Mountains are undoubtedly one of the most beautiful and fascinating panoramas of the Earth. Many strive to conquer the majestic peaks, not fully realizing how harsh such beauty is. That is why, deciding on such a brave step, extreme people face difficulties in all their manifestations.

The mountains are a rather dangerous and complex terrain, in the expanses of which there is a constant mechanism of gravity, so the destroyed rocks move and form plains. Thus, the mountains eventually turn into small hills.

In the mountains, danger can always await, so you need to undergo special training and be able to act.

Definition of Avalanches

Snow avalanches are one of the most devastating, dangerous destructive phenomena of nature.

An avalanche is a rapid, sudden, minute process of moving snow with ice, occurring under the influence of gravity, water circulation and many other atmospheric and natural factors. Such a phenomenon most often occurs in the winter/spring period, much less often in summer/autumn, mainly at high altitudes.

It is always worth remembering that the avalanche is primarily a harbinger of weather conditions. Hiking in the mountains in bad weather: snowfall, rain, strong wind - is quite dangerous.

Most often, an avalanche occurs, lasting about a minute, while passing a distance of about 200–300 meters. It is extremely rare to hide or run away from an avalanche, and only if it became known about it at least 200–300 meters away.

The avalanche mechanism consists of a sloping slope, an avalanche body and gravity.

Sloping slope

The level of the slope, the roughness of its surface greatly affect the avalanche risk.

A slope of 45–60° is usually not dangerous, as it gradually unloads during snowfalls. Despite this, such places under certain weather conditions can create avalanche accumulations.

Snow will almost always fall from a slope of 60–65°, and this snow can linger on convex sections, creating dangerous blowouts.

Slope 90 ° - the collapse is a real snow avalanche.

avalanche body

Formed from accumulations of snow during an avalanche, it can crumble, roll, fly, flow. The type of movement directly depends on the roughness of the lower surface, the type of snow accumulation, and swiftness.

Types of avalanches according to the movement of snow accumulations are divided:

• to streaming;
• cloudy;
• complex.

Gravity

It acts on the body on the surface of the Earth, is directed vertically downwards, being the main mobile force that contributes to the movement of snow accumulations along the slope to the foot.

Factors affecting the occurrence of an avalanche:

• type of matter composition - snow, ice, snow + ice;
• connectivity - loose, monolithic, reservoir;
• density - dense, medium density, low density;
• temperature - low, medium, high;
• thickness - thin layer, medium, thick.

General classification of avalanches

Avalanches of powdered, dry recent snow

The convergence of such an avalanche usually occurs during heavy snowfall or immediately after it.

Powder snow is called fresh, light, fluffy snow, consisting of tiny snow flakes and crystals. The strength of snow is determined by the rate of increase in its height, the strength of the connection with the ground or previously fallen snow. It has a fairly high fluidity, which makes it possible to easily flow around a variety of obstacles. In different cases, they can reach speeds of 100–300 km / h.

Avalanches created by snowstorms

Such a convergence is the result of the transfer of snow by a blizzard. Thus, the snow is transferred to the mountain slopes and negative landforms.

Avalanches of dense dry powder snow

They arise from snow a week old or more, which during this time is pressed, becomes much denser than freshly fallen. Such an avalanche moves more slowly, partially turning into a cloud.

landslide avalanches

They grow after the collapse of snow cornice blocks, which set in motion a large amount of snow.

Dust avalanches

An avalanche is characterized by a huge cloud or a thick layer of snow on trees and rocks. Created when dry, powdery recent snow melts. A dust avalanche sometimes reaches a speed of 400 km/h. Risk factors are: snow dust, strong shock wave.

Formation avalanches

They arise through the descent of layered snow, reach a speed of 200 km / h. Of all snow avalanches are the most dangerous.

Avalanches from hard sheet snow

A stream is formed by the descent of solid layers of snow over a weak, loose layer of snow. They consist mainly of flat snow blocks resulting from the destruction of dense formations.

Soft plastic avalanches

Snow flow is formed by the descent of a soft layer of snow on the underlying surface. This type of avalanche is created from wet, settled, dense or moderately cohesive snow.

Avalanches of monolithic ice and ice-snow formations

At the end of winter, snow deposits remain, which, under the influence of external factors, become much heavier, turning into firn, eventually turning into ice.

Firn is snow cemented by frozen water. It is formed during temperature drops or fluctuations.

Complex avalanches

Consist of several parts:

• flying cloud of dry snow;
• a dense stream of formation, loose snow.

They arise after a thaw or a sharp cold snap, which is the result of the accumulation of snow, its separation, thereby forming a complex avalanche. This type of avalanche has catastrophic consequences and can destroy a mountain settlement.

Avalanches are wet

Formed from snow accumulations with the presence of bound water. Occur during the period of accumulation of moisture by snow masses, which occurs during precipitation and thaw.

Avalanches are wet

They arise due to the presence of unbound water in snow accumulations. Appear during a thaw with rain and warm wind. They can also occur by sliding a wet layer of snow over the surface of old snow.

Mudflow-like avalanches

They arise from snow formations with a large amount of moisture, the driving mass of which floats in a large volume of unbound water. They are the result of long thaws or rains, as a result of which the snow cover has a large excess of water.

The presented types of avalanches are quite dangerous, fast-moving flows, so you should not think that some are safer than others. The basic safety rules must always be followed.

Avalanche safety

The term avalanche safety refers to a set of actions aimed at protecting and eliminating the tragic consequences of avalanches.

As practice shows, in most accidents the extreme people themselves are to blame, who, without calculating their own strength, themselves violate the integrity and stability of the slopes. Unfortunately, there are deaths every year.

The main rule for the safe crossing of mountain ranges is the full knowledge of the passable territory, with all the dangers and obstacles, so that in an emergency you can safely, carefully leave the dangerous section of the path.

People going to the mountains, the basic rules of avalanche safety, be able to use avalanche equipment, otherwise the likelihood of falling under a snow blockage and death is very high. The main equipment is avalanche shovels, beepers, avalanche probes, float backpack, maps, medical equipment.

Before going to the mountains, it will be useful to take courses on rescue operations during a collapse, first aid, making the right decisions to save lives. Also an important stage is the training of the psyche and ways to overcome stress. This can be learned in courses on working out techniques for saving people or yourself.

If a person is a beginner, it will be useful to read books about avalanche safety, which describe different situations, moments, stages of their overcoming. For a better understanding of avalanches, the best option would be personal experience obtained in the mountains in the presence of an experienced teacher.

Avalanche safety basics:

• mental attitude and preparation;
• obligatory visit to the doctor;
• listening to an avalanche safety briefing;
• taking with you a sufficient amount of food, small in volume, a spare pair of clothes, shoes;
• careful study of the route, upcoming weather conditions;
• taking a first aid kit, flashlight, compass, equipment on a hike;
• departure to the mountains with an experienced leader;
• studying information about avalanches in order to have an idea of ​​the degrees of avalanche safety during a collapse.

The list of avalanche equipment, which you need to be able to work with confidently, quickly, for your own safety and rescue of the victims:

• victim search tools: transmitter, avalanche ball, beeper, radar, avalanche shovel, avalanche probe, other necessary equipment;
• tools for checking snow flooring: saw, thermometer, snow density gauge and others;
• tools for rescuing victims: backpacks with inflatable pillows, avalanche breathing apparatus;
• tools for transporting victims, as well as medical equipment: bags, stretchers, backpacks.

Avalanche slopes: precautions

In order to avoid getting into an avalanche or if there is a high probability of an avalanche situation, you need to know a few important rules for avalanche safety and how to prevent it.

• move on safe slopes;
• do not go to the mountains without a compass, know the basics of the direction of the winds;
• move along elevated places, ridges that are more stable;
• avoid slopes with snow cornices hanging over them;
• return along the same road that went ahead;
• monitor the top layer of the slope;
• do tests on the strength of the snow cover;
• it is good and reliable to fix the insurance on the slope, otherwise the avalanche can drag a person with it;
• take on the road spare batteries for the phone and a flashlight, and also have in the memory of the mobile phone the numbers of all nearby rescue services.

If a group or a certain number of people still find themselves under an avalanche, you need to call rescuers, immediately starting the search on your own. In such a situation, the most necessary tools will be an avalanche probe, a beeper, a shovel.

Every person who goes to the mountains should have an avalanche probe. This tool performs the function of sounding snow during search operations. It is a dismantled rod, two to three meters long. In safety courses, an obligatory item is the assembly of an avalanche probe in order to assemble it in the shortest possible time when creating an extreme situation.

An avalanche shovel is indispensable when searching for victims, it is necessary for digging snow. It is more effective when combined with an avalanche probe.

A beeper is a radio transmitter that can be used to track a person covered in snow.

Only with coordinated quick action you can save your friend. After a thorough briefing on avalanche safety, a person will be mentally and physically ready to help others.

As a result, I would like to emphasize that hiking in the mountains cannot be carried out in bad weather, in the evening or at night, when crossing a dangerous area, it is necessary to use rope insurance, be sure to have beepers, flashlights, avalanche shovels and avalanche probes in the arsenal. Some part of these instruments must necessarily have a length of 3-4 m.

Observing all the rules, following the instructions, a person will protect himself from disastrous consequences and return home safely.

Write to us if the article was useful.

The materials of the site www.snowway.ru and from other open sources were used.

Avalanches: mass of snow; natural process. Preconditions for formation: accumulation of snow; gravity; friction force; slope steepness 25 - 60 ° (but it can also be 15 °); snow properties.

Snow cover.

1. Types of snow and conditions of formation: fresh snow (freshly fallen snow (fluffy, loose), freshly deposited snow, blizzard snow), old snow, firn.

