Avalanche awareness

Avalanche training Guidelines

If you take an avalanche course, make sure it follows the guidelines of the American Avalanche Association (AAA), which requires at least 24 hours of instruction, 1/2 of which is in the field. Requirements for Class I instruction are on the web site. Make sure you get an instructor who has:

In Tahoe, classes are offered by Donner Summit Avalanche Seminars


Avalanche hotline: US forest service in Truckee 530 587-2158 or online at http://www.fs.fed.us/r5/tahoe/currentconditions/

Snow Characteristics

Snow has several characteristics which make it a weak material:

  1. Snow looks like a single blanket, but it is made up of a series of layers. Every storm, and every cell within each storm, lays down a different layer. Horizontally driving snow is different from light snow that falls straight down. Avalanche likelihood is impacted by how well a layer of snow sticks to another layer of snow.
  2. Snow is inherently weak.
    • Snow consists mostly of air, held together by skeletal structures. The lightest champagne powder snow is 5-6% water. Sierra "powder" is 8-10% water. Sierra cement (or cascade concrete) is 12-15% water. Mixed rain/snow is 20% - 25% water.
    • Snow exists at or near its melting point, around 0° C, and it is therefore not a stable structure. 0° C is the melting point, not the freezing point. Water is capable of freezing at 0° C, but only if it has an object that it can stick to, like a dust particle, the side of a rock, or frozen ice. A droplet of water suspended in the air (a cloud) will not spontaneously freeze until it is -40° C. Between -40° C and 0° C, water is in a supercooled state. Ice that is created between -40° C and 0° C is called rime ice. Rime ice is formed when supercooled water floats in and comes in contact with an object.
    • Snow is infinitely variable in time and space. Snow will continue to change over time. Any material that is variable in thickness will break at its thinnest place. Terrain underneath snow can vary in height: Trees or rocks that stick up cause a thinner area of snow, which causes weakness. As a result, an avalanche often breaks from tree line to tree line.
  3. Snow behaves in ways similar to a visco-elastic material like silly putty. It flows and stretches if pully slowly, but it can also break if pulled quickly. Snow on the side of a slope can stretch from the top to the bottom under the force of gravity, and then eventually break. The snowpack is like an inflated balloon. A force spread out over a large area will cause even deformation of the snow, which can be absorbed by the snowpack. A small force, like a skier, applied to a small area, can "pop" the snowpack and cause an avalanche. Every time you turn, you put 3-8 times your body weight on the snow.

Types of avalanches

  1. Loose avalanche or point release avalanche: Starts at a point and becomes wider as it flows downhill, forming a triangle. It appears like a chain reaction, similar to sand falling down the side of a sand castle, and usually moves only the top 10cm of snow. Loose avalanches move slowly down the hill, and it is often possible to move out of the way. There are two types:
    • Loose avalanches caused by dry snow: These usually do not cause injuries.
    • Loose avalanches caused by very wet snow: They are somewhat more dangerous.
  2. Slab avalanche: More dangerous than the loose avalanche. It is easy to predict and locate. In a slab avalanche, a large slab from the snow layer releases and slides downhill. The cross section at the top that remains after the slab avalanche at the fracture line is called the crown surface. The newly exposed slope revealed after the avalanche is called the bed surface. People usually trigger the slab avalanches that kill them. These avalanches have lots of mass, momentum, and inertia. Avalanches can approach 150-200 mph. A loose avalanche can trigger a slab avalanche.
  3. Ice avalanche: resulting from falling ice flows. Ice avalanches are highly unpredictable, and do not correlate to snow, heat, weather, etc. It is best to avoid climbing on or under an ice flow. Not surprisingly, the most difficult segment of the Everest ascent is the Khumbu Icefall.
  4. Cornice avalanche: Cornices are very weak when they are first formed, because the dry snow is barely held together by static electricity that results when the polar H2O molecules rub against each other. Cornices then get stronger over time, and then get weaker again as they melt. Cornice avalanches can trigger slab avalanches.
  5. Roof avalanche: Created by snow on top of an artificial surface, with a heat source underneath. Roof avalanches are highly unpredictable, and can happen on any roof slope angle. Snow can melt at the surface from the underlying heat source, flow down the roof, and then freeze into solid ice at the eves where there is no heat. Roof avalanches have both snow and ice, and the falling ice can easily kill you, so it is best to stay away from under the eves.

