Landslides happen far more often than most people realize, with thousands of damaging events recorded worldwide each year. Global landslide databases consistently show the Americas and Asia as the most affected continents, with rainfall, human activity, and earthquakes ranking as the top three triggers. But the frequency varies enormously by region, season, and terrain, and the processes that set a slope up for failure are often quite different from the event that finally pushes it over the edge.
How Often Landslides Happen Around the World
Pinning down a single global number is tricky because many landslides happen in remote areas and go unrecorded. What databases do capture, though, tells a clear story about where the problem is worst. Analysis of the global landslide database maintained by the U.S. Geological Survey and other open inventories shows the Americas and Asia dominating the count for both events and fatalities, with distinct seasonal rhythms: the Americas tend to follow a two-peaked annual pattern tied to their rainy seasons, while Asia shows a single broad peak concentrated in monsoon months.1Natural Hazards. Spatial and temporal landslide distributions using global and open landslide databases Africa, Europe, and Oceania see fewer recorded events, though underreporting in parts of sub-Saharan Africa makes direct comparisons unreliable.
Recorded death tolls from landslides routinely reach the hundreds or thousands in a bad year. Single events in mountainous countries can kill dozens of people at once, and the cumulative toll in places like Nepal, Colombia, the Philippines, and parts of China is staggering over any multi-decade window. The economic damage extends well beyond lives lost: roads are severed, farmland is buried, hydropower infrastructure is damaged, and communities can be cut off for weeks.
Rainfall and Water in the Ground
Rain is the most common landslide trigger on the planet. The connection is straightforward in principle: water seeps into the ground, fills pore spaces in the soil or rock, and the resulting pressure reduces the friction holding the slope in place. In saturated soil, water exerts buoyancy on the solid grains, effectively making the slope material weigh less against its own internal resistance to shearing. In partially saturated soil, the suction forces that normally help hold grains together weaken as water content rises.2WIREs Water. Landslide hydrology: from hydrology to pore pressure Either way, enough water reaching the right depth at the right time can push a slope past its breaking point.
Not every rainstorm causes a landslide. What matters is the combination of intensity and duration, plus how much rain has already fallen in the days and weeks leading up to the event. A short, intense cloudburst can saturate the surface layer and trigger a shallow slide. A week of steady moderate rain can raise groundwater levels enough to destabilize a much deeper failure plane. Early-warning researchers have found that thresholds combining short-term rainfall intensity with antecedent moisture over one to twelve days do the best job of distinguishing conditions that actually produce slides from those that merely look threatening.3Earth Surface Processes and Landforms. Comparing threshold definition techniques for rainfall‐induced landslides: A national assessment using radar rainfall Some of these studies have revealed that certain regions experience landslides at lower accumulation levels than previously assumed.
Earthquakes and Seismic Shaking
Earthquakes are the second-most recognized landslide trigger. The mechanism is different from rainfall: ground shaking generates forces that act directly on the slope material, and if those forces exceed the strength of the rock or soil, a landslide initiates. Research comparing multiple earthquake events has shown a clear relationship between the size of the resulting landslides and the intensity of ground motion. Stronger shaking produces higher stresses that can overcome material strength at greater depths, generating larger and more dangerous failures.4Earth and Planetary Science Letters. Seismic and geological controls on earthquake-induced landslide size
Earthquake-triggered landslides tend to cluster within a zone around the epicenter, and their distribution follows the pattern of shaking intensity rather than just distance from the fault. A magnitude-7 earthquake in steep terrain can trigger thousands of individual slope failures across a wide area. The 2008 Wenchuan earthquake in China and the 2015 Gorkha earthquake in Nepal are reminders of how damaging seismically triggered landslides can be: in both events, a substantial share of the total casualties came from slope failures rather than building collapse.
Why Some Slopes Fail and Others Do Not
The trigger gets the headlines, but the geology underneath often determines whether a slope is vulnerable in the first place. Features like folds, faults, fracture networks, and pre-existing planes of weakness in rock can control where a landslide happens, what style of failure it takes, and how large it gets. A slope cut by closely spaced joints in an unfavorable orientation may need only a modest push from rain or an earthquake to fail, while an intact slope with the same geometry might hold up through far worse conditions. Lithological variety matters too: alternating layers of strong and weak rock create interfaces where sliding is more likely.