2. Changes in the structure of snow under the influence of wind and solar radiation, temperature, deep frost.

3. Distribution of forces in the snow cover lying on the slope: steady state, unstable state, stressed balance.

Avalanche elements: origin zone, separation line (point), transit zone, avalanche body, alluvial cone, deposition zone.

Types of avalanches.

1. By type of snow: avalanches from a snow (wind) board, avalanches from freshly fallen (fluffy and loose) and freshly deposited snow (dust avalanches), wet avalanches (from wet, moist, moistened snow).

2. According to the form of snow movement in the transit zone: snow landslide, flume, jumping avalanche.

Avalanche-prone forms of mountainous relief: open steep slopes, gently convex slopes, a concave slope and a prepass part of a hollow, cornices, couloirs, cirques, a slope with a recess. Valleys: trough-shaped, v-shaped, canyon.

exit conditions: weight gain (snowfall); reduction of friction (warming, cutting the slope with a path); vibration (loud sound, thunderstorm, tremors); impact on snow (falling cornice, stone, movement of people, wind); a sharp cold snap after heat, the formation of "deep frost".

Avalanche forecasting. The thickness of the snow cover (more than 30 cm). The steepness of the slope. The presence of obstacles on the slope (rock ledges, terraces, forest). Weather (snowfalls, rain, warming, temperature changes, wind). Time of day and position of the sun. Type and density of snow. The presence of alluvial fans, avalanches on neighboring slopes. Unexpected avalanches (gradual change in snow properties). Closure of mountainous areas in the off-season, artificial avalanches (explosions).

Avalanche equipment: avalanche cord, probe from ski sticks, probe with a hook, radio beacon, avalanche shovel (put on an ice ax). Can be used as a last resort for pot lids.

Avalanche cord marking: (tape) 15 - 25 m, bright, durable, slippery, with marking of the direction to the person and the distance to him.

When driving in avalanche-prone areas, you must adhere to the following rules:

1. In no case should you go to an avalanche section of the route with an unfavorable weather forecast, with a sharp warming, a drop in pressure, in fog, soon after a snowfall or heavy snowstorms.

2. Remember that avalanche danger is possible on all slopes steeper than 15°. If the depth of freshly fallen or old loose snow is more than 30 cm, slopes of 15 ° can be avalanche-prone. At the same time, the descent of one avalanche does not remove the danger for the same slope, since avalanches can descend several times in a row.

3. To reduce the avalanche danger, it is preferable to move along ridges, rocky ledges, groups of trees, bypass (even far) dangerous areas on a reliable terrain or above a possible separation line.

4. After snowfalls or blizzards, avoid crossing steep lee slopes even at the very top, do not go out onto snow cornices and under cornice accumulations of snow - snow bags.

5. Immediately leave the avalanche danger zone and stop further movement: a) during heavy snowfalls and poor visibility; b) during rain when there is a snow cover of 30 cm or more on the slopes; c) during strong winds and during a blizzard; d) with a sharp drop in temperature.

6. In the spring, with a cloudless night and in the absence of a hair dryer, movement is allowed in the morning from 4 to 12 o'clock.

7. Before driving, check the stability of the snow on the slope and determine the nature of the accumulation of snow on the route and on the slopes above it. When moving in an avalanche zone, it is necessary to select an observer and, before starting to cross the avalanche-prone slope, outline in advance the escape routes and escape to a pre-planned shelter from avalanches.

8. When moving along a slope that is suspected of being an avalanche hazard, in every possible way avoid moving across it or in a zigzag and go only straight up or down "on the forehead" - along the line of the circle, so as not to cut the snow layer and not cause an avalanche. Check the direction of movement along the line of fall of a thrown pebble or snowball. Crossings are allowed only on safe slopes or at least above an unstable layer, but in no case at the bottom of it and not in the middle. The narrowest places of the couloirs should be chosen, preferably above the confluence of its constituent gutters. Before such a crossing, remove the skis so as not to cut the top layer of the snow cover, free the hands from the loops of the ski poles, tie an avalanche cord to the belt, fasten all the buttons on the clothes, keep a scarf ready to cover the nose and mouth.

9. In every possible way avoid actions that can cause shaking of an avalanche-prone slope: jumps, falls, sharp turns on skis, screams, collapse of stones and cornices.

10. Do not gather in one place of the slope more than two in order to avoid overloading the snow layer and simultaneously falling into an avalanche. Apply the technique of insurance and movement from this calculation.

11. Observe the maximum distances between people both on the slopes and under them, in the area of ​​the avalanche cones. At the same time, conduct unremitting monitoring of comrades passing in the danger zone, preventing further movement of the entire group until there is a firm certainty that the last participant has passed the dangerous slope or section.

12. Avoid concave terrain, snow funnels and trays. All the time to pave the way with the expectation of being above the snow-gathering, and not under them.

13. In an avalanche-prone gorge, in cold weather, stick to the south, in warm sunny weather - the foot of the northern, shady slopes.

14. Avoid stops and halts on alluvial cones and in avalanche flumes.

15. Before a forced descent along an avalanche-prone slope, it is better to try to release the avalanche with stones, a broken cornice, or in another way.

When forced to cross an avalanche-prone slope, the following precautions should be observed:

1. Dissolve the avalanche cord, unfasten the ski bindings, remove your hands from the lanyards of sticks or ice axes, get ready to drop your backpack and other items.

2. Passing and crossing avalanche-prone slopes should be carried out quickly, but carefully, carefully forming each step in the snow, looking closely at the behavior of the snow layer, testing it from time to time.

3. Always keep a distance of 100 - 200 m, depending on the width of the dangerous slope or log. Cross avalanche sites only alone, under the supervision of a friend who monitors the slope and warns of the beginning of the avalanche movement by shouting "avalanche", after crossing the slope the roles change.

4. Walking in the avalanche zone should be done carefully, less frequently and more widely in order to disturb the snow surface as little as possible, not to fall and cause an avalanche. In order to maintain stability when walking, light pressure is first made with the foot, after which the foot is placed completely, compressing the snow.

5. If footsteps cause a dull sound like a distant rifle shot, a crackling sound or a snow board settling with a characteristic hiss, then you should immediately leave this area.

6. Observe silence so as not to weaken attention, avoid screams that are not caused by necessity.

7. A first aid kit, avalanche shovels and probes are at the last members of the group.

Snow cornices pose a great danger, since it is impossible to predict the time of their collapses.

Walking along the snow ledge, you need to:

1. Go along the cornice ridge below the line where the planes of the windward and leeward slopes intersect, and in any case, do not approach the edge of the cornice closer than 5-6 m.

2. Check the safety of the path by probing and inspecting the snowy surface.

3. Be sure to insure by binding to each other.

4. On suspicious ledges, especially after a snowfall or a blizzard, each of the group should lay their own trail (the first - above all, the last - below all along the windward slope).

5. Eaves ridges pass across in the narrowest part, with careful insurance, bringing down the ridge part to a dense firn base. When crossing from the side of the windward slope in the eaves, they make the deepest possible trench 0.5-0.6 m wide, along which they go down on the insurance one at a time.

When caught in an avalanche,:

1. Don't lose your temper. In dusty avalanches, first of all, pull a scarf over your mouth and nose or cover them with a hat, gloves to avoid suffocation from snow dust. Try to immediately get rid of unnecessary things (drop your skis, backpack, discard ski poles, etc.) so that you don’t get sucked into an avalanche.

2. If there is a reliable support under your feet and the avalanche has not yet gained speed, try to take the blow of the snow masses, letting the avalanche past you in order to be in its tail.

4. Make swimming movements with your arms and legs to keep yourself on the surface of the avalanche, while trying to get closer to the edge of the snow flow, swim to the surface before braking.

5. When diving into avalanche snow, before stopping it, you should try to cover your face with your hands and take a position face down, which saves you from quick freezing.

6. After stopping, immediately create as large a cavity as possible in the snow in front of the face.

7. Do not fall asleep, do not scream, because the scream is still not heard through the snow, and the victim will be exhausted.

8. If you manage to make a hole for air access, but you can’t get rid of the snow masses, then try to stick your hand out to the surface in order to attract the attention of the rescuers.

9. You can determine your position in space by releasing saliva.

10. Do not lose hope for salvation, since a person who has not received fatal injuries can, in some cases, lie under the snow for up to two days.

The observer fixes the "place of disappearance" in the avalanche, escorting him along the moving clods of snow until the avalanche stops, fixes the "stopping point".

The search for the person caught in the avalanche must be carried out according to a strict system that excludes multiple surveys of the same places, while other areas could remain unexplored. Before the start of search operations, it is necessary to set up an observer to warn about the recurrence of avalanches. The search begins, guided by the following rules:

1. If an avalanche hit a person from above, then he should be looked for on the periphery of the avalanche cone.

2. If the avalanche broke from under the feet of the victim, then he should be searched along the upper edge of the avalanche.

3. If a person was kept for some time on the surface of the avalanche, and then disappeared, then he should be sought below this place and at a considerable depth.

4. If there were various obstacles on the way of the avalanche (stones, walls of cracks, stumps, depressions, etc.), then the search is carried out first of all at these obstacles.

5. If the avalanche has stopped due to friction on the surface of the slope, the search should begin 5-10 m before the end of the avalanche cone.

6. If an avalanche has overflown over an obstacle (for example, a lateral moraine), then in most cases the victim is in front of him.

In the event of a possible re-avalanche, rescuers should post a slope watcher and disband their avalanche cords.

If the victim’s companions have marked the place of his disappearance, you must first quickly but carefully inspect suspicious places, moving in a line from the landmark down the surface of the avalanche in search of protruding parts of the body, clothing, avalanche cord or equipment. If an avalanche cord is found, it must be quickly and accurately dug out, avoiding a break, and the location of the backfilled one should be determined.