Avalanche prediction

Avalanche forecasting is a mix of art and science. Leaning avalanche skills involves two things: Data measurement, and the human factor that influences how those measurements are interpreted. The data measuring aspect of avalanches consists of three main areas:

  1. Terrain. The slope angle is the most important. 96% of all avalanches occur on slope angles between 30° and 45°. 50% of avalanches occur between 35° and 40°. You can avoid avalanches if you can avoid being on or underneath slope angles between 30° to 45°. The difference in risk between 28° and 30° is huge.

    If you have crampons, you can climb on slope angles above 45° with less risk, because snow sluffs off very quickly for angles of greater than 45°; however, you may be susceptible to an ice avalanche.

    Measure the slope angle using a clinometer. A transit can also be used. Clinometers are usually built into compasses. Manufacturers include:
    • Suunto (M-3G and MC-2)
    • Brunton (Classic, Elite and 8040 compasses)
    • Silva (Ultra Ranger 530, Ranger 515CL and Ranger 515CLQ)
  2. Weather. Weather aspects include wind loading:
    • How fast the snow is falling in cm/hour
    • The density of snow that is falling in % of water
    • The length of time that the snow is falling
  3. Snowpack. Weak layers in the snowpack increase avalanche risk. Slopes at different elevations have different layer characteristics.

In addition to these three measurable quantities, there is a 4th element, and that is the human factor. There are several issues related to the human factor:

  1. The human factor which impacts how the measured quantities are interpreted. Humans are the only animal capable of going against logic, and one factor that can enter into avalanche prediction is rationalized expedience: A skier who wants or needs to traverse a terrain out of convenience may unwittingly skew his analysis to produce a prediction that is less risky.
  2. The human factor that occurs after prediction. After avalanche risk is interpreted correctly, a skier may knowingly enter an area of high avalanche risk.
  3. False asumptions. An example is assuming that someone else in the group is looking out for the safety of the group.
  4. Presumed Safety. Just because someone else recently skied down a slope safely does not mean that the next person will ski down safely. Avalanches can be triggered by the 10th, 50th, or 100th person to ski down a slope.

Avalanche prediction can be simplified by reducing the data set that is used, and limiting the data that you are gathering. Not all data is important. A skier can limit data to the bullseye set of data, by observing mother nature's 5 billboards which advertise imminent avalanches:

  1. Evidence of a recent avalanche. A recently avalanched slope indicates that slopes with the same characteristics (elevation, pitch, etc) are also likely to avalanche. An area of slope that is right next to a recently avalanched slope and of similar conditions is ripe for an avalanche.
  2. Wumping sounds. The whole snowpack sinks by about 3 inches, causing a guttural reaction in skiers known as the sphincter factor. A weak layer of snow is collapsing, which causes the sheer strength between layers to be zero. The wump can travel all the way up to the top of a slope. Unfortunately, if you experience a wumping sound on a steep slope, an avalanche is imminent and likely to be upon you.
  3. Hollow feeling/hollow sounds: It is like skiing on a 50 gallon drum. Snow piling in from a strong wind creates an airy layer, which can collapse underfoot.
  4. Shooting cracks: A crack will lightning bolt to an anchor object like a tree.
  5. Heavy windloading. Most avalanches occur within 24 hours of heavy precipitation. Avalanche likelihood increases significantly with 2cm/hour of accumulation, with a snow density of greater than 10%, for 5-6 hours, with 25 mph wind. If any of those factors increases, the likelihood increases.

After an avalanche path runs its course, the snowpack will not avalanche again at that location. If there are no possible avalanche release points above or to the side of the area, the avalanched area is considered safe.

Videotape highlights

There are a few VHS videos on avalanches - some are more for entertainment, and some have educational value:


A few examples are from Nova, National Geographic, Discovery, and Burton.


The most recommended videotape for educational information on avalanches is "Avalanche Awareness: A Question of Balance" from 1988. A few highlights from the film:

Avalanche dynamics

If you are caught in an avalanche, the old conventional wisdom includes advice on what to do while you are caught up in the sliding snow: However, this advice is mostly obsoleted:

A lot of the old advice was derived from people who recounted their experience surviving an avalanche, and this data is obviously skewed - it would be ideal to also include information from people who did not survive.

The moment of an avalanche is the event horizon. Before an avalanche happens, you are in total control: you decide where to go and your risk exposure. After the avalanche, the avalanche is in control, and you have zero control over what is happening. It is better to spend time, energy, and money that will help you evaluate avalanche risk, rather than focusing on what to do after you are in an avalanche.

