Skip to content

Landslide-Dam Failure

The two-stage natural-hazard cascade in which mass wasting dams a valley, impounds a lake, and later fails — releasing a downstream outburst flood usually larger than the landslide, because the unengineered barrier is itself the hazard's storage element.

Core Idea

A landslide-dam failure is the natural-hazard cascade in which mass wasting — a landslide, rockfall, or debris flow — deposits material across a river valley, forming a temporary dam that impounds an upstream lake, which subsequently fails, releasing a rapid outburst flood downstream. The structural distinctive is that the mechanism is two-stage and cross-scale: a locally contained trigger event creates a standing regional risk — the impounded reservoir — that later generates a downstream catastrophe typically larger in energy and spatial extent than the original landslide. The barrier is not engineered: it has no spillway, is composed of poorly sorted, heterogeneous debris, and is geotechnically unstable in ways that make its time-to-failure distribution short-tailed and difficult to forecast precisely. Costa and Schuster's 1988 analysis of historical landslide dams established that a substantial fraction fail within days to months, and that overtopping, piping, and slope collapse are the three dominant failure modes. The structural commitment the concept enforces is that the barrier is itself the hazard's storage element — removing it rapidly is dangerous because it releases the stored load; leaving it is dangerous because it will release the load anyway, on a schedule the impoundment volume and barrier geometry constrain but do not pin down. The intervention logic follows from this: controlled drawdown (excavating a spillway to drain the lake before overtopping triggers uncontrolled failure), downstream evacuation and exclusion zoning during the impoundment window, and lake-level monitoring to track the race between accumulation and barrier degradation. The same lifecycle — trigger → deposit → impoundment → failure mode → outburst — applies to the glacial-lake outburst flood (GLOF) variant, in which a moraine or ice dam plays the role of the landslide deposit; the Tangjiashan barrier lake formed by the 2008 Wenchuan earthquake (impounding over 100 million cubic metres above a densely populated valley, requiring controlled excavation over several weeks and the downstream evacuation of 250,000 people) is the canonical modern case.

Structural Signature

Sig role-phrases:

  • the triggering mass-wasting event — the landslide, rockfall, debris flow (or, in the GLOF variant, the moraine/ice dam) that puts material across the valley
  • the unengineered barrier — the cross-valley deposit: no spillway, poorly-sorted heterogeneous debris, geotechnically unstable, and the hazard's storage element
  • the impounded reservoir — the upstream lake whose volume accumulates behind the barrier and routinely stores more energy than the landslide that formed it
  • the accumulation-versus-degradation race — the central dynamic: rising lake volume against the weakening barrier, the relationship that sets whether a drawdown window exists
  • the failure mode in play — overtopping, piping, or slope collapse, read from barrier composition and geometry, setting the character of release
  • the short right-skewed time-to-failure — a substantial fraction fail within days to months, so the intervention window is narrow and the urgency follows from the distribution's shape
  • the downstream outburst flood — the time-delayed, catastrophic, downstream-directed release, the secondary hazard that arrives last and largest, often in a different community
  • the drawdown intervention window — the interval between formation and natural failure in which controlled drawdown (excavated spillway) plus downstream exclusion plus lake-level/deformation monitoring can bleed the load off gently

What It Is Not

  • Not a single-stage landslide. The slope failure is only the trigger; the cascade is two-stage and cross-scale — trigger → deposit → impoundment → failure mode → outburst. A response that clears the slope but ignores the impounded lake has not closed the event; it has armed the larger one, which arrives later, downstream, and often in a different community.
  • Not a hazard whose worst harm is the landslide. Because the impounded reservoir routinely stores more energy than the mass that dammed it, the secondary outburst flood typically dominates the original landslide in magnitude and reach. The downstream threat must be sized from the lake, not the slope — the largest consequence is the one that has not happened yet.
  • Not a barrier that is safer removed. The barrier is the hazard's storage element, so removing it rapidly is dangerous precisely because it releases the stored load at once. The intervention is controlled drawdown — bleeding the load off on a chosen gentler schedule via an excavated spillway — not demolition, and the paradox is that leaving the barrier is dangerous too, since it will release the load anyway.
  • Not a steady leak. The release is time-delayed and catastrophic — overtopping, piping, or collapse producing a sudden outburst — not a continuous escape through the boundary. A barrier that slowly seeps without ever bursting is a different regime; the defining hazard is the discontinuous release of an accumulated load.
  • Not a useful buffer to preserve. A maintained intermediate capacity is normally desirable; the impounded lake is an unintended buffer whose very release is the hazard. The reasoning is about eliminating a storage element under control, not about protecting one.
  • Not a release on a known date. Time-to-failure is a short, heavily right-skewed distribution — a substantial fraction fail within days to months — not a deterministic schedule. The urgency follows from the distribution's shape and the narrow drawdown window, not from any predictable failure moment.