In the absence of a positive result after examining the avalanche, it is necessary to carry out high-speed probing with ski poles with removed rings, special probes, and an ice ax. To do this, the probers stand facing the slope and, on command, plunge the probes into the snow for the entire length. The distance between sounding points along the line should not exceed 75 cm. Then move up the slope by 70 cm and repeat the operation. As you move forward, you must carefully maintain the intervals.

If the double high-speed probing did not give a positive result, they start detailed probing with an interval of 25-30 cm. To do this, the probes move up in a strictly straight line, moving almost shoulder to shoulder, placing their legs with their feet turned. At the command of the leader, the line stops and everyone probes first at the toe of the left foot, then between the feet, and then at the toe of the right. When the sounding is completed by all participants, at the command of the leader, the line advances by 25-30 cm (the alignment is monitored by the right-flank one) and repeats the operation.

During sounding, complete silence must be observed so that the sounders can not only feel, but also hear the impact of the probe on various objects and possible sounds made by the victim. The probe should be immersed in snow strictly vertically. It is advisable to insert the probe into the snow with one hand (without gloves), slowly turn it 180° and pull it out. Inspection of the tip establishes the nature of the encountered obstacle.

It should be remembered that probing must be carried out very carefully, since the probe can cause injury when covered by an avalanche.

If the probe does not reach the ground, it is necessary to dig trenches after the first sounding. Trenches are dug along the slope of the slope from the bottom up, starting slightly below the possible location of the backfilled. The distance between the walls of adjacent trenches should not exceed 4 m, the width of the trenches is 1-1.2 m. The depth of the trenches should be sufficient so that from its bottom it is possible to reach the ground with a probe not only directly under the trench, but also obliquely between the trenches. Probing of the bottom of the trenches and the space between the trenches should begin after their penetration, but without interfering with the work on their penetration.

When the location of the victim is established, it is necessary to mark it and proceed with the excavation. You should dig quickly, but as you approach the victim, you need to be careful.

When it was possible to get to the fallen asleep, he immediately needs to provide first aid: as soon as possible, free his face with his hand, trying at the same time to free his mouth and nose from snow and dirt. After cleaning the mouth and nose from snow, it is necessary to start artificial respiration using the mouth-to-mouth or mouth-to-nose method, dig the victim out from under the snow as soon as possible and transfer him, taking into account possible injuries, to the place where further assistance will be provided to him. In this case, the victim should be put on a thick, dense bedding and covered warmly, put warm compresses or heating pads under his back, stomach and sides, continue artificial respiration when the victim regains consciousness, give him a drink and liquid nutrition.

An attempt to revive the victim can only be stopped when there are clear signs of death.

This section describes the scientific approach to predicting avalanche danger.

Types of forecasts

Currently, three types of avalanche hazard forecasts are used - a small-scale background for a mountainous area, a large-scale background for a mountain basin or a group of avalanches, and a detailed one for a given avalanche or avalanche-prone slope (local forecast).

An avalanche forecast presupposes the advance determination of a certain time interval during which snow accumulation and metamorphism processes can lead to a violation of the stability of the snow cover and the formation of avalanches. It is closely related to the forecast of meteorological conditions, since the type, intensity of precipitation, the amount of precipitation, blizzard snow transfer, air temperature and humidity, and other characteristics of meteorological conditions directly affect the state and stability of the snow cover.

The background forecast consists in assessing the avalanche danger in the mountainous area under consideration and is issued in the form of "avalanche" or "non-avalanche". The lead time of avalanche forecasts is limited by the lack of quantitative methods for long-term forecasting of precipitation intensity, thaw intensity and duration, and other meteorological indicators in the mountains. Usually it is measured in hours, and often the forecast is issued with a "zero" lead time, i.e., only a current assessment of the avalanche danger is given.

The local forecast provides for determining the indicators of snow cover stability in the avalanche initiation zone of a specific avalanche collection and the time before the expected spontaneous avalanche release, assessing the probable volume and range of the avalanche release, choosing the optimal conditions for eliminating the avalanche danger by artificially lowering the avalanche.

Avalanche forecasting methods were developed back in the USSR, starting in the 1930s, first in the Khibiny, then in the Caucasus, where they found wide practical application. In the postwar years, significant progress in forecasting avalanche danger was also achieved in the mountains of Central Asia, Kazakhstan and South Sakhalin.

The background forecast of avalanches caused by snowfalls and blizzards is the most developed. Some progress has also been made in the development of background avalanche forecasts from wet snow, based mainly on the analysis of the snow-meteorological situation and established statistical relationships between the time of onset of avalanche danger and changes in the factors that determine avalanches. It uses all available information about the structure, density and temperature regime of the snow cover and local characteristics of its stability.

Methods for local forecasts are still poorly developed, which is due to the lack of methods and equipment for obtaining reliable information about the state and properties of snow cover in the zones of avalanche origin, and the accuracy of existing methods for determining the strength characteristics and indicators of snow cover stability is low.

Forecast of avalanches caused by snowfalls and snowstorms.

Snowfalls and blizzards directly affect the stability of the snow cover, so the avalanches caused by them are called "direct action" avalanches. However, other factors also have a significant influence on the processes of avalanche formation. For a qualitative assessment of the likelihood of avalanches, 10 main avalanche-forming factors are evaluated (Snow avalanches, 1965):

— The height of the old snow. The first snowfalls are usually not accompanied by avalanches. Snow first fills the unevenness on the slope, and only after that can a flat, smooth surface appear, contributing to the sliding of new layers of snow cover. Therefore, the greater the height of the old snow before the onset of snowfall, the greater the likelihood of an avalanche formation. In this case, the ratio of the height of old snow to the characteristic dimensions of irregularities on the slope is very important. So, on smooth grassy slopes, an avalanche hazard can occur at a snow cover height of 15-20 cm, and on slopes with large rocky ledges or bushes - only at a height of old snow of 1-2 m.

— Condition of old snow and its surface. The nature of the snow surface affects the adhesion of freshly fallen snow to the old one. The smooth surface of wind-driven snow slabs or ice crust favor avalanches. The likelihood of buckling of fresh snow is increased if such a surface has been covered with a thin layer of powdered snow. A rough surface, wind sastrugi, porous crusts from rain, on the contrary, reduce the possibility of avalanche formation. The features of old snow determine the amount of freshly fallen or blizzard snow that it can withstand without collapsing, and its ability to stay on slopes without being involved in an avalanche when new snow slides over it. The presence of layers and interlayers of deep frost is especially predisposing to avalanche formation, the formation of which, in turn, is determined by the type of slope surface and the thermodynamic conditions of snow cover recrystallization processes.

- Height of freshly fallen or blizzard-deposited snow. An increase in the depth of snow cover is one of the most important factors in avalanche formation. The amount of snowfall is often used as an indicator of potential avalanche danger. For each area there are certain critical heights of fresh snow, above which there is an avalanche danger. However, one must always remember that snow depth as an indicator of avalanche danger should be used in conjunction with other avalanche factors.

— View of freshly fallen snow. The type of solid precipitation that occurs affects the mechanical properties of the snow cover and its adhesion to old snow. So, when cold prismatic and needle-shaped crystals fall out, a loose snow cover is formed, characterized by low cohesion. It is also formed when star-shaped crystals fall out in frosty calm weather. If the air temperature is about 0 °, then snowflakes during the fall can combine and fall out in the form of large flakes. The snow cover of such particles is rapidly compacted. The greatest probability of avalanche formation occurs when a cover is formed from freshly fallen fluffy and dry fine-grained snow; often avalanches form from dry compacted snow, and when wet and wet snow is deposited, avalanches rarely occur.

— Density of freshly fallen snow. The highest probability of avalanche formation is observed during the formation of snow cover of low density - less than 100 kg/m 3 . The greater the density of new snow during a snowfall, the less likely avalanches are. Increasing the density of snow reduces the likelihood of avalanches, but this rule does not apply to snow slabs formed during snowstorms.

— Intensity of snowfall (speed of snow deposition). At low snowfall intensity, the decrease in the stability index of the snow cover on the slope as a result of an increase in shear forces is compensated by an increase in stability due to an increase in adhesion and friction coefficient during snow compaction. As the rate of snow deposition increases, the effect of an increase in its mass prevails over the effect of its compaction, and conditions are created for a decrease in the stability of the snow cover and the formation of avalanches. For example, in the Tien Shan regions, with snowfall intensity up to 0.15 cm/h, avalanches are not observed, and when it increases to 0.8 cm/h, avalanches are observed in 45-75% of cases.

— Quantity and intensity of precipitation- a factor essentially corresponding to the previous one. It more accurately characterizes the increment in the mass of snow per unit area of ​​the horizontal projection of the slope, including taking into account liquid precipitation and snowstorms.

- Settling of snow. The processes of compaction and settling of the falling snow increase its adhesion and the coefficient of internal friction and thus contribute to the stability of the snow cover. Snow of low density has a low initial strength, but quickly compacts; dense snow with a high initial strength settles slowly. Snow settling is important both during a snowfall or a blizzard, and immediately after they end. Avalanche formation is sometimes influenced by the settling of old snow (for example, uneven snowfall under a solid snow slab can lead to a break in the slab and a violation of its stability).

- Wind. Wind transport leads to a redistribution of the snow cover and to the formation of hard crusts, snow slabs and puffs. The wind forms snow cornices, and below them, accumulations of loose snow. A strong wind creates a suction of air from the snow mass, which contributes to the migration of water vapor and loosening of the lower layers of snow. In the processes of avalanche formation, the wind plays an important role, especially as a factor of blizzard snow transport and accumulation.

- Temperature. The effect of temperature on avalanche formation is multifaceted. The air temperature affects the type of falling solid precipitation particles, the formation, compaction and temperature regime of the snow cover. Differences in the temperature of the snow cover in depth determine the rate and nature of metamorphism processes. Snow temperature significantly affects the characteristics of its viscous strength properties. A rapid decrease in air temperature can lead to the formation of temperature cracks in the rupture of the snow layer and the occurrence of avalanches.

