Landslides happen on every continent and in every climate zone where slopes exist, but they cluster in mountainous terrain, volcanic regions, coastal bluffs, and river valleys where gravity, water, and weak rock or soil converge. The causes split roughly into two categories: conditions that make a slope vulnerable over time (steep terrain, certain rock types, deforestation) and triggers that push a weakened slope past its breaking point (intense rainfall, earthquakes, rapid snowmelt). What makes landslide science tricky is that these two categories interact in ways that can be hard to predict, turning a hillside that looked stable for decades into a disaster zone after a single storm or wildfire.
Rainfall Is the Single Most Common Trigger
Water does more to destabilize slopes than any other force. When rain soaks into soil, it fills the tiny spaces between particles, building up what geologists call pore water pressure. That pressure essentially pushes soil grains apart, reducing the friction that holds the slope together. During extreme precipitation events, the sheer intensity and duration of rainfall can overwhelm a hillside regardless of how deep the soil is or how wet it was before the storm hit.1CATENA. Exploring extreme rainfall-triggered landslides using 3D unsaturated flow, antecedent moisture and spatially distributed soil depth That finding matters because it means even slopes in relatively dry regions can fail catastrophically if a single rainfall event is intense enough.
Typhoons and hurricanes add an extra wrinkle. The rapid drop in atmospheric pressure during these storms creates its own destabilizing effect on slopes, separate from the rain itself. Research on hillslope mechanics during typhoons shows that while rainfall-driven pore pressure builds gradually and with a time delay, the effect of a sharp atmospheric pressure drop hits almost immediately, no matter how deep the soil is or how fast water can move through it.2Natural Hazards and Earth System Sciences. Finite-hillslope analysis of landslides triggered by excess pore water pressure: the roles of atmospheric pressure and rainfall infiltration during typhoons So during a powerful typhoon, slopes get hit from two directions at once: water infiltrating from above and a sudden pressure imbalance in the ground.
Prolonged wet seasons are dangerous for a different reason. Weeks of moderate rain can saturate a slope gradually, priming it so that even a modest additional rainfall tips the balance. This is why many fatal landslides happen not during the heaviest rain of a storm but hours or even days after the peak, once the water has had time to seep deep into the ground.
Earthquakes and the Landscape They Leave Behind
Seismic shaking is the second major landslide trigger worldwide, and it tends to produce the most dramatic events. During the 2005 Kashmir earthquake, researchers found that landslide locations were controlled by a combination of strong ground shaking, steep slopes, proximity to faults and rivers, and the type of rock present.3Seismica. The influence of ground shaking on the distribution and size of coseismic landslides from the Mw 7.6 2005 Kashmir earthquake Ridge geometry also played a role: the largest landslide triggered by that earthquake, at Hattian Bala, occurred where a ridge changed direction, trapping seismic energy and amplifying the shaking at the point where the slide initiated.
Earthquake-triggered landslides tend to be especially dangerous because they strike with no rain-based warning. A hillside that has been dry and stable for years can fail in seconds during strong shaking. They are also common in remote mountain areas where roads and bridges are limited, meaning the landslides themselves can block rescue routes and isolate communities that need help. The Himalayan arc, the Andes, Central America, Japan, Taiwan, and the Philippines are all seismically active zones where earthquake-triggered landslides are a recurring threat.
Why Some Rock and Soil Fail More Easily
Not all slopes are created equal, and geology is the main reason. Research comparing landslide-prone areas with stable slopes in the same watershed found that soils in landslide zones consistently had higher moisture content, more plasticity, and weaker shear strength. Soils formed on one type of volcanic rock formation showed high water-retention capacity with relatively low resistance to shearing, making them especially slide-prone. Soils on a different, harder formation in the same region had greater shear strength and more stability.4Acta Solum. Geological Controls on Soil Shear Strength and Slope Instability in the Watugede Sub-Watershed
In practical terms, this means two neighboring valleys can have wildly different landslide risk based on the rock they sit on. Clay-rich soils that swell when wet and shrink when dry are notorious for slope failures. Volcanic ash deposits can be loose and porous, soaking up water until they liquefy. Layered sedimentary rock with weak planes between layers can slide along those planes like a deck of cards being pushed sideways. This is why geologic mapping is one of the most basic tools for landslide hazard assessment: if you know what rock and soil types underlie a slope, you can estimate how strong or weak the ground is before anything else happens to it.
Wildfire Changes the Rules
A burned hillside is a landslide waiting to happen, sometimes for years. Fire does two things to soil that matter for slope stability. First, the heat creates a water-repellent layer at or just below the surface, so rain that would normally soak in instead runs off in sheets. Second, the loss of vegetation removes root systems that were physically holding soil in place, and eliminates the canopy that slowed raindrop impact and let water absorb gradually.