Slope angle is the most obvious factor, but it is not the whole story. Gentle slopes can fail if the materials are weak enough, especially in clay-rich soils or weathered rock. Conversely, very steep slopes in hard, intact granite can stand for millennia. The relationship between steepness and stability depends on what the slope is made of and how fractured or weathered those materials are.
Seasonal and Cold-Climate Drivers
In regions with cold winters, freeze-thaw cycles and snowmelt add a distinct seasonal dimension to landslide timing. When snow accumulated over winter melts in spring, the infiltrating water can alter the physical properties of the upper soil layers and change hydrodynamic conditions within the slope. Monitoring of loess slopes (those made of fine, wind-deposited silt common in parts of China and Central Asia) has documented significant deformation stages timed to spring warming, with total surface displacement of over 26 millimeters during a single snowmelt period following a winter of heavy snowfall.5Scientific Reports. Triggering mechanics and early warning for snowmelt-rainfall-induced loess landslide
Freeze-thaw cycles compound the problem. When water in soil pores freezes, it expands and creates small cracks. When it thaws, the now-loosened material absorbs more water, which freezes again in the next cycle, progressively weakening the slope. Research on expansive soil slopes has found that the combination of freeze-thaw cycles and snowmelt infiltration significantly increases creep deformation during the spring melting period, with more soil grains becoming involved in shear deformation over time.6Canadian Geotechnical Journal. Insights into the shallow landslide mechanism of expansive soil slope induced by freeze–thaw cycles and snowmelt infiltration This is why spring is landslide season in mountainous cold-climate regions from the Himalayas to Scandinavia to Alaska.
Human Activity as a Trigger
People make landslides happen, and they do it more often than you might expect. Road construction through hilly terrain is one of the most direct causes. Cutting into a slope to build a road removes the material that was supporting the ground above, creating a new, steeper face that may not be stable. In the Himalayas, where road-building has expanded rapidly in recent decades, road cuts are a recognized contributor to landslide frequency alongside rainfall and snowmelt.7Journal of the Geological Society of India. Road Cut Slope Stability Analysis at Kotropi Landslide Zone Along NH-154 in Himachal Pradesh, India Mining operations, reservoir filling, and even heavy agricultural irrigation can have similar destabilizing effects.
Deforestation is a slower-burning but widespread contributor. Trees and their root systems do two important things for slope stability: they pull water out of the soil through transpiration, and their roots physically bind soil particles together, adding what engineers call root cohesion. When forests are cleared, both effects diminish. Research using remote sensing and physically based models has found a higher propensity for landslides in deforested areas, with the largest increases in landslide area associated with large, connected tracts of cleared forest within a few years of cutting, when root decay outpaces any regrowth.8Water Resources Research. Deforestation Effects on Rainfall‐Induced Shallow Landslides: Remote Sensing and Physically‐Based Modelling Even patchy, small-scale forest clearing can cause landslides many years later, sometimes over a decade after the trees were removed.
The protective role of vegetation is not uniform across land-use types. Detailed modeling on China’s Loess Plateau found that farmland exhibited the highest landslide susceptibility, followed by artificial and secondary forests, with susceptibility escalating after rainfall events.9PubMed Central. Quantifying Root Cohesion Spatial Heterogeneity Using Remote Sensing for Improved Landslide Susceptibility Modeling The type, density, and root depth of vegetation all matter, so replacing a natural forest with a crop field or even a young plantation of shallow-rooted trees can leave a slope substantially more vulnerable.
Volcanic Slopes and Lateral Collapses
Volcanoes are landslide factories. Their slopes are steep, built from layers of loose ash and lava that may be weakened by hydrothermal alteration (hot, acidic fluids circulating through the rock and turning strong minerals into soft clay). A volcanic landslide can happen during an eruption, triggered by an earthquake, an explosion, or the injection of magma into the flank, or it can happen between eruptions as the weakened material simply gives way. Both intrinsic factors like internal structure and external factors like climate contribute to the instability.
What makes volcanic landslides especially dangerous is their scale and their cascading effects. A large flank collapse can produce a debris avalanche that travels tens of kilometers, and it can trigger secondary hazards including eruptions, tsunamis, lahars (volcanic mudflows), and river blockages that form natural dams. The 1980 eruption of Mount St. Helens began with one of the largest landslides in recorded history, which unroofed the magma chamber and triggered the catastrophic lateral blast. History is full of similar events at volcanoes worldwide.