In the United States, attempts were made to use information about avalanche-forming factors for rapid assessment and forecasting of avalanche danger. For this purpose, each of the listed factors was evaluated according to a ten-point system, depending on its predisposition to avalanche formation, then these points were summed up. The possible scores are 0 to 100. The higher the score, the more likely the avalanches, 0 means no avalanche danger, and 100 means the highest probability of avalanches.

Similar methods for assessing avalanche-forming factors for background forecasts of avalanche danger are also used in some avalanche-prone regions of Russia. To predict avalanches, the time of snowfalls for the Northern Tien Shan region, in addition to the listed 10 factors, the characteristics of synoptic processes and the stability of the snow mass are also used. When analyzing the synoptic processes leading to snowfalls and avalanches, the most typical situations were identified and their quantitative assessment in points was given. The stability of the snow mass is estimated on the basis of measurements of the resistance of snow to shear at the experimental site and the determination of the stability index of the snow cover in the zone of avalanche origin. Based on the analysis and statistical processing of avalanche observation materials and the meteorological conditions accompanying them, the probability of avalanche descending in points was estimated depending on the avalanche-forming factors.

The total score shows the degree of avalanche danger, with an increase in the sum, the likelihood of avalanches increases. The scoring of avalanche-forming factors begins with the accumulation of 7-8 cm of new snow at the observation site of the snow avalanche station. Then periodically, at certain intervals, the calculation is repeated. With a known rate of increase in snow thickness, the time before the onset of an avalanche danger is determined as the time to reach the critical snow height.

Empirical graphs of the relationship between avalanches and snowfall intensity, air temperature during snowfall, wind speed, and other factors are often used to predict avalanches.

Similar empirical graphs are constructed to identify the relationship between avalanche formation and a combination of wind speed and air temperature, wind speed of a given direction with an increase in air temperature, total blizzard transport and time, etc. hazards depending on the intensity of blizzard transfer (Practical allowance…, 1979). Forecasting is based on the data of snow storm observations, which are simultaneously monitored for the temperature distribution in the snow mass and for air temperature.

The validity of forecasts based on empirical dependencies is determined primarily by the amount and reliability of the meteorological information used and how clearly these dependencies characterize avalanche activity. To improve the reliability of forecasts, it is necessary that meteorological sites be located in the altitudinal zone of the highest frequency of avalanches; Particular attention should be paid to identifying the factors that most strongly affect avalanche formation in a given area, and to use them in a comprehensive manner for a probabilistic-statistical assessment of an avalanche situation. It is also important to timely analyze the processes of atmospheric circulation that precede avalanches from freshly fallen and blizzard snow. This allows you to increase the lead time of forecasts.

Prediction of avalanches caused by snow cover metamorphism.

To predict avalanches, it is necessary to take into account not only the current meteorological conditions, but also the characteristics of the entire previous part of the winter. It is especially important to know the temperature regime, stratigraphic structure, density and strength characteristics of snow in the zone of avalanche origin. It is dangerous to conduct direct observations of the snow cover in this zone; therefore, its characteristics are determined on the basis of remote observations, measurements on the experimental site and route snow measurements in avalanche-safe places near the avalanche origin zone.

The most dangerous are slopes with relatively shallow but significantly recrystallized snow cover.

A layer of deep hoarfrost at some point does not withstand the load of the snow slab on it, its sharp precipitation occurs. Due to the inhomogeneity of the settlement, the formation of cracks in the slab and the violation of its stability are possible. Particularly unfavorable conditions occur during heavy snowfall or during the deposition of snowstorms, when there is an additional load on a potentially unstable layer of deep frost.

It is dangerous when snowfall is comparatively high temperature air forms a fluffy cover, on which snowstorm snow is subsequently blown, forming a snow slab, where fluffy snow rapidly recrystallizes.

The heterogeneity of the snow mass, especially the presence of crusts or weak layers in it, creates the possibility of avalanches coming down at almost all stages of snow cover development. Therefore, special attention should be paid to such signs.

Snow recrystallization avalanches typically occur when there are potentially unstable single or multi-layer snow slabs on a slope. In some areas, they are in a locally unstable state and are kept on the slope due to edge forces. Violation of the stability of these slabs can be caused by various unforeseen reasons (collapse of a snow cornice, falling stone, passage or passage of a skier-snowboarder, uneven snow settling under the slab, etc.). It is almost impossible to predict the time of avalanches. Therefore, they confine themselves to assessing the probability of avalanches and determining the time when it is most expedient to produce artificial snow fall from avalanche-prone slopes.

In order to obtain quantitative characteristics of the snow cover in order to calculate its local stability on avalanche-prone slopes, the snow mass is drilled in pre-selected areas with a frequency of 10 days. At this time, the stratification of the snow mass, the layer density, the strength limits of the snow for shear at the contacts of the layers and for rupture are determined. If there are areas of snow slabs with a small margin of stability, then it becomes necessary to take into account the possibility of reducing the local stability index of the snow cover due to further recrystallization processes. If areas of local instability of the plates are revealed, then this indicates an avalanche danger.

To calculate changes in the local stability index between snow cover surveys, calculations of the intensity of recrystallization and probable changes in the strength properties of snow are carried out using information about meteorological conditions and snow cover temperature. In the same way, predictive estimates of the probable decrease in the stability of the snow cover are determined based on the forecast of meteorological conditions and the temperature regime of the snow mass.

Particular attention is paid to the forecast of avalanches in the event of an expected sharp drop in air temperature and during snowfall. A decrease in temperature causes additional tensile stresses in the snow slab in the places of kinks, which can cause the formation of a tear crack and a violation of the stability of the slab. Even a small snowfall can create an additional load sufficient for the brittle destruction of deep frost, breaking the continuity of snow plates and the formation of avalanches.

Wet snow avalanche forecast.

Mass avalanches from wet snow usually occur in the spring, when the snow begins to melt. Such avalanches are also possible in winter as a result of thaws and rain falling on the snow cover. The prediction of such avalanches is based on the analysis of observations of the temperature, heat exchange, and moisture content of the snow cover. The forecast problem is solved on the basis of the analysis of avalanche-forming factors and their critical values.

Based on the analysis of the meteorological situation during the periods of formation of avalanches from wet snow in the Western Tien Shan, the following provisions have been developed that are recommended to be used when developing forecasts (Practical allowance ..., 1979):

- Avalanches from freshly fallen wet snow are formed as a result of intense warming with the passage of air temperature through zero. Avalanches occur if during the snowfall preceding warming, the amount of solid precipitation was 10 mm or more.

- The daily forecast of avalanches from fresh snow is made up of two types: “avalanche-hazardous” and “non-avalanche-hazardous” - using empirical graphs of the relationship between avalanche formation and air temperature. The curves on these graphs determine the critical values ​​of daytime air temperature, which determine the onset of avalanche danger. The forecast is made in advance (12 hours in advance) and is updated according to the actual air temperature.

- A necessary condition for avalanches from old wet snow is a stable (more than 24 hours) transition of air temperature to positive values. The beginning of the avalanche danger period is determined by an empirical schedule, similar to the prediction of avalanches from freshly fallen sleet.

— The forecast of avalanches during the period of rainfall is carried out according to a graph that characterizes the relationship of avalanche formation with the night and maximum air temperature on the days of rainfall on the surface of the snow cover.

Under the conditions of the Inner Tien Shan, the relationship between the water content of the snow cover by the time the air temperature passes through 0° to positive values ​​and the sum of maximum daily values ​​for the period from its passage through 0° to the avalanches turned out to be the closest. For the forecast, a graph of the relationship between the time of avalanches and the intensity of solar radiation is also used.

In some areas, empirical graphs of the relationship between the time of the onset of wet avalanches and the intensity of the increase in air temperature are used; avalanche formation with snow adhesion, snow load and the sum of positive air temperatures and other empirical dependencies. Wet snow avalanche forecasting methods require further improvement.

According to materials - Avalanche science / K.F. Voitkovsky - M., MSU publishing house, 1989

The main purpose of developing an avalanche classification scheme is to establish uniform descriptive terms that can be used in the exchange of information on natural disasters, safety and control measures. Another purpose is to group avalanche events for statistical analysis, for example, to identify the relationship between avalanches and the factors that cause them - terrain, weather conditions, snow cover characteristics. It is also necessary to develop decisions on planning and implementation of protective measures.

Currently, international morphological and genetic classifications are used to describe and systematize the characteristics of avalanches and to predict avalanche danger.

The International Morphological Classification of Avalanches allows the transmission of information about avalanches in coded form, where the symbols for the criteria are given in the form of: capital letters (A, B, C, D, E, F, G, H), and the symbols for characteristics - in the form of numbers. In addition to numeric characters (1-5), it is proposed to use numbers: 0 - when there is no information about the characteristic, 7 or 8 - for mixed characteristics and 9 - to refer to a special note. For example, the code AZ B2 C1 D9 E1 F4 G1 H4 indicates that the avalanche was formed from a soft snow slab as a result of a break in the new snow cover, an avalanche of dry snow moved along the flume and formed an air wave (9 refers to a special remark clarifying the characteristics of the movement path avalanches), avalanche deposits are finely lumpy, dry, containing tree branches.

Genetic classification

Genetic classification relates avalanche phenomena to the conditions in which they are formed, for example, the shape of the slope, the weather, and the properties of the snow cover. Several genetic classifications have been proposed, but all of them are unsatisfactory, since the process of avalanche formation is so complex that it does not allow attributing the cause of formation to one or two factors.