Research into post-wildfire debris flows found that the type of flow you get depends on how far along the soil is in recovering from the fire. Immediately after a burn, when the soil still resists water infiltration strongly, debris flows tend to be triggered by short, intense rainstorms that generate rapid surface runoff. As time passes and the water-repellent layer begins to break down, the rainfall intensity needed to trigger a runoff-driven flow goes up, but at the same time, longer-duration rains become more dangerous because water can now penetrate deeper, building up subsurface pore pressures.5Journal of Geophysical Research: Earth Surface. Postwildfire Soil‐Hydraulic Recovery and the Persistence of Debris Flow Hazards So the hazard does not simply go away as a burned area recovers; it shifts from one failure mechanism to another.
Laboratory studies of the water-repellent layer itself show that erosion starts from many small failure patches simultaneously, and the size of sand grains in the burned soil determines how quickly those patches develop. Fine sand is the most vulnerable to failure, whether the water-repellent layer sits at the surface or slightly below it.6Canadian Geotechnical Journal. Kinematics of post-wildfire debris flow initiation mechanism: impact of hydrophobic layer spatial variability and particle size This means fire-scarred terrain with fine-grained soils needs the most cautious management and monitoring in the years following a burn.
Thawing Permafrost and Retreating Glaciers
Mountain regions at high latitudes and high elevations face a landslide problem that is getting worse with warming temperatures. Permafrost, the permanently frozen ground that cements loose rock and soil on steep slopes, loses its structural role when it thaws. And glaciers, which physically buttress valley walls by pressing against them, leave behind over-steepened slopes when they retreat. The removal of that lateral support initiates stress-release fractures and contributes to a period of heightened landslide activity that can persist for decades or longer.7Geomorphology. Landslide response to climate change in permafrost regions
This is not just an Arctic concern. High Mountain Asia, which includes the Himalayas, the Karakoram, and the Tibetan Plateau, is projected to see the greatest increase in landslide activity over areas currently covered by glaciers and glacial lakes, potentially worsening cascading hazards for communities downstream.8Geophysical Research Letters. Changes in Extreme Precipitation and Landslides Over High Mountain Asia A “cascading hazard” here means one event triggering another: a landslide falls into a glacial lake, the lake overflows or bursts its natural dam, and a flood rushes downvalley. These chain-reaction events can travel tens of kilometers from the original slide.
How Climate Change Reshapes the Global Picture
Beyond the direct effects of permafrost thaw and glacier retreat, warming temperatures drive changes in rainfall patterns that are expected to increase landslide risk broadly. Where global warming produces more frequent and more intense severe rainfall events, rapid-moving landslides become more likely. These fast-moving slides and debris flows are the ones that cause the most fatalities because people have so little time to react.9Earth-Science Reviews. Landslides in a changing climate
The connection is not uniform around the world. Some regions are projected to get drier overall, which would reduce rainfall-triggered landslides but might increase the risk of fire-related slides. Other regions are seeing shifts in the timing of wet seasons, so slopes that historically had months to dry out between storms may stay saturated longer. The real concern is that landslide hazard zones are shifting into areas that historically had less risk and therefore have less infrastructure, fewer warning systems, and communities that have less experience recognizing the danger signs.
Human Activity as Both Cause and Amplifier
People reshape slopes constantly through road construction, mining, terracing, deforestation, and urban development. Cutting into a hillside to build a road removes material at the base that was supporting the slope above. Clearing forests strips away root networks and exposes bare soil to erosion. Poorly managed irrigation or drainage can saturate slopes that would otherwise stay dry.
A large-scale statistical analysis of fatal landslides in mountain regions worldwide found something that should reframe how we think about landslide risk: land-use and land-cover changes have a substantially greater influence on the density of fatal landslides and landslide fatalities than physical factors like topography and precipitation, especially in lower-income countries. When land-use change was low, landslide impacts were marginal regardless of a country’s income level.10PubMed Central. Wealth and land-cover change govern landslide fatalities on world’s mountains That is a striking result: it suggests that for many of the world’s most vulnerable communities, the single most effective way to reduce landslide deaths is not better engineering or better weather forecasting, but better land-use planning.
This finding connects to a broader pattern. Landslide vulnerability is strongly correlated with the economic development of a region.11PubMed Central. Human Vulnerability to Landslides Wealthier countries tend to have stricter building codes, more robust drainage infrastructure, better monitoring networks, and the resources to relocate communities out of high-risk zones. Lower-income countries often cannot afford those measures and face additional pressures from rapid urbanization that push settlements onto unstable terrain. Unlike some natural hazards where gender and age heavily stratify who gets killed, landslide losses do not show the same strong demographic patterns, which suggests the hazard is more about location and exposure than individual physical vulnerability.
