Snow Avalanches: Slab Failure, Fracture Propagation and the 30-45 Degree Danger Zone

A snowpack is a stack of unequal layers, not a block

Snow falls in discrete storms, and each storm can bond to the layer beneath it well or poorly depending on temperature, wind and how much time passed before the next snowfall buried it. The result is a snowpack that is really a stack of distinct layers with very different mechanical properties -- some strong and cohesive, some weak and crumbly -- and the most dangerous kind of avalanche happens when a strong, cohesive slab sits directly on top of a persistently weak layer that cannot support it.

Forecasters and researchers formalise the question 'will this slope avalanche' as a stability ratio: the weak layer's shear strength divided by the shear stress the overlying slab's weight actually places on it. A ratio comfortably above 1 means the layer can support the slab; as new snow, wind loading, or a skier's added weight push the ratio down toward 1, the slope moves from stable toward marginal to genuinely dangerous.

stability ratio = tau_strength / tau_stress
tau_stress increases with: new snowfall, wind-loading, added weight (a skier, a cornice fall)
ratio ≤ 1  →  weak layer cannot support the slab above it -> failure

The fracture that outruns the trigger

A slab avalanche does not begin as a slow slide -- it begins as a fracture, often triggered at a single, localised weak point (a skier crossing a thin spot in the weak layer, a cornice collapsing onto the slope), that then propagates sideways through the weak layer at speeds that can reach tens of metres per second. Only once the fracture has spread far enough that a large enough slab has lost its support does the whole slab begin sliding as a coherent block, releasing along the fracture and often along flanking and downslope fractures as well -- which is why an avalanche can release well above or to the side of the actual trigger point, and why the crown, flank and bed-surface fracture lines of a released slab are so clearly visible afterward.

Why 30 to 45 degrees is the danger zone

Slope angle governs avalanche risk in a distinctly non-monotonic way. Below roughly 25 to 30 degrees, gravity's component along the slope is usually too small to overcome friction and cohesion even in a weak snowpack, so slab avalanches are rare. Above about 45 to 50 degrees, snow tends to slough off in smaller amounts continuously as it accumulates, rather than building up into the large, cohesive slab needed for a major release. The 30 to 45 degree band sits in the sweet spot where slopes are steep enough for gravity's shear stress on a weak layer to be significant, yet gentle enough for a substantial slab to accumulate and hold together before it fails -- which is exactly why avalanche forecasters and backcountry travellers watch that angle range most closely, and why slope-angle shading is a standard layer on avalanche terrain maps.

Wind, aspect and why yesterday's storm matters more than today's

Wind does not just add snow, it redistributes it, scouring windward slopes and depositing dense, cohesive wind slabs on the lee side -- often the single most dangerous combination in avalanche terrain, since wind-loaded slopes both accumulate a strong slab quickly and are frequently the very slopes most attractive for backcountry travel (sheltered, well-filled-in terrain). Aspect (which compass direction a slope faces) also matters for how weak layers form in the first place: shaded, north-facing slopes (in the northern hemisphere) stay colder and support the growth of large, weak, faceted snow crystals or surface hoar for much longer than sun-exposed slopes, where daily melt-freeze cycles tend to bond layers together instead. This is why a persistent weak layer buried weeks ago by an earlier storm, not the snow that fell yesterday, is very often the layer forecasters are most worried about -- it has had time to become both weaker and more widespread.

Frequently asked questions

Try it live

Everything above runs in your browser — open Snow Avalanche and change the parameters while it is running. Nothing is installed, nothing is uploaded, the whole model lives in one tab.