Scope of Application

Landslide-dam failure lives within the natural-hazards and geomorphology subfields of earth and environmental science — across the whole family of debris-impoundment events regardless of what put the barrier in the valley; its reach is bounded to that physical-hazard substrate, and the substrate-neutral blockage-release dynamic it instantiates (which recurs in volcanism, embolism, and pent-up demand) is carried by an emergent blockage-release pattern, not by this geomorphic concept.

  • Mountain-river barrier lakes (canonical) — earthquake- or rainfall-triggered rock avalanches and landslides dam a valley, as at Tangjiashan after the 2008 Wenchuan earthquake (>100 million m³ impounded, weeks of controlled drawdown, 250,000 evacuated downstream).
  • Glacial-lake outburst floods (GLOFs) — moraine- or ice-dammed lakes in the Himalaya, Andes, and Alaska (Imja Tsho, Cordillera Blanca) run the identical lifecycle with a moraine/ice dam filling the deposit slot, with dedicated GLOF early-warning monitoring.
  • Volcanic-debris dams — lahar deposits choke valleys after eruptions (Mount St. Helens, Pinatubo), impounding lakes whose outburst dominates the eruption's local damage.
  • Ice jams and log jams — seasonal river-ice or log-debris impoundments on northern rivers produce the same outburst behaviour at smaller scale.

Clarity

Naming landslide-dam failure makes legible that a landslide need not be a single-stage event. The label binds the dam-formation and the dam-failure into one cascade — trigger, deposit, impoundment, failure mode, outburst — so an emergency manager stops treating them as two unrelated incidents and recognises that the deposit persists as a standing secondary hazard for weeks to years, and that the downstream outburst typically dominates the local landslide it grew from. The sharp inference the concept licenses is that a landslide response which clears the slope but ignores the impounded lake has not closed the event; it has merely armed the larger one.

The clarifying core is the recognition that the barrier is itself the hazard's storage element. That single framing organises the otherwise paradoxical intervention logic: removing the barrier rapidly is dangerous because it releases the stored load, yet leaving it is dangerous because it will release the load anyway, on a schedule the impoundment volume and barrier geometry constrain but do not pin down. Holding that in view turns the problem into a race a practitioner can actually reason about — accumulation against barrier degradation — and makes the operative question crisp: is there an intervention window between formation and natural failure long enough to drain the lake by controlled drawdown, and is the downstream population cleared of the outburst path while it lasts? The unengineered nature of the barrier — no spillway, heterogeneous debris, a short and hard-to-forecast time-to-failure — is what makes that window narrow and the monitoring of lake level and barrier deformation load-bearing rather than optional.

Manages Complexity

The events the concept covers look superficially unalike: an earthquake-triggered rock avalanche burying a Sichuan river, a moraine-dammed glacial lake in the Himalaya, a lahar deposit choking a valley below a volcano, a seasonal ice jam on a northern river. Treated separately they are four hazard literatures with four trigger physics. Landslide-dam failure collapses them onto one lifecycle — trigger → deposit → impoundment → failure mode → outburst — so the analyst stops reasoning from the trigger and reasons instead from a fixed sequence whose stages are shared regardless of what put the barrier there; the glacial-lake outburst variant differs only in that a moraine or ice dam fills the deposit slot. Within that lifecycle the central compression is recognizing the barrier as the hazard's storage element, which converts an otherwise paradoxical intervention problem into a single legible race: impoundment volume accumulating against barrier degradation. From a small parameter set — impoundment volume, barrier geometry and composition, the failure mode in play (overtopping, piping, or collapse), and the resulting short, right-skewed time-to-failure distribution — the analyst reads off the one decision that matters: is there an intervention window between formation and natural failure long enough to drain the lake by controlled drawdown, and can the downstream population be cleared of the outburst path while it lasts. The branch structure is fixed and small. Remove the barrier fast and you release the stored load now; leave it and it releases the load anyway on a schedule the volume and geometry constrain but cannot pin down; the resolution is controlled drawdown plus downstream exclusion plus lake-level and deformation monitoring to track the race. And one qualitative ordering reads straight off the framing: because the impounded reservoir routinely stores more energy than the landslide that formed it, the secondary outburst dominates the primary event — so a response that clears the slope but ignores the lake has not closed the hazard, it has armed the larger one. The heterogeneous family of debris-dam catastrophes thereby reduces to one staged typology and a handful of barrier-and-impoundment parameters off which the outburst magnitude, the failure timing, and the existence of a drawdown window all follow.