When an avalanche occurs, kinetic energy is released, causing friction between snow crystals, which melts the outer part of snow crystals. After the avalanche comes to a stop, the mixture refreezes. The resulting mixture is 3-4 times as dense as when it started out, and feels like concrete, which makes it very difficult to dig yourself out. The refreezing process also causes the snow to expand by about 3%. The expansion makes it feel like you are in a body cast, and reduces the space that your lungs have to expand. Because of this entombing process, it doesn't make sense to try to do anything like swim to the surface or get rid of things when you are caught up in an avalanche. The density of an avalanche while it is traveling down the mountain is 1-5%, which means you are likely to sink to the bottom.

Of the people killed in avalanches, 2/3 die from suffocation, and 1/3 die from trauma.

Route selection: what you fall back on if you encounter shades of gray.

Statistics from the Colorado Avalanche Information Center

Fatalities per country, 1986-2004

The difference in fatalities in different countries is mainly attributed to different attitudes about risks of backcountry skiing:

Country Fatalities
 France  556
 Austria  471
 Switzerland  435
 USA  416
 Italy  361
 Canada  228
 Germany  37

At Alpine Meadows, there are about 300 avalanche points where avalanche risk is determined. In 1982, 7 people were killed at Alpine Meadows in an avalanche that occurred within ski boundaries. Alpine Meadows was found not liable, because they demonstrated that they followed procedure and had taken adequate precautions.

In the United states, avalanche control is done for all areas within the ski area boundary. It is still possible to get caught in an avalanche in-bounds at a US ski resort, however the in-bound fatality rate since 1982 is zero. Class A ski resorts have one full time avalanche forecaster.

In Europe, there is no such thing as out of bounds: there is only on-piste and off-piste. No avalanche control is done on areas off-piste - that includes in the trees between runs. Prince Charles and a friend went off-piste to escape the paparazzi at a ski resort, triggered and avalanche, and his friend was killed. In Europe, the ski patrol charges a heafty sum for a rescue.


Over the past few years, there has been an increasing trend of more avalanche accidents, due to an increase in the skill level of the sport, without an increase in avalanche awareness. Also, backcountry snowpack is one of the most difficult environments to understand, even more difficult to analyze than whitewater rivers.

Age group correlation, 1951-2003

The 20 to 29 age group has the highest avalanche fatality rates:

Age group Rate of Fatalities
>50 30
45-49 27
40-44 38
35-39 75
30-34 60
25-29 110
20-24 86
15-19 58
10-14 18
<10 9

Seasonal correlation, 1951 - 2003

A small number of avalanche accidents happen in summer: When mountaineering in the summer, avalanche awareness is still required when hiking snowy areas.

Month Rate of Fatalities
Nov 40
Dec 75
Jan 140
Feb 150
Mar 120
Apr 70
May 30

Fatalities by US state, 1986-2004

Colorado has by far the highest avalanche accident rate, due to two factors:

Fatalities (total = 416)
CO 114
AK 82
WA 29
UT 53
MT 51
CA 15
WY 31
ID 22
NH 7
NV 3
OR 7
NY 2
AZ 1
NM 1

Local Trend Correlation

The number of fatalities in the California/Nevada area has reduced in the past few years.

Sport type correlation, 1986-2004

Climbers and skiers in the backcountry have the largest number of avalanche fatalities. Snowboarders have a correspondingly small number of fatalities. The awareness among snowboarders has gone up significantly in the past few years. Snowmobilers have a high rate of avalanche accidents, due to a widespread lack of avalanche awareness. In particular, one snowmobile game is "high-mark," in which a group of snowmobilers see who can get the highest on a steep slope. The one who wins usually triggers an avalanche.

Fatalities (total = 416)
Snowmobilers 121
BC  skiers 81
Climbers 64
Misc rec 41
Lift Skiers, out of Bounds 41
BC Snowboarders 25
Snowboarders, out of bounds 14
Residents 9
others @ work 7
Ski patrollers 6
Motorists 4
Lift Skiers (in area) 3

Training correlation, 1981-1997

The highest rates of avalanche accidents happen to people who have the greatest avalanche training (not necessarily experience), and also to people who have the highest skill in a snow sport:

  Skill level
Training  novice intermediate advanced
advanced  0%  0%  38%
some  0%  7% 28%
none  7%  8%  12%

Avalanche forecasting should be considered a skill sport: In addition to training, it takes practice to become proficient.

