The Avalanche Problem Types: A Field Guide to What's Trying to Slide

01Why the bulletin talks about 'problems', not just a number

The avalanche problem types are the recurring ways a snowpack fails, and every modern bulletin is built on them. The core set is: storm slab (new-snow), wind slab, persistent slab (and its worst form, deep persistent slab), wet snow (wet-loose and wet slab), loose dry snow (sluff), cornice fall, and glide. Europe's EAWS groups these into five families; North America names nine. Each has its own fingerprint — where it lives, what warns you, how it behaves, and how you manage it. This guide walks through all of them. At the top level, every one of them is either a slab (a cohesive layer that fractures and releases all at once) or a loose-snow release (a single point that fans out as it runs) — a distinction worth holding in mind as you read.

Open any modern avalanche bulletin in the Alps — France, Switzerland, Austria, Italy — and you will see two things side by side: a danger level from 1 (Low) to 5 (Very High), and one or more avalanche problems (Fr. problème avalancheux, De. Lawinenprobleme, It. problemi valanghivi). The number gets the headlines. The problem is what actually keeps you alive.

Here is the trap. A danger level is a rating of overall hazard — it blends how likely a slide is, how big it could be, and how widespread the trouble is. But two days both rated 3 — Considerable (the middle rung of the EAWS 1–5 danger scale, and the level at which most avalanche accidents happen — see how to read an avalanche bulletin) can demand completely opposite behaviour. On one, the issue is fresh storm snow that will stabilise by tomorrow afternoon; you wait, then ski almost anything. On the other, the issue is a persistent weak layer buried two weeks ago that you can trigger from 50 metres away on a slope that looks bomber. Same number. The first rewards patience; the second rewards staying off entire categories of terrain for weeks.

That is the whole point of the framework. The avalanche problem tells you what is trying to slide, where it lives, and how it behaves — and behaviour is what you manage. The response is non-linear and problem-specific: you cannot simply ski 'one notch more carefully' as the number rises. You have to ask a different first question.

The danger level sizes the threat. The problem type tells you how to think. Always read both — and let the problem drive your terrain choices.

This is exactly why Snow Trace surfaces the official avalanche bulletin on the map: so you can read the active problem and the snowpack summary beside your route, not just the headline digit. (The rating is issued by your regional avalanche service and surfaced here, not forecast by us.)

The European Avalanche Warning Services (EAWS) define a harmonised set of problems used across the Alps; North America (CAA / AAA, Avalanche Canada, the US centres) uses an almost identical taxonomy but splits a few of them out separately. We start with the five core problems that map cleanly between both systems, then cover the extra problems North America names on their own — deep persistent slab, loose snow and cornice fall:

1. New snow / storm slab — fresh snow overloading a weak interface.
2. Wind slab (plaque à vent, Triebschnee) — wind-drifted snow on lee slopes.
3. Persistent weak layer (couche fragile persistante) — buried surface hoar, facets or depth hoar.
4. Wet snow — meltwater destroying bonds.
5. Gliding snow — the whole pack creeping on smooth ground.

For each, we give: what it is, where and when it forms, the field signs, how it behaves (trigger sensitivity, spatial predictability, persistence, typical size), and how to manage it. Use the interactive 'fingerprint' tool to compare them at a glance, and read each problem against the Aspect & Elevation rose and your Slope Angle habits.

Park the time-probe at day 2 and flip between problems: storm slab has already fallen to ~24% and reads "Healing" — the mountain is mending it for you — while the persistent weak layer is still ~69% and reads "Still armed", essentially where it started. Two days after each was born, one is nearly a wait-and-ski problem and the other has not budged. Persistence is not a footnote; it is a shape you can feel by holding the clock still and switching problems.

021. New snow / storm slab — the overloaded interface

What it is. During and just after a snowfall, a slab of fresh, cohesive snow loads the surface faster than it can bond to whatever lies beneath — an old crust, a sun-baked surface, or a weak interface within the new snow itself. When the load wins, you get a storm slab. This is the most familiar avalanche problem and the one most directly tied to a single, visible driver: it is snowing, and it is snowing a lot.