Landslides Under the Ocean
Landslides do not only happen on dry land. Submarine landslides occur on continental slopes, underwater canyon walls, and volcanic island flanks, and they can move enormous volumes of sediment. Their most feared consequence is tsunami generation: a large mass of material sliding downslope on the seafloor displaces the water column above it, sending waves outward. Modeling has shown that the resulting tsunami amplitude increases with the volume of the slide and decreases with the water depth at which it occurs, meaning shallow-water landslides are especially threatening to nearby coastlines.10Journal of Marine Science and Engineering. Three-Dimensional Modeling of Tsunami Waves Triggered by Submarine Landslides Based on the Smoothed Particle Hydrodynamics Method
Some of the largest mass movements on Earth are submarine. The Storegga Slide off the coast of Norway, roughly 8,000 years ago, moved about 3,000 cubic kilometers of sediment and generated a tsunami that left deposits many meters above sea level along the Scottish and Norwegian coasts. Submarine landslides are harder to detect and monitor than their terrestrial counterparts, making them a persistent concern for offshore infrastructure like pipelines and communication cables as well as for coastal communities.
Climate Change and Permafrost Thaw
Warming temperatures are already changing landslide patterns in cold regions. In areas underlain by permafrost, thawing reduces the cohesion of hillslope materials and increases how easily water can move through the ground, making slopes more susceptible to failure.11Geomorphology. Landslide response to climate change in permafrost regions The expected trajectory is that as permafrost thaws, landslides will transition from being driven primarily by melting ground ice and perched groundwater to being driven by rainfall and freeze-thaw fracturing, essentially shifting to resemble landslide regimes in non-permafrost terrain but with added instability during the transition period.
Field observations in Alaska have documented a feedback loop that accelerates this process. When a landslide initiates in permafrost terrain, the ground disturbance exposes subsurface ice to warmer conditions, increasing heat flux into the ground. Measurements at multiple sites showed that shallow permafrost present adjacent to landslides was absent within the slide area, indicating that permafrost thaws faster inside an active landslide than in the surrounding undisturbed ground.12Geophysical Research Letters. Ongoing Landslide Deformation in Thawing Permafrost Two recently initiated landslides were still mobile, with maximum surface lowering of about a meter over the study period. This positive feedback means that as climate warming continues, new landslides in permafrost regions may remain active and growing for years or decades.
The implications stretch beyond the immediate physical hazard. Permafrost soils contain vast stores of organic carbon, and landslides that mobilize that material can release it into rivers and the atmosphere. The transition between frozen and thawed states is projected to increase both the frequency and magnitude of landslides in high-latitude and high-elevation regions over timescales ranging from seasons to centuries.11Geomorphology. Landslide response to climate change in permafrost regions
Early Warning and Prediction
Predicting exactly when and where a landslide will happen remains one of the harder problems in geoscience, but practical warning systems exist and are getting better. Most rely on rainfall thresholds: if a certain amount of rain falls within a certain time window, alerts go out. The challenge lies in setting those thresholds accurately enough that they catch real events without producing so many false alarms that people stop paying attention.
Modern approaches use radar-derived rainfall data, which can cover large areas in near real-time, and combine it with historical landslide inventories to calibrate the thresholds. A system developed for China’s Sichuan Basin, for instance, uses deep-learning-processed radar rainfall data alongside Bayesian probability analysis to set thresholds at various exceedance levels, accounting for the effect of cumulative antecedent rainfall on landslide initiation.13PubMed Central. Establishing radar-derived rainfall thresholds for a landslide early warning system: a case study in the Sichuan Basin, Southwest China Similar systems are operating in Italy, Norway, Hong Kong, and other landslide-prone regions.
Ground-based monitoring adds another layer. Instruments that track surface displacement, groundwater levels, and soil moisture can detect the creeping deformation that often precedes a catastrophic failure. When monitoring data shows accelerating movement, authorities can evacuate before the slope lets go. The limitation is cost: you can instrument a known problem slope, but you cannot instrument every hillside in a country.
What Can Be Done to Reduce the Risk
Engineering solutions range from simple to massive. Retaining walls, gabion baskets (wire cages filled with rock), soil nailing, and drainage systems are standard tools for stabilizing road cuts, construction sites, and known hazard zones. These work well for specific locations but are expensive to build and maintain, and they address the symptom rather than the cause.
