lower-income countries often suffer smaller absolute losses that nonetheless represent a crippling share of national GDP. A volcanic eruption that destroys a single highway in a small island nation may cut off trade routes that the entire economy depends on, while the same event near a major city in a wealthy country, though costlier in dollar terms, draws on national resources for rapid recovery. Geohazard prevention investment, as a result, yields disproportionately high returns in regions with the fewest resources to recover from disasters without it.