Where and when it forms. All aspects, all elevations — anywhere snow accumulates. The intensity matters more than the total: 30 cm in three hours is far more dangerous than 30 cm over two days, because the snowpack has no time to adjust. And what a given depth actually weighs depends on how dense the new snow is — its snow-to-liquid ratio: a wet, heavy 30 cm loads the slope far harder than dry powder. Sensitivity peaks during the storm and for the first 24–48 hours after it ends, then fades comparatively fast as the new snow settles, sinters and gains strength.

Field signs.

• Active, ongoing snowfall, especially heavy rates (>2–3 cm/hour).
• Recent natural avalanches on similar slopes — the single most important sign.
• Whumpfing and shooting cracks as you move.
• A soft, cohesive upper slab that you can pick up as a block.
• Poor bonding to an old surface in a hand pit or quick column test.

How it behaves. Trigger sensitivity is high during the active window — skiers and natural triggers both work easily. Spatial predictability is comparatively good: the problem is widespread but consistent, so a representative slope tells you a lot about its neighbours. Persistence is low — this is a fast-healing problem, often gone within a day or two of clearing weather. Size scales with the depth of new snow.

How to manage it.

1. Wait it out. Patience is the cheapest tool you own. Give a big storm 24–48 hours and re-assess.
2. Avoid steep, unsupported and freshly wind-affected terrain during and right after the storm.
3. Stick to lower-angle slopes (under ~30°) and mind run-outs and connected start zones above you.
4. Re-test bonding to the old surface before committing to anything steep.

Storm slab is the problem that rewards doing nothing. The mountain heals it for you — if you give it time.

032. Wind slab — snow moved by the wind, not the sky

What it is. Wind can transport snow many times more efficiently than snowfall can deliver it. As it does, it shatters and packs the crystals into a dense, cohesive layer that drapes over lee and cross-loaded slopes. The result — plaque à vent, Triebschnee — is a wind slab: a slab sitting on a weaker layer beneath, often with no fresh snow falling at all. You can build a serious wind-slab problem under a blue sky.

Where and when it forms. On the lee side of ridges, in gullies, behind terrain features and on cross-loaded slopes where wind blows across rather than over. It is most pronounced at and above treeline, where wind has free rein, and it forms wherever the wind has recently been strong enough to move snow — which can be daily in the high Alps.

Field signs.

• Pillows, drifts and smooth, rounded loading on the lee side of features.
• A surface that feels dense, 'chalky' or hollow — a drum-like whumpf underfoot.
• Cornices, sastrugi and visible snow plumes off ridgelines.
• Stiff slab over softer snow; shooting cracks that radiate from your skis.
• Recently stripped, scoured windward slopes paired with loaded lee bowls.

How it behaves. Trigger sensitivity is high, especially while the slab is fresh. The defining trait is localised, patchy distribution: wind slabs hide in specific pockets, so spatial predictability is moderate-to-poor — the loaded gully is deadly while the ridge two metres away is bare. Persistence is low-to-moderate: fresh wind slab can be reactive for days, particularly when it sits on a weak interface. Size is usually small-to-medium but enough to bury or carry you into worse terrain.

How to manage it.

1. Read the wind. Know the recent wind direction, then deliberately picture which slopes were loaded and which were scoured.
2. Avoid fresh, smooth, convex lee loading and the immediate area under cornices.
3. Approach suspect slopes from windward or via supported ground; test the slab before committing.
4. Because the problem is pocketed, one good test is not a green light for the next slope — re-evaluate at every aspect and feature change against the Aspect & Elevation rose.

Wind slab is a problem of where, not whether. The skill is spotting the loaded pocket — and walking around it.

043. Persistent weak layer — the buried memory of bad weather

What it is. Sometimes a fragile layer forms at the snow surface — feathery surface hoar, sugary faceted crystals, or, near the ground, coarse depth hoar. Then it gets buried by later snow and preserved, sometimes for weeks. A slab builds on top while the weak layer below refuses to bond. This is the persistent weak layer (couche fragile persistante) — and it is responsible for the majority of avalanche fatalities involving skiers. EAWS labels it the old snow problem (Ger. Altschneeproblem), the exact term the Alpine bulletins you read will use.