Bioengineering offers a cheaper, more sustainable alternative for some situations. Planting deep-rooted vegetation on unstable slopes restores root cohesion and reduces soil moisture through transpiration. Research has highlighted bamboo species as particularly promising for landslide mitigation due to their rapid growth, dense root networks, and tolerance of degraded soils.14Advances in Bamboo Science. A global perspective on a bioengineering approach to landslide mitigation using bamboo diversity In practice, many projects combine civil engineering structures for immediate stabilization with bioengineering measures for long-term sustainability, using techniques like brush layering, jute netting, and grass seeding alongside harder infrastructure.15Universal Technical Journal of Lumbini International Academy of Science and Technology. Civil Engineering structure and Bioengineering Technique for landslide reduction and there cost comparison
Land-use planning is arguably the most effective mitigation of all, yet the hardest to implement. Keeping people and infrastructure off susceptible slopes prevents exposure entirely. Landslide hazard maps exist for many regions and can guide zoning decisions, but political and economic pressures often push development into risky areas anyway, especially in rapidly growing cities in the developing world.
Who Is Most Vulnerable
Landslide vulnerability tracks closely with economic development. Research has found a strong correlation between a region’s level of economic development and the vulnerability of its population to landslide impacts.16PubMed Central. Human Vulnerability to Landslides Wealthier countries tend to have better building codes, more robust infrastructure, effective early-warning systems, and the resources to relocate communities away from hazardous slopes. Poorer countries and communities often lack all of these. The result is that while landslides happen everywhere steep terrain and water intersect, the death toll is overwhelmingly concentrated in low- and middle-income countries.
Informal settlements on steep hillsides in cities like Freetown, Rio de Janeiro, and Tegucigalpa face especially high risk. These communities are built on slopes that would be off-limits under stricter zoning, using construction methods that offer little resistance to ground movement, and their residents often lack the resources or options to relocate. When a major rainfall event hits, the consequences can be devastating. Addressing landslide risk in these contexts requires not just better geoscience but investment in housing, infrastructure, and economic opportunity.
Landslides on Mars
Earth is not the only planet with a landslide problem. Mars hosts some of the largest landslides ever documented in the solar system, concentrated in Valles Marineris, a canyon system that dwarfs the Grand Canyon. These long-runout landslides have traveled distances exceeding 50 kilometers, and they share a common morphology: an inner zone of rotated blocks near the source and an outer zone of thin, flat-surfaced debris spread across the canyon floor.17Icarus. Spatiotemporal evolution, mineralogical composition, and transport mechanisms of long-runout landslides in Valles Marineris, Mars
What makes these Martian landslides puzzling is their extreme mobility. Their runout distances are roughly twice what laboratory experiments with dry granular flows would predict, even after accounting for the canyon’s topography.18Geophysical Research Letters. Mobility and topographic effects for large Valles Marineris landslides on Mars Compositional analysis has detected hydrated silicate minerals in many of the outer deposit zones, and landslides containing those minerals sometimes show longer runout and greater lateral spreading, suggesting that water-bearing minerals may have acted as a lubricant or otherwise reduced friction.17Icarus. Spatiotemporal evolution, mineralogical composition, and transport mechanisms of long-runout landslides in Valles Marineris, Mars Understanding what made these ancient Martian slopes fail and travel so far is not just a curiosity; it informs models of how large landslides behave in general, including worst-case scenarios on Earth where scale and low-friction conditions conspire to produce unexpectedly long runout distances.
How Landscapes Recover After a Landslide
A fresh landslide scar looks lifeless, but vegetation reclaims even badly damaged slopes given time and the right conditions. Studies tracking primary succession on earthquake-triggered landslide scars have identified certain hardy species as early colonizers. On scars from the 2016 Kumamoto earthquake in Japan, for example, the herbaceous plant giant butterbur and the introduced tree species Japanese larch were among the first to establish, achieving high frequency and coverage in the recovering landscape.19Ecological Engineering. Assessing primary vegetation recovery from earthquake-induced landslide scars
Recovery is not guaranteed everywhere on a scar, though. The same research found that vegetation recovered best in stable areas with little ongoing erosion, on moderate slopes, and where biological material from the surrounding intact landscape could serve as a seed source. Extremely steep sections, areas still actively eroding, and surfaces receiving intense direct sunlight were the slowest to revegetate. These findings matter for restoration planning: rather than planting uniformly across a landslide scar, targeting the zones most likely to support growth can produce better outcomes with fewer resources. Over years and decades, successful revegetation restores root cohesion to the slope, reducing the chance of a repeat failure and completing a cycle that links vegetation, slope stability, and the landslide process itself.