Size classification

Avalanches can be classified according to their size (mass or volume of moving snow) or their destructive power. The following is a conditional classification scheme - five gradations of destructive action of avalanches (this scheme is widely used in the western part of Canada):

1) a small amount of snow that cannot harm a person;

2) can harm a person;

3) can cause damage to buildings, cars, break several trees;

4) can destroy large vehicles, forests on an area of ​​up to 4 thousand km2;

5) an unusual, catastrophic phenomenon - the destruction of settlements and the destruction of forests over a vast territory is possible.

Definition of avalanche danger

The information used to make decisions about the choice of a safe site for the construction of roads, buildings, ski slopes, as well as the choice of methods for controlling avalanches, comes from determining the location and size of avalanches, the frequency of avalanches and the assessment of potential damage. Avalanche collections can be recognized by the features of the relief (slope, flumes, characteristic points of origin), vegetation,. as well as on snow deposited by an avalanche. In the heavily forested mountains of southern British Columbia and Alberta, avalanche flows can be identified by examining the age and species of trees on different parts of the slopes. Features of the relief and vegetation can be identified on aerial photographs, however, ground studies are also necessary for clarification. The height of the trees must be accurately estimated, and the possible nature of the movement of avalanches should be taken into account. It should be kept in mind that not only avalanches affect the growth of trees, but also fires, mudflows, logging, soils, solar radiation and wind. Estimating the frequency of descent, the type and size of avalanches is very difficult; The most reliable method is the use of long-term data. The data show that on average every 12 to 20 years there is a winter or several consecutive winters with catastrophic avalanches. Often, the observation period may not be long enough and not include winters with maximum amounts of snowfall; in this case, historical data should be supported by data on tree age and damage, as well as analysis of climate data. The most important factor in planning the placement of structures outside the reach of avalanches is the maximum range of the release of avalanche material. In forested areas, very large avalanche deposits are often deciphered due to the presence of clear boundaries between trees of different ages and different species. These boundaries are best identified when comparative analysis old and new aerial photographs. The historical approach to the methodology for assessing the place of deposition, frequency and maximum range of avalanches is considered in the work.

snow avalanches- one of the natural natural phenomena capable of causing loss of life and significant damage. Among other dangers, avalanches are distinguished by the fact that human activity can become the cause of their collapse. Ill-conceived nature management in mountainous regions (cutting down forests on slopes, placing objects in open, avalanche-prone areas), access to snow-covered slopes of people, shaking the snow mass from equipment lead to increased avalanche activity and are accompanied by casualties and material damage.

The facts of the death of people in avalanches have been known since ancient times - in the works of Strabo and his contemporary Livy, accidents in the Alps and the Caucasus are described. The largest avalanche disasters are associated with military operations in the mountains - the crossing of the troops of Hannibal and Suvorov through the Alps, the war between Italy and Austria in 1915-1918. AT Peaceful time Avalanches that took on the character of a natural disaster occurred in 1920 and 1945. in Tajikistan, in 1951 in Switzerland, in 1954 in Switzerland and Austria, in 1987 in the USSR (Georgia), in 1999 in the Alpine countries. Only in Switzerland in 1999 the damage from avalanches exceeded 600 million Swiss francs. On the territory of the Russian Federation, cases of mass deaths in avalanches and significant destruction have been repeatedly noted. The most famous are the tragic events of December 5, 1936 in the Khibiny, when the village of Kukisvumchorr was destroyed by two successive avalanches. Limited information about catastrophic avalanches is contained in the USSR Avalanche Cadastre .

Cases of a one-time mass death of people are confined to avalanches on settlements, individual structures and vehicles. Significant destruction occurs most often during periods of mass avalanche formation, when a large number of avalanche sources are triggered over a large area over a short period of time.

In the 40-60 years, avalanches most often overtook their victims in buildings and on roads. Modern research statistics of deaths in avalanches show that the bulk of the dead are people who freely move within avalanche-prone areas - lovers of "untrodden paths". In the US, these are snowmobilists (35%), skiers (25%) and climbers (23%); in Canada, skiers (43%), snowmobilists (20%), climbers (14%): in Switzerland, skiers and climbers (88%). Most of the tragedies are provoked by the victims themselves. And only in the winter of 1998-1999. the balance has changed - 122 victims of avalanche disasters in the world (63% of the total number of victims) at the time of the collapse of avalanches were indoors and on the road. In Russia in recent years, accidents have been associated with moving through avalanche-prone areas - the death of climbers (North Caucasus), tourists (North Caucasus, Khibiny), skiers (North Caucasus), border guards (North Caucasus), vehicle passengers (Transcaucasian transport highway). Schoolchildren in the neighborhood are tragically regularly avalanched settlements. The size of avalanches is not critical to possible damage. The statistics of the victims claims that almost half of them die under small avalanches that travel no more than 200 meters.

Snow avalanche on a train running at this time

The consequences of an avalanche on the railway line

Thus, the main tasks of anti-avalanche measures are determined: protection from individual avalanche sources that threaten specific economic objects and prevention of people moving through economically undeveloped territories, where any mountain slope can pose a threat, into avalanches.

52 degrees (slope under the eaves). At steepness above 45 degrees, the risk of an avalanche decreases. Avalanche steepness - from 30 to 45 degrees. Most avalanches descend on slopes of 38 degrees. When the slope is less than 26 degrees, the probability of an avalanche is reduced. An angle of 45 degrees is easy to determine using two ice axes of the same length. Also 26 degrees is a ratio of about 1 to 0.5.

The warning reads: Beware of avalanches!

The need to organize anti-avalanche protection is determined by the scale of the phenomenon: the area of ​​avalanche-prone territories in the Russian Federation is 3077.8 thousand square kilometers. (18% of the total area of ​​the country), and another 829.4 thousand square kilometers. are classified as potentially avalanche-prone. In total, avalanche-prone areas on Earth occupy about 6% of the land area - 9253 thousand square kilometers. .

The avalanche danger forecast is part of a set of measures aimed at protecting the population and economic facilities from avalanches in mountainous areas. The definition adopted in glaciology of "avalanche forecast" (avalanche danger forecast) implies the prediction of the period of avalanche danger, the time and scale of avalanches . The use of the forecast to ensure life safety is determined by certain conditions and requires the creation of an information and methodological base.

Organization of anti-avalanche measures

The cardinal solution to prevent damage from avalanches is to prohibit the construction and placement of people in avalanche-prone areas. For certain reasons, this option is not always acceptable. A whole range of anti-avalanche measures has been developed and applied with varying degrees of success. Identification of avalanche-prone territories and determination of the parameters of the phenomenon, organization of an avalanche time forecast service, construction of protective structures, preventive avalanche release - these actions are aimed at preventing damage from avalanches. The nature of their influence on the process of avalanche formation is different. Engineering structures of various types prevent the formation of avalanches; preventive descent and some types of protective structures provide a controlled descent of avalanches (collapse time, size, direction of movement and range of release); survey work and forecasting the timing of avalanches contribute to the organization economic activity in avalanche-prone areas and preventing people from entering dangerous territories at a certain point in time. The greatest efficiency is achieved, as a rule, with a combination of various anti-avalanche measures.

An important factor in the choice of protective equipment is their cost. Engineering structures providing high reliability require significant material costs. For example, in Switzerland, from 1952 to 1998, about 1.2 billion Swiss francs were invested in the construction of anti-avalanche facilities. The cost of survey work and the forecast of the descent time are much lower. Thus, the budget of the avalanche center in Gallatin (Gallatin National Forest Avalanche Center, USA) in the 1998/99 season was $89,600 , and the maintenance of a similar unit in La Sala (La Sal Avalanche Forecast Center, USA) cost much less - about $ 17,000.

Comparison of the cost of anti-avalanche measures, carried out in the USSR in the 80s, gave the following results:

- forecast and preventive descent of avalanches, 1 km 2 of avalanche-active slopes per year - 10-20 thousand rubles;

- building slopes with reinforced concrete shields, 1 km 2 of avalanche-active slopes - 15,000-45,000 thousand rubles;

- compilation of maps of avalanche danger of different scales, the cost per 1 km 2 of avalanche-active slopes is 0.00015 -0.03 thousand rubles.

In the 1980s, the peak period of avalanche research in the USSR, the collection and processing of avalanche information in Russia was carried out by about 40 subdivisions of the State Committee for Hydrometeorology. The oldest organization in Russia engaged in snow avalanche research, the Avalanche Protection Department of the Apatit p/o (now the Avalanche Safety Center), carried out snow avalanche support on the territory of the Khibiny mountain range. The study of the distribution of snow cover in avalanche centers, the physical and mechanical properties of snow, and observations of avalanches that came down were carried out in areas of intensive economic development - along highways and railways, in mountain resorts, mining enterprises. Stations were organized to collect information, where constant observations of the snow and meteorological situation were carried out. With a certain frequency, avalanche patrol routes, overflights of avalanche-prone areas, and expeditions to avalanche-prone areas were carried out on vehicles.

(avalanche circle) - Avalanche Hazard - Low, Moderate, Severe, High, Very High

(Terrain + avalanche ring) - areas of high avalanche danger marked on the map. Although some sections of the gully do not pose a high avalanche risk, there are snow layers on its upper slopes that are under load. Any avalanche will go down the ravine. Therefore, traverses at its foot are not the best idea. Also, even if your route does not present an avalanche danger - how about a descent, is it just as safe?

The task of the snow avalanche divisions was to provide the population, governing bodies, organizations and enterprises of the regions whose territory is subject to the impact of snow avalanches with a forecast of avalanche danger. For the production of forecasts, observational data from a network of meteorological and aerological stations of the territorial departments of the hydrometeorological service were used. The work of the avalanche forecast service, as well as the entire hydrometeorological service, was based on the territorial-administrative principle. Figure 1, as an example of the organization of anti-avalanche work, shows a scheme of snow avalanche maintenance of the territory of the central regions of the Magadan region by units of the Kolyma Territorial Administration of Hydrometeorology and Environmental Control in the 80s.