Where and when it forms. Weak layers love a cold, clear, settled spell — high pressure that lets the surface metamorphose into facets or grow surface hoar before the next storm buries it. They persist longest on cold, shaded aspects (roughly N–E) and, for depth hoar, on early-season snow over higher, colder terrain. The problem can outlive several storm cycles.

Field signs. This is the dangerous part: the field signs are unreliable and intermittent. A clean-looking slope can hide a primed weak layer.

• Whumpfing and remote triggering — slides released from a distance or from flat ground.
• Recent persistent-slab avalanches, often stepping down to a deep, smooth, faceted bed surface.
• Column tests that produce sudden, clean (planar) shears on an identifiable buried layer.
• A documented weak layer in the bulletin's recent history — read the snowpack summary, not just the number.

How it behaves. Trigger sensitivity is moderate but treacherous — the slab can feel solid until the one trigger point connects, often the thin spot at a rock or convexity. The two defining traits are extreme: spatial predictability is the worst of any problem (a single safe slope tells you almost nothing about the next), and persistence is the longest — measured in weeks, not hours. Avalanches tend to be large, breaking wide and deep, frequently with no warning underfoot.

How to manage it.

1. Respect time, not signs. You cannot reliably test your way onto a persistent-slab slope. The honest answer is usually terrain avoidance.
2. Stay off and out from under steep slopes on the suspect aspect/elevation band until the bulletin confirms healing — which can take a long, frustrating while.
3. Treat 'it skied fine yesterday' as meaningless. Low spatial predictability means yesterday's safe slope is not today's.
4. Build your slab-and-weak-layer mental model with Understanding Snowpack Stability, and use the Aspect & Elevation rose to define exactly which slopes to avoid.

Persistent weak layers are the problem you cannot outsmart with cleverness — only with discipline. When one is active, the win is the run you chose not to ski.

054. Wet snow — when meltwater takes the bonds apart

What it is. Add liquid water — from sun, warm air or rain — and the bonds between snow grains weaken. Wet-loose avalanches start from a point and fan out as the surface loses cohesion; wet slabs release a cohesive block when water reaches and lubricates a buried layer. Both are driven by warming and water input, which makes them, unlike most dry problems, partly predictable by the clock.

Where and when it forms. Classic spring conditions: sunny aspects (S, SE, SW) in the afternoon, during heat waves, with the first rain-on-snow events, or whenever an overnight refreeze fails to set up. The risk follows the sun around the compass through the day and climbs with elevation as the warm air rises. This is the world of the melt-freeze cycle — see Spring Corn & the Melt-Freeze Cycle.

Field signs.

• Roller balls / pinwheels and the first natural wet-loose sluffs on sun-warmed slopes.
• Snow that no longer supports you — boot or ski punching through a failing crust.
• A weak or absent overnight refreeze; isothermal, soggy snow at depth.
• Water visibly running, dripping cliffs, slush underfoot by late morning.

How it behaves. Trigger sensitivity rises through the day as water content increases — moderate, but strongly time-dependent. Spatial predictability is moderate: it tracks the sun and aspect well, but a saturated wet slab can surprise you in scale. Persistence is short within a day (a good overnight refreeze resets it) but the pattern recurs every afternoon of a warm spell. Wet-loose slides are usually small-to-moderate; wet slabs can be large and destructive.

How to manage it.

1. Ski with the clock. Start early on a firm refreeze, finish before the surface goes to mush. Be off consequential slopes by the time roller balls appear.
2. Follow the sun: move to shaded or not-yet-warmed aspects as the day heats up.
3. Treat a poor or missing overnight refreeze as a hard stop signal for sun-exposed terrain.
4. Mind terrain traps and run-outs — even a small wet slide carries enormous mass.

Wet snow is the rare problem that runs on a timetable. Your defence is the alarm clock, not the snowpit.