Abstract Reasoning

Landslide-dam failure licenses a set of reasoning moves built on the recognition that the barrier is the hazard's storage element and that the hazard is two-stage and cross-scale — letting the hazard analyst reason from barrier-and-impoundment parameters to outburst magnitude and timing, and frame the response as a race between accumulation and barrier degradation.

Diagnostic — read the standing secondary hazard, not just the trigger. The defining inference is that a landslide which deposits across a valley has not produced a single-stage event but armed a larger one: the deposit is diagnosed as a standing barrier impounding a reservoir, and the reservoir is read as a hazard that will persist for weeks to years and typically dominate the landslide it grew from. A second diagnostic infers the failure mode from barrier composition and geometry — overtopping read from a filling lake approaching the deposit crest, piping from seepage through poorly-sorted heterogeneous debris, slope collapse from an unfavourable height-to-width ratio and saturation — because the mode in play sets the character and timing of release. A third diagnostic infers outburst magnitude from impoundment volume rather than from the original landslide: because the reservoir routinely stores more energy than the mass that dammed it, the secondary flood is predicted to exceed the primary event, so the analyst sizes the downstream threat from the lake, not the slope. The governing inference throughout: a response that clears the slope but ignores the impounded lake has not closed the hazard.

Interventionist — act on the storage element, and predict release-on-a-schedule versus release-on-its-own. The paradoxical intervention logic is organised by the storage framing: removing the barrier rapidly is dangerous because it releases the stored load now; leaving it is dangerous because it will release the load anyway, on a schedule the volume and geometry constrain but do not pin down. The licensed move is controlled drawdown — excavating a spillway to drain the lake before overtopping triggers uncontrolled failure — predicting that the load is released on a chosen, gentler schedule instead of catastrophically. Paired with it: downstream evacuation and exclusion zoning during the impoundment window, predicting the population is clear of the outburst path whenever release comes, controlled or not; and lake-level and barrier-deformation monitoring (gauges, satellite altimetry, time-lapse imagery), predicting that the race between accumulating volume and degrading barrier stays observable so the drawdown can be timed and the warning issued before failure. The interventionist invariant: the lever acts on the barrier-as-storage and on the downstream exposure, and the decisive question is whether an intervention window exists between formation and natural failure long enough to drain the lake while the population is held clear.

Boundary-drawing — storage-and-release, not phase transition and not steady leakage. The concept is a cascade with a storage stage between trigger and release, which distinguishes it from a bare propagating cascade (no impoundment) and locates its distinctive feature precisely. It has a tipping element — the barrier's failure — but its structure is storage-and-release, not phase transition, so it should not be reasoned about as a continuous order-parameter change. It is bounded away from steady leakage: the release is time-delayed and catastrophic, not a continuous escape through the boundary, so a slowly seeping barrier that never bursts is a different regime. And the impounded lake is an unintended buffer whose release is the hazard, the inverse of a desirable maintained buffer — so the reasoning is about eliminating a storage element, not preserving one. A scope boundary marks the family the lifecycle covers: trigger → deposit → impoundment → failure mode → outburst applies to earthquake-, rainfall-, volcanic-, glacial-, and ice-triggered barriers alike, with the glacial-lake outburst variant differing only in that a moraine or ice dam fills the deposit slot.