Time to rescue vs death, 1951-2003

The faster a victim is dug out of an avalanche, the greater the chance of survival:

 Time to recover  Recovery
15 minutes 80%
30 minutes 50%
60 minutes 30%
90 minutes 20%
120 minutes 15%
180 minutes 7%

Depth vs death

The deeper the victim, the greater likelihood of death:

Depth  Recovery rate
< 1 foot 85%
2 FT 59%
3-3.9FT 30%
4 FT 40%
5-5.9 FT 30%
6 FT 21%
>7 FT 1 person

Method of rescue vs death, 1951-2003

 Method of Rescue  Found Alive  Found Dead
 Attached object / body part located 140 54
 Hasty search / spot probe 26 46
Coarse/fine Probe 23 163
 Transceiver 55 83
 Dogs 6 50
 Voice 30 1
 Inside vehicle 30 11
 Inside structure 23 30

When digging for a victim, try to be quiet - you might be able to hear the victim. Sound travels very well into the snow (the victim can hear the rescuer), but sound travels very poorly out of the snow (the rescuers cannot always hear the victim).

An avalanche cord may not be effective because:

Dogs are not effective because it takes a long time to carry them in: They cannot usually hike in, and usually must be brought in by helicopter.

Type of rescue vs survival, 1951-2003

   Self  Party  Rescue team
 Alive 53 243 58
 Dead  0 117 361

The highest proportion of survivors are found by the victim's own party. It takes too long to bring in a rescue team.


In the backcountry, there are three items that are required for avalanche rescue:

  1. Beacon
  2. Probe
  3. Shovel

The probe and shovel can be used for measurement. In addition, it is useful to have a clinometer and a thermometer.


There are two types of beacons: Analog and digital. Both work with each other. There are also a few beacons that can operate in both modes.

For people who don't practice often, a digital beacon is probably the better option, since it's easier to use. However, an analog beacon works as well as a digital beacon if you practice. All modern beacons were standardized to use the 457 KHz frequency in 1996. Prior to 1996, a different frequency was used that was less effective. Verify that everyone in your party is using a beacon that uses the 457 KHz frequency. Beacons have only a rescue function: they have no measurement or other use. If a beacon has an LCD readout, make sure it is visible in glaring sunlight.


Probes have cm marks, and can be used to measure snowpack depth. Get a probe that is at least 8 feet long. The resistance you feel as the probe penetrates layers can also give a rough indication of the snowpack layering. Probes are constructed like tent poles, and can be extended easily. They range from light to heavy (depending on price). Avoid probes with compression buttons - they get jammed with ice.

You can buy ski poles that can be taken apart and re-connected as probe poles, however:

However, it does not hurt to have dedicated probe poles andconvertible ski poles as a contingency. Keep your probe in your backpack and within easy reach - you are more likely to use the probe for routine measurements if it's easy to get. Don't store a folded-up probe in a shovel handle.


Get a shovel with a flat profile, to make it easy to do sheer tests. the larger the blade, the faster you can dig someone out. Avoid plastic shovels - avalanche debris is too hard for plastic shovels, and plastic shovels are not good for shear testing.

If you are going to lend rescue gear to other people in your group, give them the good stuff, because they may use it to rescue you. It is also a good idea to become an expert in teaching your friends how to use the equipment.

Ava-Lung & Airbags

Black Diamond developed the Ava-Lung, a vest worn over all other clothes. It gathers air directly from the snow through mesh panels which then is available to breathe through a snorkel. It also exhales the air in a different place that where it gets the air to prevent an ice mask from forming. The Ava-Lung will theoretically keep an avalanche victim alive longer before he suffocates. However, in the event of an avalanche, there may be a problem getting the snorkel into your mouth and keeping it there. There has been at least one incident where a guide was able to make effective use of the Ava-Lung.

An ABS pack consists of 2 balloons that expand. They provide flotation to keep you on top of an avalanche. About 60-70 people in Europe have been able to make effective use of ABS airbags in avalanches.

The Ava-Lung and the ABS packs should not be relied upon for safety. Instead, it is better to avoid avalanches in the first place. However, these devices can be effectively used when you are placing control of a situation in the hands of a guide. Even experienced guides can mess up big time. On January 20, 2003, seven people, including Craig Kelly, died in an avalanche on a guided tour as part of the Selkirk Mountain Experience (SME) in Canada. The skiers were guided by Ruedi Beglinger, considered to be a master guide.

Subjective justification

If you are about to enter an area and you are thinking to yourself: "now is a good time to set my beacon on transmit" or "now is a good time to get the ava-lung snorkel ready," re-evaluate the risk. If you have a gut feeling that a situation does not feel right, see if you can quantify the situation with actual measurements. Making observations to back up your gut feeling is sometimes necessary to avoid peer pressure if you are in a group that has a higher tolerance for risk than you.


