At the White Pine trailhead in Little Cottonwood Canyon, at ten of seven on a January morning, a man in a headlamp thumbs open the day's avalanche forecast, sees the word CONSIDERABLE glowing orange, decides that's just the number they always give, and starts walking. He is wrong on both counts, and the proof is printed four inches below the color he already swiped away. The forecast is a document a professional got up at four to write, specifically so this man would not have to learn its contents with his body. He read the headline and pocketed the letter.

A North American avalanche forecast is one of the densest free documents in public life, and most of the people it is written for read it the way they read a nutrition label — one number, one color, a decision already made on the drive up. The color is the least useful thing on the page. Learn to read the rest of it the way it was built — from the problem up, not the rating down — and the same screen that looked like a weather app starts naming, with unsettling specificity, which slopes will try to kill you today: on what aspects, at what elevation, at what hour, and how sure the author is about all of it.

The Number Is Not the Message

The danger scale runs one to five — Low, Moderate, Considerable, High, Extreme — green through black, and it feels like the whole story because it's the only part in color. It is a headline. Here is the fact that should reorganize how you read it: the great majority of avalanche fatalities happen at the two middle ratings, Moderate and Considerable, and almost none at Extreme. On an Extreme day everyone stays in the parking lot. On a Considerable day the snow is often superb and the slopes still look entirely skiable, and competent, experienced, well-equipped people die on them, because Considerable means human-triggered avalanches are likely and the danger is real and deniable at the same time. The rating tells you how loud to be careful. It does not tell you what to do. For that you have to keep reading, which is precisely where most people stop.

The Rose

Below the number sits an octagon most people register as decoration. It is the most useful graphic on the page. The eight wedges are the eight compass aspects — north around to northwest — and the octagon is stacked in three tiers for the three elevation bands: below treeline, near treeline, above treeline. Each of those twenty-four little zones is shaded to its own danger. Your day has aspects and elevations too. That couloir faces northeast and tops out above the trees; that mellow apron is a south-facing, below-treeline roll. Lay your actual plan over the rose and the regional forecast becomes a forecast about your specific slope. If your line lands in the orange, the document on your phone stopped talking about the mountains in general some time ago. It is talking about you.

The rating is the volume. The rose is the address.

The Problem Is the Instruction

Under the rose, the forecast names the day's avalanche problem — sometimes two or three — and this is the part that reads like instructions, because it is. Each problem comes with a location (which aspects, which elevations), a likelihood (from unlikely to almost certain) and a size (a D2 is big enough to bury, injure, or kill you; a D3 will destroy a car). More importantly, each problem is a different animal that wants to be handled a different way. Mistaking one for another is how a good decision quietly becomes the wrong one.

  • Wind slab: fresh snow drifted onto the downwind side of ridges, near and above treeline. Predictable, localized, often stiff and hollow underfoot — the most findable and the most walk-around-able of the bunch.
  • Storm slab: new snow that hasn't yet bonded to the old surface beneath it. Touchy during and right after the storm; frequently heals itself with a day or two of patience.
  • Persistent slab: a buried weak layer — sugary facets, feathery surface hoar — that refuses to heal on the storm's schedule. Low odds, high consequence, miserable to test. This is the one that gets experts.
  • Wet loose and wet slab: spring's department, governed by the thermometer and the clock. Roller-balls and pinwheels are the snowpack telling you what time it is.
  • Cornice: the overhanging drift curling off the ridgeline. It breaks back farther than you'd guess, weighs tons, and tends to take the slope beneath it along for the ride.

Which of these you'll meet depends on where you are, and a good reader knows the local dialect. The maritime snowpack of the Cascades and the Sierra is deep, dense, and quick to forgive; it mostly hands you storm and wet problems you can wait out over coffee. The thin, cold, continental snowpack of the Colorado San Juans spends half the winter rotting its own base into sugar and hands you deep persistent slabs that stay armed and lethal for months after the last flake fell. The intermountain ranges in between — the Wasatch, the Tetons — collect a little of everything, and hoard the worst of the persistent stuff on their coldest, shadiest, most tempting aspects. The forecast is regional because the snow is. Learn your own snowpack's favorite way to hurt you, and half the document is read before you open it.

The Sentence Under the Number

Then there is the prose — the bottom line up top, the longer discussion below it — and it rewards being read for tone as much as content. Forecasters write in a deliberate, level register, and when that register shifts you should feel it like a change in cabin pressure. A paragraph that starts repeating itself, a weak layer named for the fifth morning running, a word like stubborn or unusual or reactive slipping in: that is a professional raising their voice inside the only volume they're permitted. Read the trend — is the danger climbing with the afternoon sun, easing as the storm snow settles, or holding? Read the timing, because a forecast can honestly say Low at dawn and mean it, and mean High by two o'clock, and that whole arc lives only in the sentence. And read the confidence, which forecasters state plainly and which almost nobody weighs. "We have low confidence in the persistent slab problem" is not throat-clearing. It means the snowpack is lying to everyone, including the people paid to interrogate it, and that uncertainty is its own hazard, owed the same respect as a number.

"Low confidence" isn't the forecaster hedging. It's the forecaster telling you the snow is lying to everyone, them included.

Back at the White Pine trailhead, the better version of that morning costs ninety more seconds. You read past the color to the rose, and there's the orange, sitting on exactly the northeast, above-treeline bowl you drove up here to ski. You read the problem, and it's a persistent slab — the patient one, quietly lethal under good skiing for nine days now. You read the sentence, and the forecaster, four hours and two thousand vertical feet ahead of you in the dark, has already set down the word that matters: stubborn. Ninety seconds. Then you change the plan or you keep it, but you do it having read the whole letter — which is the entire, and only, thing the person who wrote it was up at four hoping you would do.