Predictive / order-of-events. The concept commits the analyst to a forecast about sequence and timing. It predicts a short, heavily right-skewed time-to-failure distribution — a substantial fraction of landslide dams fail within days to months — so the intervention window is narrow and the urgency follows from the distribution's shape, not from any single deterministic date. It predicts the ordering of consequences: the local landslide first, the standing impoundment hazard next, the downstream outburst last and largest, so the worst harm arrives after and below the initiating event, often in a different community. And it predicts the counterfactual that justifies costly intervention: an unmodified barrier left to overtop or pipe will release a flood orders of magnitude larger than the original landslide damage — the reasoning that converts "a slope failed upstream" into "evacuate a quarter-million people and excavate a spillway downstream," as at Tangjiashan after the 2008 Wenchuan earthquake.

Knowledge Transfer

Within natural hazards the concept transfers as mechanism across the whole family of debris-impoundment events, regardless of what put the barrier in the valley. Earthquake-triggered rock avalanches (Tangjiashan), rainfall-driven landslides, volcanic lahar deposits (Mount St. Helens, Pinatubo), glacial-lake outburst floods behind moraine or ice dams (Cordillera Blanca, Imja Tsho), and seasonal ice or log jams all run the identical lifecycle — trigger → deposit → impoundment → failure mode → outburst — with the GLOF variant differing only in that a moraine or ice dam fills the deposit slot. The full apparatus carries intact: the barrier-as-storage-element framing, the accumulation-versus-degradation race, the failure-mode triad (overtopping, piping, collapse), the short right-skewed time-to-failure distribution, and the intervention menu (controlled drawdown/spillway, downstream evacuation and exclusion zoning, lake-level and deformation monitoring, geotechnical stability assessment). The geophysics of the trigger changes but the storage-and-release reasoning and its operative vocabulary need no translation, because every member genuinely has physical material accumulating behind a physically unstable barrier — this is one mechanism sampled across hazard types, not an analogy stretched across them.

Beyond the physical-hazard substrate the transfer is bimodal, and the split is worth stating exactly. Lifted whole, "landslide-dam failure" travels only by analogy: its operative vocabulary (controlled breach, spillway, lake-level monitoring, geotechnical seepage) presupposes physical accumulation behind a physical barrier whose stability is a geotechnical property, so invoking it for a financial, public-health, or governance system borrows the dam-burst shape while dropping the storage element that gives the original its predictive structure. A pent-up-volatility-released-by-a-regulatory-barrier story is more cleanly carried by pressure_release or volatility-clustering/credit-cycle primes; the bare trigger→storage→catastrophic-release skeleton is partly held already by cascade (propagation), tipping_points (abrupt change), and escape_and_leakage (boundary failure). But there is a genuine shared-abstract-mechanism residue that the named concept's metaphor obscures and that should be carried instead of it: the blockage-release dynamic — a barrier holding back an accumulating load and releasing it discontinuously when it fails, as distinct from steady leakage on one side or a continuous phase transition on the other. That dynamic really does recur across distinct substrates as co-instances, not metaphors: volcanic plugs and eruptive episodes, thrombus formation and embolism, suppressed-infection rebound when control measures relax, pent-up demand released after rationing. Where the cross-domain lesson is needed — a delayed, larger release stored behind an unstable hold, with an intervention window to bleed the load off gently before it bursts — the honest move is to reach for that substrate-neutral blockage-release pattern (filed as an emergent-candidate side capture), not the geomorphic concept, whose controlled-breach toolkit and outburst-flood vocabulary stay home in earth and environmental science. The mechanism that travels is the storage-and-discontinuous-release abstraction; the named landslide-dam concept and its geotechnical cargo do not (see Structural Core vs. Domain Accent).

Examples

Canonical

The Tangjiashan barrier lake is the textbook modern case. The May 2008 Wenchuan (Sichuan) earthquake shook loose a massive rock-and-earth avalanche that fell across the Jian River, forming an unengineered dam that impounded a lake growing toward and beyond 100 million cubic metres above a densely settled valley. The dam had no spillway and was made of loose, heterogeneous debris — the classic short, hard-to-forecast time-to-failure. Chinese engineers raced the accumulating lake against the degrading barrier: they cut a controlled sluice channel across the deposit to drain the lake in a managed release before it could overtop and burst, while evacuating roughly a quarter of a million people from the downstream flood path during the impoundment window. The controlled drawdown succeeded, and the potential outburst — far larger than the landslide itself — was averted.

Mapped back: The avalanche is the triggering mass-wasting event; the spillway-less debris pile is the unengineered barrier acting as the hazard's storage element, holding the impounded reservoir. The excavated sluice is the drawdown intervention window exploited during the accumulation-versus-degradation race, and clearing 250,000 people is action on the downstream outburst flood whose magnitude is set by the lake, not the slope.