065. Gliding snow — the whole pack on the move

What it is. On smooth ground — grassy slopes, rock slabs, smooth bedrock — the entire snowpack can slowly slide downhill as a unit, lubricated by a film of meltwater at the base. Tension cracks open across the slope: glide cracks, mouth-like fractures that gape over days. When a glide crack finally lets go, the whole pack releases in a glide avalanche. The terrifying signature: there is no reliable way to predict the timing.

Where and when it forms. Over smooth, anchoring-free ground at any aspect, wherever liquid water reaches the base — common on grassy alpine slopes and rock slabs. It appears both in spring melt and, increasingly, mid-winter when a warm, deep snowpack stays wet at the ground. The aspect rules of other problems barely apply: glide follows the ground surface, not the sky.

Field signs.

• Open glide cracks — the only real warning, and an ambiguous one.
• Snow visibly creeping or bulging above the crack.
• Known glide-prone terrain (smooth grass, slabby rock) carrying a deep, wet pack.

How it behaves. This problem is the strange one. Trigger sensitivity to people is very low — you almost never trigger a glide avalanche yourself; it goes when it goes. But spatial predictability is as poor as it gets and timing is effectively random: a glide crack can sit open for weeks, then release at 3 a.m. or 3 p.m., loaded or unloaded. Persistence is long — the crack is a standing threat for as long as it exists. Glide releases involve the full depth of the pack and can be very large.

How to manage it.

1. Avoid, do not predict. This is the one problem where the official advice is to not try to forecast it at all.
2. Do not travel below, above or through an open glide crack — minimise time spent in its run-out.
3. Route-find around glide-prone ground entirely when cracks are present.
4. Accept that a clean-looking crack proves nothing about when; treat the whole feature as armed.

Glide is the humbling problem. It does not care about your snowpit, your wind read, or your clock. The only safe move is to be somewhere else.

07The 9 North American problems: deep persistent slab, cornice fall & loose snow

The five families above cover what is trying to slide almost anywhere on Earth. But if you read a North American forecast (Avalanche Canada, the US centres via Avalanche.org), you will see nine avalanche problems, not five. North America simply splits a few of the families into named problems of their own, because each behaves distinctly enough to manage differently. Here is what those extra names mean — and where they fit the framework above.

Deep persistent slab. This is the persistent weak layer at its worst: the weak layer sits near the ground (usually basal depth hoar or old facets), often buried by a metre or more of snow. It is the same beast as a persistent slab, but deeper, less reactive to a light skier, and far more destructive when it goes — a full-depth, terrain-spanning break that can be triggered remotely from low-angle terrain or from a thin spot at a rock. It has the longest memory of any problem, sometimes staying primed for the entire season — a direct consequence of the buried snowpack structure rather than the day's weather. Management is pure discipline: identify the suspect aspect/elevation band and stay off and out from under it, for weeks. When one is present, 'it feels supportive' is not evidence of anything.

Loose dry snow (sluff). A point-release of dry, unconsolidated snow that starts small and fans out as it entrains more snow — the dry cousin of the wet-loose slides in the wet-snow section. On its own a sluff is usually small, but it matters for two reasons: it can knock you off your feet in steep, exposed terrain (couloirs, above cliffs and terrain traps), and it is a visible clue that the surface snow is soft and unbonded. Manage it by controlling your sluff on steep lines (ski off to the side, let it run, don't ski on top of it) and by respecting what is below you — a small slide over a big consequence is still a big problem.

Loose wet snow. The wet, point-release counterpart, driven by meltwater rather than fresh cold snow. It is already folded into the wet-snow problem above (roller balls, pinwheels, the first natural sluffs on sun-warmed slopes) and follows the same clock-and-refreeze management. North America just gives it its own line on the forecast.

Cornice fall. A cornice is the overhanging mass of hard, wind-built snow that curls over the lee edge of a ridge — the same wind that builds a wind slab builds the cornice above it. Cornices fail with warming, new loading or simply their own growth, often with no trigger at all, and a falling cornice both threatens anyone below and can act as a heavy trigger that sets off the slope beneath it. Management is spatial and simple: give cornices a wide margin, never walk out onto a corniced ridge (the fracture line is set back from the visible edge), and do not linger on slopes directly beneath one, especially as the day warms.