The methodological center for conducting snow avalanche observations and organizing a service for the temporary forecast of avalanche danger on the territory of the USSR was the Central Asian Research Institute. V.A. Bugaev (SANIGMI) in Tashkent. A variety of avalanche information from all over the country flocked here, and annual reports from avalanche stations were received. SANIGMI developed the theoretical foundations for avalanche hazard forecasting and applied forecasting methods for various avalanche-prone areas of the USSR (often in collaboration with employees of local avalanche departments). The problem laboratory of snow avalanches and mudflows of Moscow State University served as a methodological center for the development of methods for assessing avalanche danger and its mapping. Specialists of Moscow State University have developed a specialized methodology for assessing avalanche danger and recommendations for serving in border avalanche-prone mountainous areas, and organized observations of avalanches. Snow avalanche research was also carried out by research and production organizations of the Ministry of Railways, Gosstroy and other departments.

The activities of organizations that carried out snow avalanche work were regulated by various governing documents. .

Snow avalanche research is carried out in many countries of the world. In some of them, data collection is carried out according to the network principle - the organization of the issuance of the National Avalanche Bulletin of Switzerland provides for the daily collection of data from 80 observers and 61 automatic stations (Fig. 2) . In the United States, there are 12 avalanche centers in the Forest Service alone (Fig. 3).

Abroad, the most popular manual for organizing snow avalanche operations are various editions of the Avalanche Handbook, specialized manuals have been developed.

Avalanche factors

Many years of experience in avalanche research has made it possible to identify certain patterns in the process of avalanche formation, to identify the leading factors in the collapse of avalanches, and to evaluate the parameters of the phenomenon. The collapse of avalanches occurs when the stability of the snow layer on the slope is disturbed, caused by the influence of external factors and processes inside the snow mass, occurring under the influence of external factors. Avalanches can occur on slopes with an inclination angle of 15 degrees and with a snow cover thickness of 15 cm. However, such cases are extremely rare. In the USSR, to identify areas where avalanches are possible, when compiling maps of medium and small scales, their boundaries were drawn along isolines of a snow cover thickness of 30 cm, and 70 cm isolines limited areas where avalanches form frequently and pose a significant danger. The most favorable for avalanche formation are recognized slopes, the angle of inclination of which is 25-40 o. Detailed large-scale studies using field observations and calculations, the study of geomorphological, geobotanical, soil and hydrological features in various regions make it possible to identify areas where the formation, movement and stop of avalanches occur.

In the process of studying avalanche collapse, the leading factors common for various mountain regions were identified and the nature of their impact on avalanche formation was determined (Table 1).

Table 1

Classification of avalanche-forming factors:

Factors	Impact on avalanches
A. Fixed Factors
1. Conditions of the underlying surface
1.1. Relative height, general topographical situation:	Determine the depth of dissection (the height of the fall of avalanches) and snow cover depending on the latitude of the place and the absolute height and orientation of the ridges
zone of ridges and high plateaus	Strong influence of wind on snow distribution, snow eaves, local avalanches from snow boards
zone between the ridges and the upper forest line	Blizzard snow accumulation, extensive area of ​​avalanche formation from snow boards
zone below the upper forest line	Reducing the influence of wind on the redistribution of snow, a decrease in the number of avalanches from hard boards, the prevalence of avalanches from soft boards
1.2. slope steepness	Determines the critical snow height
> 35o	Loose snow avalanches often form
>25o	Avalanches often form from snow boards
> 15o	Snow flow, lower limit of avalanche formation
< 20 o	The flow of snow, the deposition of avalanche snow. Possible occurrence of avalanches from water-saturated snow descending from slopes of very low steepness
1.3. Slope Orientation:	Affects snowiness, types of avalanches
in relation to the sun	On shaded slopes, an increase in avalanches from snow boards, on sunny slopes, an increase in the number of wet avalanches (with equal snow reserves)
in relation to the wind	On the leeward slopes, increased snow deposition, an increase in the number of avalanches from snow boards, on the windward slopes, the opposite effect
1.4. Surface configuration	Affects snow content, types of avalanches, critical snow height
flat slope	Non-canalized avalanches (wasps) from snow boards and loose snow
trays, funnels, carts	Places of snow concentration, canalized (chute) avalanches mainly from snow boards
Changes in the steepness of the slope along the longitudinal profile	On convex slopes, there is often a line of avalanche separation from snow boards, on steep slopes - points of emergence of loose avalanches, a significant effect on the critical snow height, jumping avalanches
Ledges in relief	Under them, avalanches of loose snow often occur.
1.5. Surface roughness	Affects critical snow thickness
Smooth surface	Small critical thickness, surface layer avalanches
Protruding obstacles (rocks, transverse ridges)	Large critical thickness, full depth avalanches
Vegetation	Grass - contributes to the breakdown of snow, avalanches of full depth; bushes - until they are completely covered with snow, they prevent avalanches from coming down; forest - if dense enough, it prevents the emergence of avalanches
B. Variables
2. Current weather (up to 5 days ago)
2.1. Snowfalls:	Increasing load. Increasing the mass of unstable material.
Type of new snow	Fluffy snow - loose avalanchesCohesive snow - avalanches from snow boards
Daily snow growth	Increase in snow instability with increasing snow cover thickness. Breakaway is possible in both new and old snow.
Snowfall intensity	Progressive instability at higher intensity, increased number of fresh snow avalanches, increased risk of avalanches on gentle slopes
2.2. Rains	Promotes the descent of wet loose or soft reservoir avalanches; possible occurrence of water-snow flows and snow-soil landslides
2.3. Winds	Create local snow overload on slopes, form snow boards and unstable stratigraphy
Direction	Increased risk of formation of formation avalanches on leeward slopes; cornice formation
Speed ​​and Duration	With their increase, the probability of local collapse of reservoir avalanches increases.
2.4. Thermal conditions	Ambiguous influence on the strength of snow and stresses inside the snow mass. Both a decrease and an increase in temperature can lead to instability
Snow temperature and free water content	Increasing the temperature to the melting point results in free water in the snow, which can cause it to become unstable.
Air temperature	The same effect for slopes of all exposures, strong cooling contributes to the development of instability due to gradient metamorphism
Solar radiation	On the slopes of solar exposure, the development of instability due to the development of radiation thaws
thermal radiation	Cooling of the snow surface at night and in the shade, which is significant in a cloudless sky, contributes to the formation of surface and deep frost.
3. Conditions in the old snow cover (integral influence of previous weather conditions and weather for the entire winter season)
3.1. Total snow height	Not a major avalanche hazard. Smoothing the roughness of the slope surface. Affects the mass of an avalanche descending on the ground. Influences the process of gradient metamorphism.
3.2. Stratigraphy	The stability of the thickness on the slope is controlled by the presence of weakened layers, taking into account stresses
Old surface layers	Condition - looseness (surface frost), brittleness, roughness - are important during subsequent snowfalls
The internal structure of the snow cover	Complex structure, weakened layers, ice crusts lead to the development of instability

It should be noted that the process of avalanche formation is influenced not only by the above factors, but also by their combination. Already during the deposition of snow on the earth's surface, the influence of many processes is carried out. The shape and size of snow crystals, the nature of occurrence and the density of the surface layer are determined by air temperature, wind direction and speed, the shape and parameters of the underlying surface. The predominance of one or another type of metamorphism in the snow mass, the nature of its evolution are a function of the action of a wide variety of factors.

On the basis of long-term observations, quantitative indicators of meteorological factors of avalanches (precipitation intensity, snow cover growth, wind speed, etc.) and characteristics of the avalanche regime for individual mountain regions have been identified, which make it possible to assume with a certain degree of probability the possibility of snow avalanches, the relief is assessed as avalanche factor. The simplest forecasting methods are based on comparing the current and predicted values ​​of snow and meteorological characteristics with critical values .

An analysis of the factors leading to the collapse of avalanches made it possible to identify the genetic types of avalanches and classify them. The need for genetic classification for avalanche forecasting is explained by the fact that the forecaster must clearly understand what exactly he is going to predict and what factors should be paid attention to in the first place. This may be taking into account external factors that determine the occurrence of additional loads and the presence of moisture in the snow cover. , separation according to the action of external and internal processes in the snow cover , typification of the structure of falling snow and the nature of its separation , the influence of external factors on the balance of forces in the snow cover lying on the slope.

Schematic photo of an avalanche on a ski slope

The development of a unique genetic classification is complicated, among other things, by the fact that avalanches can be caused by a combination of a number of factors. For example, in many regions of Russia, the collapse of avalanches, conventionally classified as avalanches of freshly fallen or snowstorm snow, occurs due to the destruction of the deep layer of snow cover, in which for a long time before snowfall or snowstorm there was a process of loosening, that is, according to some signs, they can also be attributed to avalanches of long-term development. An analysis of the available methods shows that the number of predicted types of avalanches is less than that proposed by most researchers. A simplified scheme for differentiating avalanches was proposed by the creators of " methodological recommendations according to the forecast of snow avalanches in the USSR ":

• freshly fallen snow;
• snowstorm;
• old snow;
• others.

The uncertainty of the last group is explained by the mixed genesis of many avalanches. In the future, when specifying the genetic type of avalanches, the definition specified by the developer of the forecast methodology will be used.

It should be noted that many foreign researchers do not pay attention to special attention classification of avalanches according to their genesis, with emphasis on the study of the structure of the falling snow layer. For example, the terms soft board or hard board are widely used. .

Avalanche Forecast

Avalanche forecast for general view includes an indication of the place and time of the avalanches.