A strong, heavy layer over a weak, light layer of snow represents a high avalanche risk. In the Sierra, storms often start with heavy, dense snow and finish with light, airy snow, called a right-side-up snowpack. However, if two storms happen back to back, the interface between the light layer from the first storm and the heavy layer from the second can be weak. Sometimes double storms happen so close together that you can't discern one storm from the next: these situations result in extremely high probability of avalanche. Back-to-back storms in the 1999-2000 season in Tahoe caused avalanches in Crystal Bay at Lake level.

Snow metamorphism

Snow starts changing the moment it lands. Snowpack can change from weak to strong and back over time. The warmer the temperature, the faster the changes. Elements of snowpack can include the following characteristics:

One question to ask is how deep is the snow:

Snow is more like air/atmosphere than like the ground: Skis and snowboards are actually "wings" that fly through the air.

Evaluating new snow:

There are three types of metamorphosis:

  1. Rounded - gains strength
  2. Faceted - weakens
  3. Melt-freeze - both

Rounding - dull / lackluster. Rounded crystals are created when the air is warm and the snowpack is deep. These conditions mean that the air in the snowpack is stagnant.

Faceting - Sparkles / looks sugary. Faceted crystals are created when the air is cold and snowpack is shallow. These conditions mean that the air in the snowpack is rising.

Melt/freeze - Melt-freeze crystals are created when the air goes above and below 0° C.

Snow strength determination

Snow strength is determined by three factors:

  1. Air temperature: Lay a thermometer on the surface in the shade.
  2. Snow depth - use a probe to measure
  3. Temperature at the snow-ground interface - this temperature will be 0° C outside of the arctic circle.

In a layer of snow, if the temperature gradient decreases by 1° C or more every 10 cm starting from the ground, faceting results, otherwise rounding results. Example: If the depth of the snowpack is 200 cm, the temperature of the snow at the snow surface must be colder than -20° C for faceting to take place. This analysis is somewhat simplified because it only works on a layer-by-layer basis. Some layers have gradients, and some don't.

The Sierra often has deep snowpack and warm temperatures, which creates rounded crystals. Most winds in the Sierras come out of the Southwest, pulling up warm air from the tropics (starting out with heavy, dense snow), then bring in lows from the North (ending in light, low density snow). In Colorado, the snowpack is typically shallow and cold, which tends to create faceted crystals. Therefore, Colorado is more susceptible to weaker faceted layers, and avalanche danger is typically higher. However, early in the season in the Sierra, you can have shallow snowpack and cold temperatures.

The dew point is the temperature at which the water vapor in air starts to condense. As water vapor rises, it reaches the dew point.

Radiation balance

If the solar radiation is S, and the terrestrial radiation is T, then rounding/faceting is determined as follows:

If it snows early, the snow can create a faceted base, which means that every subsequent snowpack will slide easily.

Avalanche terrain risk scale

Factors include: Forces, Obstacles, Rescue Time (FORT)

The scale represents the avalanche potential, not the expected number of avalanches. Risk is also classified as low, moderate, considerable, high, and extreme.


  1. Snow Sense by Jill Fredston & Doug Fesler, 1999. ISBN 0-9643994-0-7. This small book is condensed and to the point. It also focuses on decision making, and you can take the book out into the field with you and refer to it. It was one of the first books to provide avalanche prediction guidelines, based partly on information gleaned from interviews with avalanche survivors.
  2. Avalanche Safety for Skiers, Climbers and Snowboarders 2nd Edition by Tony Daffern, 1999. ISBN 0921102720. A good reference. Published by Mountaineers Books
  3. Sledding in Avalanche Terrain: Reducing the Risk, by Bruce Jamieson, 1999. ISBN 0969975872. A good guide for snowmobilers.
  4. The Avalanche Handbook by David McClung & Peter Schaeter, 1993, ISBN 0898863643. Highly technical, engineering-oriented handbook, only useful for professionals. It's not for recreational backcountry skiers.
  5. Snowy Torrents: Avalanche Accidents in the United States 1984, ISBN 0933160135
  6. Snowy Torrents: Avalanche Accidents in the United States 1980-1986 by Nick Logan & Dale Atkins, 1996, ISBN 1884216528. A series of short case studies of avalanche accidents. From reading this book, it becomes obvious that people make the exact same mistakes over and over again.