Applied / In Practice

Himalayan glacial-lake outburst flood (GLOF) management runs the same skeleton with a moraine dam in the deposit slot, as a standing engineering programme rather than a single emergency. Imja Tsho, one of the fastest-growing glacial lakes in Nepal's Everest region, had swelled behind an unstable end-moraine until it threatened Sherpa villages far down-valley. Rather than wait for the moraine to fail catastrophically, an internationally funded project led by the Nepal Army in 2016 engineered a controlled outlet and lowered the lake level by roughly three metres, reducing the impounded volume, alongside downstream early-warning sensors and community evacuation drills.

Mapped back: The retreating glacier feeds the lake as the triggering source; the end-moraine is the unengineered barrier whose seepage and slope define the failure mode in play. Lowering the lake is a deliberate drawdown intervention that bleeds the load off gently, and the warning sensors track the accumulation-versus-degradation race so that a downstream outburst flood — which would arrive last and largest in a different community — can be pre-empted rather than absorbed.

Structural Tensions

T1: Intervening on the barrier versus waiting on it (the storage element makes both moves dangerous). Because the barrier is the hazard's storage element, every option carries the same load. Remove it fast and you release the impounded reservoir now; leave it and it releases anyway, on a schedule the geometry constrains but cannot pin down. Controlled drawdown is the prescribed thread between these, but the thread is thin: excavating a sluice across an already-unstable heterogeneous deposit can itself weaken it and precipitate the very failure the operation is racing to forestall. The intervention is not a safe third path outside the paradox — it operates on the same unstable structure, and its safety depends on the barrier holding long enough and firmly enough to be cut. The tension is that action, inaction, and the "controlled" middle all act through a storage element whose destabilization is the catastrophe. Diagnostic: Is the barrier stable enough that cutting a drawdown channel bleeds the load off gently, or does the excavation itself risk tipping a deposit already near failure?

T2: The visible landslide versus the larger flood that has not happened (where response attention goes). The concept's governing inference is that the worst harm is not the slope failure in front of the responders but the outburst that arrives last, largest, downstream, and often in a different community — so the threat must be sized from the lake, not the slope. That is correct and counter to instinct. The trouble is that the correct target is invisible on every axis that drives response: the immediate landslide has vivid victims and a clear scene, while the dominant hazard is a flood that has not occurred, in a place not yet harmed, on a schedule no one can name. Committing evacuation, exclusion zoning, and engineering effort to a downstream community for an event that has not happened, over the salient upstream disaster, cuts against the political, attentional, and resource pulls of the moment. The tension is that the largest consequence is the least legible one. Diagnostic: Is the response sizing and staging the downstream outburst from the impounded volume, or has the vivid, already-realized landslide captured the attention and resources the not-yet flood requires?

T3: A narrow window that demands precise timing versus a failure time that is a distribution, not a date. The short, heavily right-skewed time-to-failure is what makes the drawdown window narrow and the urgency real. But the same distribution denies the precision the response needs: a substantial fraction fail within days to months, yet no single date can be forecast, so both the drawdown operation and the downstream evacuation must be timed against an interval whose end is fundamentally uncertain. Hold the population clear too long and you sustain a quarter-million-person displacement, with its compliance erosion and economic cost, against a barrier that may drain or hold; act too late and the outburst arrives before the clearance completes. Monitoring the accumulation-versus-degradation race narrows the uncertainty but cannot eliminate it, because the release is set by a stochastic barrier physics, not a countdown. The tension is that precise timing is required and precise prediction is unavailable. Diagnostic: Is the evacuation-and-drawdown schedule built around the observable state of the barrier and lake, or around a false confidence in when the failure will actually come?

T4: The crisp storage-and-release regime versus a barrier that passes through the other regimes. The concept draws sharp boundaries — storage-and-release, not a continuous phase transition, and not steady leakage — because the correct intervention (bleed the load off; eliminate the storage) depends on which regime governs. But a real barrier does not stay in one box: a piping failure begins as benign-looking seepage through the poorly-sorted debris and accelerates into catastrophic release, so the reassuring steady leak can be the opening phase of the discontinuous burst the taxonomy files as a different regime. Reading the current seepage as "just a slow leak, a stable regime" is precisely how the run-up to piping is misdiagnosed. The tension is that the taxonomy's clean separation of steady escape from catastrophic release is what makes the reasoning tractable, while the physical barrier can migrate between those regimes and does its most dangerous work at the transition. Diagnostic: Is the observed seepage a stable steady-leak regime, or the accelerating early phase of a piping failure the storage-and-release framing warns is about to become discontinuous?