At the initial stage of studying avalanches in a certain area, it is necessary to identify the places of possible avalanches, calculate their parameters, and determine the avalanche regime. For these purposes, materials of snow avalanche observations, indirect signs of avalanche danger, statistical dependencies, mathematical models are used, archives are studied and surveys of local residents are conducted. Based on the received and calculated data, avalanche hazard maps are compiled. The research result is defined as spatial forecast avalanche danger - forecast of avalanche "climate" . In terms of area coverage, it can be local (for an individual avalanche source or a group of them) and background (for a mountainous region or their combination). Accordingly, large-scale maps are used to represent the local forecast, and medium- and small-scale maps are used for the background forecast.

Large-scale maps may contain the following information: contours of snow collections indicating the places of avalanche separation and transit zones, the boundaries of the distribution of avalanches of various probability, isolines of dynamic characteristics, the boundaries of the propagation of an air wave, the frequency of avalanches.

In Western Europe, the form of information presentation on large-scale maps often has an applied character - different color shading characterizes the frequency and strength of an avalanche impact and determines the possible use of a given territory: from a complete prohibition of ground construction to permission for construction using protective structures and the absence of any restrictions.

It should be noted that during the winter period of 1998/99. many avalanches in the Alpine region entered the white (calculated as safe) zones and caused significant damage. An example is the largest avalanche disaster in Austria in the post-war period on February 23 in Galtür, when an avalanche descended from a slope that was considered safe, claimed the lives of 31 people. The conclusion about safety was based on the absence of information about avalanches from this slope in historical annals. These events indicate the imperfection of methods for assessing avalanche danger - spatial forecast.

On an average scale, a characteristic of avalanche-prone slopes is given - the frequency of avalanches, their volumes, and genetic types. Small-scale maps serve to identify areas in which special surveys are necessary in the design of building structures and other survey work. They contain an estimate of the degree of avalanche activity ( tab. 2 ).

table 2

Gradations of avalanche activity:

Maps can show an assessment of possible damage from avalanches, recommendations for choosing anti-avalanche measures with an assessment of their effectiveness.

Temporal The aspect of avalanche danger forecasting involves determining the possibility of avalanches in a given area within a specified period of time. Three types of avalanche forecasts are distinguished by the area of ​​the territory covered:

1. background small-scale, compiled for a mountain system or individual river basins with an area of ​​at least 250 km 2;
2. background large-scale for the territory of a mountain basin, usually with an area of ​​25-30 km 2 or large avalanches;
3. detailed large-scale, compiled for a single avalanche or avalanche slope

The classification of forecasts given in the scientific literature into short-, medium- and long-term ones does not use fixed time intervals for such their separation. An analysis of works on forecasting avalanche danger shows that, in practice, a forecast can be made for a day, 48 hours, 72 hours, for the winter season, for a long period of time.

Avalanche danger forecasts are created using methods specially developed for a region or a separate source that determine the algorithm for identifying avalanche danger. A number of methods provide for the forecast of an avalanche period - a period of time during which the effect of the avalanche formation factor will persist. As a rule, this approach is used to predict avalanches during snowfalls and blizzards. Avalanches are predicted from the moment critical conditions are reached until the end of the snowfall (blizzard) and for a period of one to two days after they end - as long as the instability of the snow cover remains. Avalanche forecasts are consultative in nature, as the forecaster must build his forecast based on assumptions such as “if the intensity of warming continues for several days”, etc. At the same time, periodical forecasts have a significantly higher accuracy compared to daily forecasts. However, the uncertainty of the time of avalanches that accompanies this type of forecast makes its use inconvenient for the consumer.

A number of prognostic centers make a forecast for several days, indicating the degree of danger for each day.

To prevent damage or unnecessary costs for organizing anti-avalanche measures, the forecast may be updated during the validity period. For example, the Swiss national avalanche bulletin is published daily at 17:00, in case of significant changes in snow and meteorological conditions, a new text of the bulletin is published at 10:00 am.

The lead time (the time between the compilation of the forecast to the start of its action) of the forecast, which is incorporated in many forecasting methods, is zero. In practice, this means a statement of the fact that critical conditions for avalanches have been reached. The main reasons for this situation lie in the transience of the occurrence of an avalanche situation (from several hours to a day), the constant change in meteorological conditions, the impossibility of continuous and widespread collection of the necessary information. A very significant point that determines both the quality of the forecast and its lead time is the unique spatial and temporal variability of the structure and properties of the snow cover. The diagnostic scheme is converted into a predictive one when the inertial forecast of meteorological elements is used in calculations. The limitations of the lead time when the methodology is oriented towards the use of meteorological forecasting are supplemented by the lack of accurate methods for quantitative precipitation forecasting and the interval form of forecasting a number of meteorological elements. To achieve greater lead times and improve the quality of the forecast, avalanche specialists often create their own methods for forecasting the meteorological characteristics necessary for their work. As an example, we can cite a forecast of precipitation of more than 15 mm / day for the Zailiysky Alatau.

In separate forecasting methods , using information about the state of the snow cover in the region of the avalanche separation zone, the time of avalanche collapse is calculated.

As new snow and meteorological information becomes available, the forecast is subject to revision.

The subject of the forecast of a number of methods is the quantitative characteristics of avalanches - volume, range of release, number of avalanches . For the background forecast, the places of descent are specified - specific avalanche centers, altitude intervals of avalanches and slopes of a certain exposure.

The subject of the forecast may be a massive avalanche, when avalanches occur in more than 1/3 of the avalanche centers of the territory for which the forecast is made.

Methods for long-term forecasting of avalanche danger take into account possible climate changes. The objects of forecasting are the duration of the avalanche period, the number of days with avalanche snowfalls and a number of avalanche-indicating characteristics - the thickness of the snow cover, the number of days with a negative average daily air temperature.

The avalanche danger forecast may have an alternative and probabilistic character. With an alternative forecast, two formulations are possible: “avalanche risk” and “non-avalanche risk”. In the USSR, this approach to assessing avalanche danger was used in most cases. The thin point of such forecasts is the avalanches that do not threaten the population and economic facilities. . At the same time, according to a non-avalanche situation, a situation is considered when there is no avalanche coming down, or slight movements of snow up to 10 m 3 in volume, which do not pose a danger to people and economic facilities. An alternative forecast provides for the collapse of spontaneous avalanches. The forecast is considered justified if at least one avalanche came down (except for the cases of the forecast of mass avalanches). The possibility of artificial collapse of avalanches can be negotiated separately.

The probability of avalanches can be estimated as a percentage, which is used extremely rarely due to the inconvenience of interpreting the forecast by the user, and on a certain scale. The concept of the European Avalanche Hazard Scale was developed in 1985. . In 1993, after extensive discussion, the scale was adopted for use in practice by services avalanche forecast a number of Western European countries (Table 3). The degree of danger is assessed in five progressively increasing levels, which are described in terms of the stability of the snow cover on the mountain slopes, the likelihood of avalanches and their volumes, and the nature of the impact on life in the mountains. The condition of the snow (its stability) is assessed in relation to possible additional loads.

Table 3

European avalanche scale:

	Degree of avalanche danger	Snow cover stability	Probability of avalanches	Recommendations for land transport routes and settlements	Recommendations for people outside avalanche-protected areas
1	Minor	The snow cover is well fixed on the mountain slopes and is stable	Collapse is only possible with very significant additional loads on some very steep slopes. Only snow movements can occur spontaneously	No threat	Safe conditions
2	Moderate	Snow cover on steep slopes is moderately fixed, on other slopes it is good.	Collapse is possible with significant additional loads, primarily on the specified slopes, spontaneous collapse of avalanches is unlikely	Mostly favorable conditions	Careful choice of travel path, especially on the indicated steep slopes of the indicated exposure and altitude levels
3	Significant	Snow cover fixed on steep slopes fixed either moderately or weakly	Avalanches are possible with a slight additional load on these slopes. The collapse of individual medium-sized and less often large-sized avalanches and	Unprotected areas are dangerous. Precautions needed	Relatively unfavorable conditions. It is necessary to avoid movement in the area of ​​​​the indicated slopes.
4	Big	Snow cover is loosely fixed on most slopes	Collapse is possible on most slopes with little additional load	Most unprotected areas are dangerous. It is recommended to take precautions	Unfavourable conditions. It takes a lot of experience to get around. Restriction of movement on slopes.
5	Very large (exceptional)	Snow cover is unstable	Numerous spontaneous avalanches are expected to collapse on any slopes	Big threat. Precautions required	Very unfavorable conditions. Refusal to move recommended

Forecasts developed in accordance with the European avalanche danger scale always, even at a low degree of avalanche danger, provide for the possibility of the collapse of artificial avalanches. In the USA and Canada, when predicting avalanche danger, their own developments are used - the American avalanche danger scale has 4 levels, the Canadian one has five. The scale adopted by American experts takes into account the possibility of the formation of only natural avalanches. The undoubted advantage of all approaches is the presence of recommendations for the population in avalanche areas (French and Italian forecast services do not include such recommendations in the forecast formulation).

An unresolved issue in the probabilistic approach to assessing avalanche danger is the impossibility of accurately checking the correctness of the forecast. This is hindered by qualitative indicators in assessing the number of avalanches and their volumes.

Separately, it should be said that, unlike most other dangerous weather phenomena, an unjustified forecast of an avalanche danger does not mean that an avalanche will not come down later!

The generally accepted form of presenting an avalanche forecast is an avalanche bulletin (Fig. 4). When a mass avalanche was expected, the prognostic centers of the USSR prepared storm warnings, which were communicated to consumers in an emergency way. In a number of countries, the avalanche bulletin is supplemented by an avalanche hazard map of the territory. Maps and expert opinions (reports) present a forecast of avalanche danger for a long period (Fig. 5).

