Steep, rocky terrain should still be avoided in the mid to upper elevations on northwest to east facing slopes. While the odds of triggering one of these avalanches a week ago were perhaps 75%, now they may be 10%. What odds are acceptable when the stakes are so high? We interviewed Jenna Malone not long ago and she described it as Thinking in Bets (Why Forecasting is Poker and not Chess). Mark gives the state of the snowpack below. The profile is in Big Cottonwood but we found similar, if inconsistent results in the Provo mountains over the past week.

If you choose to ride slopes where this layer exists, stack the odds in your favor by choosing slopes with a clean runout zone free of trees and rocks that would cause trauma if you do trigger an avalanche.

The southwest winds will increasingly scour and erode any available snow and deposit it in the form of scattered and pockety soft and hard wind drifts in the mid and upper elevations. In some areas, these drifts may land on weak snow surfaces formed over the past week and may be surprisingly sensitive. These areas are more likely sun and wind sheltered terrain where the weak snow surfaces are not as susceptible to decay. I wouldn't be surprised to hear that some are triggered at a distance today or tomorrow.

Drifts are often smooth, scalloped, and rounded and are often formed to the lee of terrain features such as ridgelines, sub-ridges, and rocky outcroppings. Soft wind slabs you're "in the snow" with your sled or skis; hard wind slabs you're "on top of the snow". Hard slabs are often trickier because they allow you to get well on to them before the collapse where the wind pillow/teardrop lens shape tapers down.

Cornice drops and ski cutting can be effective techniques in mitigating wind slab avalanches, but they are inherently dangerous and sometimes provide misleading results. A recent paper (with link to a video) on ski cutting by Bruce Jamieson and others can be found HERE. A take-home point - People can be injured and potentially killed while ski cutting.