T5: Autonomy versus reduction (a geomorphic hazard or the blockage-release dynamic it instantiates). Landslide-dam failure is a fully specified natural-hazards concept, and within its physical family — earthquake, rainfall, volcanic, glacial, and ice-triggered barriers, GLOFs included — it transfers as mechanism, one process sampled across trigger types with the controlled-breach toolkit intact. But its operative vocabulary (spillway, seepage, lake-level monitoring, geotechnical stability) presupposes physical accumulation behind a physical barrier, so lifted whole beyond the substrate it travels only by analogy. What genuinely carries is a substrate-neutral residue the geomorphic name obscures: the blockage-release dynamic — an accumulating load held behind an unstable barrier and released discontinuously when it fails, distinct from steady leakage and from a continuous phase change — which recurs as co-instances in volcanic plugs, embolism, and pent-up demand, and sits near the parents cascade, tipping_points, escape_and_leakage, and pressure_release. The tension is between a concept whose controlled-drawdown machinery is real and reused within earth science and the recognition that its cross-domain lesson belongs to that blockage-release abstraction, not to the dam. Diagnostic: Resolve toward the blockage-release dynamic and its parents when the "barrier" is not physical material behind a geotechnical hold; toward landslide-dam failure when reasoning about an actual impounded reservoir behind an unengineered deposit.

Structural–Framed Character

Landslide-dam failure sits toward the structural end of the spectrum — best read as mixed-structural, closely analogous to isostasy: a genuine natural mechanism that runs observer-free, wearing irreducibly geotechnical vocabulary, with only a mild human-hazard-management overlay to distinguish it. Four of the five criteria point structural; the fifth, and a faint tint on the first, hold it off the pole.

Evaluative weight is low and mostly structural, with a small framed tint. The two-stage cascade — trigger, deposit, impoundment, failure mode, outburst — is a describable physical process that praises and blames nothing, and the storage-and-release physics is evaluatively neutral. The faint framed pull is that the concept is organized as a hazard: "failure," "risk," and the whole intervention logic (drawdown windows, evacuation) index human stakes and management in a way isostasy's pure balance does not. But this is a framing around the mechanism, not a verdict inside it, so the criterion still reads mostly structural.

Human-practice-bound points structural at the level that matters: the phenomenon runs entirely observer-free. A landslide dams a valley, a lake fills, the barrier pipes or overtops, and an outburst flood descends whether or not any human is present to manage or even witness it — the Tangjiashan cascade would have completed with no engineers on scene. What is practice-bound is the response apparatus (controlled drawdown, exclusion zoning, monitoring), an overlay on the natural mechanism rather than constitutive of it. So the mechanism is not human-practice-bound; only its management is.

Institutional origin is structural: the cascade is a fact of geomorphology, not an artifact of a survey or agency — Costa and Schuster analyzed and catalogued it, they did not invent it, exactly as Airy and Dutton named a thing nature already does. Import-vs-recognize is also structural within range: across earthquake-, rainfall-, volcanic-, glacial-, and ice-triggered barriers the same lifecycle is recognized intact, one mechanism sampled across trigger types, not an analogy stretched.

What keeps it off the structural pole is vocab-travels, which it fails exactly as isostasy does. The operative vocabulary — spillway, seepage, piping, geotechnical stability, moraine, lake-level, outburst flood — is irreducibly earth-science and does not float free of the physical-barrier substrate; lifted whole to a financial or governance system it becomes analogy, renaming every component.

The portable structural skeleton is the blockage-release dynamic — an accumulating load held behind an unstable barrier and released discontinuously when it fails, distinct from steady leakage on one side and continuous phase change on the other. As the entry argues, that dynamic (sitting among cascade, tipping_points, escape_and_leakage, and pressure_release, and flagged as an emergent blockage-release pattern) is what landslide-dam failure instantiates in a geomorphic register, not what makes "landslide-dam failure" itself travel: the cross-domain reach belongs to that substrate-neutral abstraction, while the domain-accented specifics — the controlled-breach toolkit, the outburst-flood vocabulary, the geotechnical failure modes — stay home. Its character: a real, evaluatively-light, recognized-in-nature storage-and-discontinuous-release mechanism, structural in skeleton but pinned by geotechnical vocabulary (and a light hazard-management framing) to mixed-structural rather than a free-floating prime.