So the taxonomies agree more than they differ. Europe folds sluff and loose-wet into new-snow and wet-snow, treats cornice fall as part of the wind picture, and reads deep persistent slabs as the dangerous end of the persistent-slab problem. North America pulls them out by name. Either way, the management questions are the same — which is exactly the point of the framework.

08Comparing the avalanche problems: same number, five different days

Lay the five problems side by side and the logic of the framework becomes obvious. They differ on the behaviours that actually drive decisions, not on how scary they sound.

(Typical size is the destructive scale, D1–D5: roughly D1 is a harmless sluff, D2 could bury or injure a person, and D3+ can destroy a car, a building or a stretch of forest.)

Now re-read that 'Considerable' day. If the active problem is storm slab, your first move is wait, then ski mellow. If it is wind slab, your first move is find and avoid the loaded pockets. If it is a persistent weak layer, your first move is stay off whole aspect/elevation bands for as long as it takes — patience measured in weeks. If it is wet snow, your first move is check the watch and the refreeze. If it is glide, your first move is route around the cracks and stop trying to predict. Five different first questions. One danger number.

This is why the response is non-linear. You do not get safer by uniformly 'skiing more carefully' as the rating rises. You get safer by identifying which problem is in play and switching to its specific playbook. Tie every one of these back to terrain: the Aspect & Elevation rose tells you which slopes each problem lives on, and disciplined Slope Angle management — staying under ~30° when in doubt — is the single most reliable defence across all five. Bake those aspect and angle limits into the route itself when you plan your line and turnaround time, so the safe choice is settled at the kitchen table rather than on the slope. When you cannot identify the problem, treat the day as if the worst plausible one is active.

One practical aid worth its weight: the most important field sign for several problems is evidence of recent avalanches and conditions on your actual line — and community trip reports on Snow Trace often flag exactly that, alongside recent snow-station totals that hint at loading. Treat them as dated, anecdotal observations to corroborate the bulletin, never as a substitute for it.

The expert move is not knowing more facts. It is asking the right first question — and the avalanche problem is what tells you which question that is.

Notice how persistent weak layers and glide max out spatial unpredictability and persistence — the two traits you cannot test your way around — while storm slab, which feels scariest mid-storm, actually scores lowest on persistence. The bars that matter for decisions are unpredictability and persistence, not raw trigger sensitivity.

09Educational use only — and how to actually use this

This article is educational only. It is not avalanche forecasting, not a substitute for formal training, and not a basis for go/no-go decisions. The fingerprint scores in the interactive are illustrative teaching values meant to show how the five problems differ in character — they are not measurements and they are not bulletin data.

Before you travel in avalanche terrain:

1. Read your local bulletin every day (Météo-France / Meteomont / SLF / lawinen.report and your national service). Read the problems and the snowpack summary, not only the danger number — see how to read an avalanche bulletin.
2. Take a recognised avalanche course** and carry — and know how to use — a transceiver, shovel and probe. The best gear is useless without practice and good decisions that keep you out of the slide in the first place (see why the first 15 minutes decide everything).
3. Travel with judgement, not articles. Conditions change by the hour and by the slope. Nothing here replaces local knowledge, current observations and a trained partner.

Plan the day in one place, then trust your eyes on the snow. On Snow Trace you can read the official bulletin on the map next to your route, check recent snow-station data and community trip reports, and pre-plan aspect and slope angle before you leave the trailhead. It helps you read the mountain and identify the active problem — it does not make the decision for you. The mountains are indifferent; humility is the only edge that scales. Ski long, ski curious, and come home.

10Sources & further reading

Sources & further reading. This article reflects the consensus of the major avalanche-safety organisations and the standard references. Always defer to your local daily avalanche bulletin and hands-on training over any single article:

• **EAWS** — European Avalanche Warning Services: the standard danger scale and avalanche problems
• **Avalanche.org** — US National Avalanche Center: forecasts and free avalanche education
• **SLF** — Swiss Institute for Snow and Avalanche Research
• McClung & Schaerer, "The Avalanche Handbook" — the standard technical reference