Landslides Under the Ocean
Not all landslides happen on land. Submarine landslides occur on continental slopes, around volcanic islands, and near underwater canyon walls. They can be enormous, sometimes involving thousands of cubic kilometers of sediment. When that mass of material displaces the water above it, it generates tsunamis. The size and speed of the resulting waves depend on the landslide’s volume, its initial acceleration, the density of the material, and the angle of the slope.12Marine and Petroleum Geology. A review of approaches for submarine landslide-tsunami hazard identification and assessment
Submarine landslide tsunamis behave differently from those generated by earthquakes. They tend to produce shorter-wavelength waves that spread radially and dissipate more quickly with distance, but can be devastating to nearby coastlines. The 1998 Papua New Guinea tsunami, which killed over 2,000 people, was generated at least in part by a submarine landslide rather than by the earthquake alone. Identifying submarine landslide hazards is difficult because the slopes in question are deep underwater and often poorly mapped, making this one of the more challenging frontiers in landslide science.
Can We Predict When a Slope Will Fail?
Predicting the exact timing of a landslide remains one of the hardest problems in geoscience, but monitoring technology has improved considerably. Satellite-based radar interferometry can detect subtle ground deformation over large areas, measuring movement of millimeters per year. A study of thirty historical landslides in central and western China found that about a third were captured by satellite radar showing accelerated deformation before failure, and roughly a quarter provided enough data to predict their failure time.13Engineering Geology. Can satellite InSAR innovate the way of large landslide early warning? Those numbers highlight both the promise and the limitation: the technology works impressively for some slides, but more than half of the landslides studied showed no clear warning signal that the satellite could pick up.
On the ground, monitoring tools include inclinometers drilled into slopes to track internal movement, rain gauges tied to threshold-based alert systems, and networks of GPS stations that detect surface creep. Japan, Italy, and Hong Kong have some of the most sophisticated landslide warning systems in the world, combining real-time rainfall data with slope-specific hazard models to issue alerts when dangerous thresholds are crossed. But these systems require sustained investment and technical infrastructure that many landslide-prone regions simply do not have.
The Places Most at Risk
If you mapped every reported fatal landslide over the past few decades, clusters would appear along the Himalayan arc from Afghanistan through Nepal and into Myanmar; across Southeast Asia, particularly in the Philippines and Indonesia; through Central America and the northern Andes; in parts of East Africa; and in southern Europe around the Alps and the Italian peninsula. These hotspots share some combination of steep terrain, heavy seasonal rainfall, seismic activity, and dense human settlement on or near unstable slopes.
But landslides are not exclusively a developing-world problem. The Pacific Northwest of the United States, the Norwegian fjord coast, the Japanese archipelago, and the European Alps all experience damaging landslides regularly. What differs is the consequence: wealthier regions tend to lose property and infrastructure, while lower-income regions lose lives, often in much greater numbers for comparable-sized events.
Urban expansion into hillside terrain is one of the fastest-growing risk factors worldwide. Informal settlements on steep slopes around cities in Latin America, Africa, and South Asia are particularly vulnerable because the construction itself often destabilizes the ground (through excavation, added weight, and altered drainage) while the residents have the fewest resources to evacuate or rebuild.
Local Knowledge and Its Underused Role
In many landslide-prone communities, especially in rural mountainous areas, local and indigenous knowledge systems contain generations of observation about slope behavior, warning signs, and land management practices that reduce risk. Research in Kanungu District, Uganda, found that factors like farming experience, access to weather information, and perception of climate changes influenced whether households adopted indigenous landslide management practices. The study also found that marginalization was a major barrier: local knowledge was often sidelined in official disaster management approaches.14Nova Geodesia. The complementary role of indigenous knowledge systems in landslide disaster management in Kanungu District, Uganda
This matters because in remote areas where satellite monitoring and engineering solutions are financially out of reach, community-based risk awareness may be the most practical first line of defense. Recognizing changes in spring flow, new cracks in the ground, tilting trees, or unusual sounds from a hillside are observational skills that have saved lives long before modern instrumentation existed.
Historical Catastrophes and What They Teach
Some of the most devastating landslides in recorded history have reshaped how scientists and governments think about slope hazards. The 1618 Piuro landslide in the Italian Alps buried an entire village, killing an estimated 1,000 to 2,000 people, and the event left such a mark on collective memory that it influenced artistic representations for centuries.15Copernicus Publications (EGUsphere). A multidisciplinary approach to improve and share the understanding of landslide hazard in mountain environments: the PIURO 1618 disaster Understanding the geomorphic dynamics of landscapes like the one that produced the Piuro disaster remains central to modern risk assessment, because similar conditions exist today in valleys across the Alps and other mountain ranges.
More recent catastrophes have driven policy changes. After the 1966 landslides and floods that devastated parts of northeastern Italy, the Italian government created one of Europe’s first national-scale landslide inventories. The 1985 Armero tragedy in Colombia, where a volcanic lahar killed more than 23,000 people, led to widespread reforms in volcanic hazard communication. Each major disaster tends to expose a gap, whether in monitoring, in land-use regulation, or simply in public awareness, and the response fills that gap for a while until the next event reveals a new one.