Structural Core vs. Domain Accent

This section decides why landslide-dam failure is a domain-specific abstraction and not a prime, and it carries the case for its domain-specificity — so it is worth being exact about which layer travels and which stays home.

What is skeletal (could lift toward a cross-domain prime). Strip the geomorphology and a thin relational structure survives: a load accumulates behind an unstable barrier that itself is the storage element, and the barrier releases the stored load discontinuously when it fails — a delayed, larger release, distinct from steady leakage on one side and continuous phase change on the other, with an intervention window in which the load can be bled off gently before it bursts. The portable pieces are abstract — an accumulating quantity, a barrier that both holds and constitutes the hazard, a race between accumulation and the barrier's degradation, and a discontinuous release on a short right-skewed time-to-failure distribution. This is the blockage-release dynamic, and it is genuinely substrate-portable: it recurs as co-instances (not metaphors) in volcanic plugs and eruptive episodes, thrombus and embolism, suppressed-infection rebound, and pent-up demand released after rationing. The catalog holds it near cascade (propagation), tipping_points (abrupt change), escape_and_leakage (boundary failure), and pressure_release (stored load discharged) — and, the entry flags, the storage-and-discontinuous-release residue is distinct enough that it is captured as an emergent blockage-release pattern rather than fully carried by any one of those. But this is the core landslide-dam failure shares, not what makes it distinctive.

What is domain-bound. Everything that makes the concept landslide-dam failure in particular is earth-science furniture, and none of it survives extraction. The trigger is a specific mass-wasting event — a landslide, rockfall, debris flow (or moraine/ice dam in the GLOF variant); the barrier is an unengineered cross-valley deposit — no spillway, poorly-sorted heterogeneous debris, geotechnically unstable; the impounded quantity is a literal lake, and the release is a downstream outburst flood. The failure modes are physical and read from barrier composition and geometry — overtopping, piping, slope collapse. The intervention menu is irreducibly geotechnical — controlled drawdown via an excavated spillway, downstream evacuation and exclusion zoning, lake-level and barrier-deformation monitoring by gauges, satellite altimetry, and time-lapse imagery. The empirical cases are worked as such — Tangjiashan's >100 million m³ behind a Wenchuan-earthquake avalanche with 250,000 evacuated, Imja Tsho's engineered moraine outlet, Mount St. Helens and Pinatubo lahar dams, Costa and Schuster's historical catalogue. The decisive test: remove the physical accumulation behind a physical barrier whose stability is a geotechnical property — take the "barrier" to be a regulatory hold or a financial constraint — and the spillway, seepage, and lake-level vocabulary loses its referents, and what remains is the bare blockage-release dynamic, no longer landslide-dam failure.

Why this does not clear the prime bar. A prime is a relational structure whose vocabulary travels and whose cross-domain transfer is recognition of the same mechanism, not analogy. Landslide-dam failure's transfer is bimodal. Within the physical-hazard family it travels intact as recognition, because every member genuinely has material accumulating behind an unstable barrier: the barrier-as-storage framing, the accumulation-versus-degradation race, the failure-mode triad, the short right-skewed time-to-failure, and the controlled-breach intervention menu carry without translation across earthquake-, rainfall-, volcanic-, glacial-, and ice-triggered barriers — one mechanism sampled across trigger types, not an analogy stretched. Beyond that substrate, lifted whole, the named concept travels only by analogy: its operative vocabulary presupposes physical accumulation behind a geotechnical hold, so invoking "landslide-dam failure" for a financial, public-health, or governance system borrows the dam-burst shape while dropping the storage element that gives the original its predictive structure. And when the bare structural lesson is wanted cross-domain — a delayed, larger release stored behind an unstable hold, with a window to bleed the load off before it bursts — it is already carried in more general form by the substrate-neutral blockage-release pattern and its neighbours: propagation by cascade, abrupt change by tipping_points, boundary failure by escape_and_leakage, stored-load discharge by pressure_release. The cross-domain reach belongs to that abstraction; "landslide-dam failure," as named, adds only the geomorphic commitments — the unengineered debris barrier, the outburst flood, the geotechnical failure modes, the controlled-drawdown toolkit — and those stay home in earth and environmental science.

Relationships to Other Abstractions

Local relationship map for Landslide-Dam FailureParents appear above the current abstraction, mutual partners to the right, and children below. Node labels state whether each abstraction is prime or domain-specific; colors identify relation types.Landslide-Dam FailureDOMAINDomain-specific abstraction: Mass Wasting — is part ofMass WastingDOMAINPrime abstraction: Stress and Rupture — is part ofStress andRupturePRIMEPrime abstraction: Blockage Release Dynamics — is a kind ofBlockageRelease DynamicsPRIMEPrime abstraction: Cascade — is a kind ofCascadePRIME

Current abstraction Landslide-Dam Failure Domain-specific

Parents (4) — more general patterns this builds on

  • Landslide-Dam Failure is a kind of Blockage Release Dynamics Prime

    Landslide-dam failure is the geomorphic specialization of blockage-release dynamics.

  • Landslide-Dam Failure is a kind of Cascade Prime

    Landslide-dam failure is a two-stage natural-hazard cascade.

  • Landslide-Dam Failure is part of Mass Wasting Domain-specific

    Landslide-dam failure contains the mass-wasting event that creates its cross-valley barrier.

  • Landslide-Dam Failure is part of Stress and Rupture Prime

    The impounded barrier stage contains load accumulation, threshold failure, and abrupt release.

Hierarchy paths (13) — routes to 8 parentless roots

Not to Be Confused With

  • Glacial-lake outburst flood (GLOF). The variant in which a moraine or ice dam fills the deposit slot — a lake impounded behind an unstable end-moraine or glacier tongue that later bursts. It runs the identical trigger → deposit → impoundment → failure → outburst lifecycle and shares the whole toolkit, so it is a sibling instance of the same concept, not a rival to distinguish away. Tell: is the barrier a moraine/ice dam fed by a retreating glacier (GLOF) or a mass-wasting debris deposit (the canonical landslide case)? Both are the same storage-and-release mechanism with different deposit material.

  • Engineered-dam failure (dam break). The collapse of a constructed dam — designed with a spillway, built of specified material, instrumented and regulated. Landslide-dam failure turns on the barrier being unengineered: no spillway, poorly-sorted heterogeneous debris, a short and hard-to-forecast time-to-failure. The absence of design is exactly what makes the barrier the uncontrolled hazard. Tell: was the barrier built and provisioned with a spillway (engineered failure), or emplaced by a landslide with no outlet at all (landslide-dam failure)?

  • Single-stage landslide or debris flow. The mass-wasting event by itself — a slope failure that buries or destroys without impounding a valley. This is only the trigger of the cascade; if no cross-valley barrier and lake form, there is no two-stage outburst hazard. Tell: did the deposit block a river and impound an upstream reservoir (landslide-dam failure), or simply run out without creating a standing lake (ordinary landslide)?

  • Ordinary (meteorological) flood. A river flood driven by rainfall or snowmelt filling a channel beyond capacity. The landslide-dam outburst is a dam-break flood — the sudden discontinuous release of a stored reservoir, typically larger than the rainfall that a catchment could deliver and set by impoundment volume, not by precipitation. Tell: is the flood a discharge of an accumulated impounded lake released by barrier failure (outburst), or a hydrometeorological rise from precipitation/snowmelt (ordinary flood)?

  • The blockage-release dynamic and its parents (cascade, tipping points, escape/leakage, pressure release). The substrate-neutral residue — an accumulating load held behind an unstable barrier and released discontinuously when it fails — that landslide-dam failure instantiates in a geomorphic register. It recurs as co-instances in volcanic plugs, embolism, and pent-up demand. Tell: if the "barrier" is not physical material behind a geotechnical hold — a regulatory, financial, or biological constraint — reach for the blockage-release pattern (near cascade, tipping_points, escape_and_leakage, pressure_release), not the landslide-dam concept. (Treated fully in an earlier section.)

Neighborhood in Abstraction Space

Landslide-Dam Failure sits in a sparse region of the domain-specific corpus (69th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Geologic Landforms & Crustal Deformation (12 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-07-12