A large avalanche at Mt. Timpanogos, Wasatch Range, Utah

The correctness of the forecast is checked by observations at stationary posts, in routes along roads and railways, during air overflights of the territory, according to reports from individual citizens and organizations, according to the results of a survey of the population of avalanche-prone areas.

Methodological support for the forecast of avalanche danger

Put on a scientific basis, regular observations of snow avalanches were started in the early 1930s in the USSR (Khibiny mountain range) and in Switzerland. The accumulated experience and data made it possible to start forecasting the avalanche danger of territories in a few years. Initially, forecasts were made on the intuition of researchers. The intuitive approach to assessing the possibility of avalanches has been maintained for quite a long period of time. For example, from the standpoint of inductive logic, an avalanche forecasting system was built in the USA and Canada. By the end of the 1930s, the first forecasting methods appeared. I.K.Zelenoy created and put into practice the methodology for forecasting avalanches during snowstorms. Subsequently, when avalanche observations covered many mountainous regions of various countries of the world, numerous methods were developed to help avalanche forecasters, using various methods for determining avalanche danger. Such techniques have been created for many mountainous regions of the country. However, by the end of the 1980s, less than half of the forecast methods mentioned in 63 had been tested and applied in practice. At this point, only the Sakhalin, Irkutsk and Kolyma departments of the hydrometeorological service and the avalanche protection shop of the Apatit plant have introduced predictive models into production. Since then, judging by the publications in the specialized literature, the situation has not improved much.

The reasons for this state are in various aspects of the activity and interaction of industrial and scientific organizations. In the literature on avalanche research, methods for forecasting avalanche danger, created in industrial and scientific and industrial organizations of the hydrometeorological service, which have received practical application after production tests, and theoretical studies of scientific organizations, most often not used in forecasting, have been published.

Methods for determining avalanche danger were created separately for the border territories of the USSR. Their use was carried out in the border troops of the country.

It should be noted that many experts are skeptical about the possibility of using the methodology developed for a particular mountain region in other regions. This is hindered by differences in climate, prevailing weather conditions, terrain, and the nature of the underlying surface of the slopes. In such cases, additional studies are carried out aimed at determining the boundaries of the application of the methodology, identifying new leading factors, etc. .

According to the practice adopted in the hydrometeorological service, newly created methods are checked on independent material, undergo production tests, and after that they are recommended (not recommended) for practical use. The term for the development of a methodology, including the collection, processing of information and production tests, is several years. Their assessments are taken to be the justification of forecasts, the warningness of the predicted phenomenon and the well-known criteria of A.M. Obukhov and N.A. Bagrov.

The main requirement for the quality of forecasts: the sum of the general justification and warning of the presence of the phenomenon in percent must be greater than the sum of the natural frequency of occurrence of cases with phenomena from 100%.

The final version of the forecast presented to the consumer is made by a specialist, using, in addition to the methods, his own experience, intuition and additional data that are not taken into account by the methods.

The main methodological principles of avalanche danger forecast are formulated:

• - the principle of proportionality between the territory covered by the forecast and its lead time, for example, the background forecast should have a lead time not less than the actual time for organizing anti-avalanche measures;
• — continuous monitoring of changes in the situation;
• — taking into account, when developing new forecasting methods, the prehistory of the development of the snow and meteorological situation in time;
• - detailed avalanche warning has a limit, which is provided by the ability to collect individual information in each avalanche source, in addition to background data.

The creation of a methodology that will be used to make an avalanche hazard forecast includes several stages:

• creating a training sample,
• choice of predictors,
• their transformation,
• choice of forecasting method,
• assessment of the reliability of recognition (justification) of the forecast.

Selection of predictors

The quality of the forecast is ensured by the choice of a set and the optimal number of predictors - indicators that determine the formation of avalanches in a particular area and at a fixed point in time. These may include (Table 1) snow cover characteristics, indices of atmospheric processes, values ​​of meteorological and aerological elements, and relief parameters. In the practice of avalanche hazard forecasting, measured, normalized (if different from the normal distribution) and calculated values ​​(precipitation intensity, air temperature change, etc.) are used, as well as generalized indicators that take into account several initial variables and describe a certain process (the product of wind speed by the duration of its action, characterizing the amount of snow swept).

Thus, at the initial stage of developing a forecast methodology, the task is to select the most informative features from the set of features that provide the required statistical reliability of the methodology and forecast accuracy. The information content of a single feature is understood as a measure of the amount of information contained in it, relative to another. At the same time, according to a number of researchers, for the analysis (in particular statistical) of most avalanche situations, there is no need to form bulky data arrays with a large number avalanche signs. Increasing the volume of data usually does not provide a gain in the lead time and accuracy of forecasts.

The selection of features (predictors) can be carried out on the basis of physical considerations and methods of mathematical statistics. The choice of predictors for forecasting methods should take into account the area of ​​the territory for which the forecast is made and the variability within its values.

As an indicator of the information content of predictors used in the forecast of avalanche danger, the following are used:

• - double t- Student's criterion;
• is the Mahalanobis distance;
• is the Fisher separability index.

Correlation analysis of pairwise independent predictors makes it possible to eliminate interdependent values ​​and thereby reduce the number of predictors. In the work, signs were taken as independent, the correlation coefficients of which are less than 0.6 modulo. Principal component analysis, used as a way to reduce factors, allows the use of interdependent predictors. The most commonly used rotation is the varimax method (which maximizes the variance of the original space of variables).

The order of the signs according to the degree of informativeness is determined using the procedure of "sifting » . When compiling an alternative forecast, a classification is made into two classes: a class with the presence of avalanches and a class with no avalanches. Initially, the composition of the general predictor vector includes all features that determine the physical model of the phenomenon under consideration and take into account its features. The predictor providing the maximum value of the Fisher separability index is selected from the total number of predictors, then the value for this predictor is calculated in pair with each of the remaining predictors, and so on. The procedure continues until the growth of the separability index stops with the addition of each next predictor. Thus, a group of predictors is determined that most fully describes the conditions of avalanche formation.

The assessment of the nature of the influence of each feature separately is made by comparing its average value in two classes. To compare the degree of information content of the features, the Mahalanobis distance is calculated. And to check the significance of the difference in the average values ​​of the parameters in each of the classes, a double t-Student's criterion. The significance of the difference indicates the isolation of the classes and the possibility of a good classification.

It has been established, for example, that when forecasting using discriminant analysis, the optimal ratio between the number of features and the length of a series of observations in a class with a phenomenon should be no more than 1/10. Usually their number lies in the range of 5-10.

When choosing predictors, one can follow the rule formulated in the work using the method of principal components:

• the first principal component can be defined (expressed) as a "force effect" (load) on the snow layer;
• the second - as a "temperature background" of an avalanche;
• the third is "readiness of the snow mass to vanish".

Long-term studies and analysis of works to identify the leading factors of avalanche formation made it possible to identify the most significant predictors for avalanches of various genetic types (Table 4).

Table 4

Sets of the most significant predictors for avalanches of various genetic types:

Types of information	Genesis of avalanches
(options)	From fresh snow	From the blizzard snow	thermal loosening	Sublimation loosening
Air temperature	+	+	+	—
Snow thickness	+	(+)	+	(+)
Water equivalent of snow	(+)	—	(+)	(+)
Snow Density	(+)	(+)	(+)	(+)
Snow moisture	—	—	+	—
Snow temperature	—	—	+	(+)
Air humidity	(+)	—	—	—
Blizzard transfer	—	+	—	—
sunshine duration	—	—	(+)	—
Acoustic emission from snow	+	+	(+)	(+)
Wind speed	(+)	+	—	—
Avalanche times	+	+	+	(+)
Power of loose horizons	(+)	—	—	(+)
Crystal size	—	—	(+)	(+)
Atmosphere pressure	—	+	—	—

+ — sign is informative

(+) - informative conditionally

- uninformative

It has been established that predictors such as the increase in the height of fresh snow and/or the amount of precipitation are well recognized and can be universal for many mountainous regions when predicting avalanches from fresh snow. Snowstorms in different regions can also be predicted using a limited set of predictors. At the same time, wet avalanches, even within the same mountainous region, can have significantly different predictors.

Detailed forecast methods are based primarily on the use of data on snow cover in a particular source, while background methods are most often based on aerosynoptic and meteorological information.

Differentiation of avalanche conditions

The classification of avalanche formation conditions that precedes the forecast procedure, which is traditional for developments in the USSR, contributes, in the opinion of a number of authors, to an increase in its quality. Because many avalanche forecasting techniques are designed for avalanches of certain genetic types, this process allows you to compare the current situation with typical ones, assign it to a certain class and focus on the leading factors and the application of certain methods.

The selection of predictors for classifying the conditions of avalanche formation is carried out similarly to the selection for forecasting methods. To differentiate the conditions of avalanche formation, the following are used:

• - regression analysis;
• — discriminant analysis;
• — Principal component analysis.
• — method of pattern recognition;

The mechanism for attributing the situation to the occurrence of dry or wet avalanches is described in the work. At the first stage, the training sample of dry and wet avalanches was formed according to the genesis determined by the avalanche station. Next, the procedure for determining the information content of predictors, constructing a discriminant function, and determining the probability that each event belongs to a particular class were carried out.

The calculated principal components in the work made it possible to obtain the equations of the discriminant function, separating fresh snow avalanches into dry and wet ones with a justification of more than 90%. At the same time, the affiliation of wet avalanches with a separation along a line and from a point showed the correctness of identification, respectively, 84 and 63%, although the separation of dry avalanches was recognized with high reliability (91-95%).

A number of methods for forecasting avalanche danger contain conditions from the moment of occurrence of which their application begins. Thus, the date of the beginning of the avalanche season can be taken as the achievement of a snow cover thickness of 30 cm on the meteorological site. For the Tom River basin, the first avalanche danger forecast, compiled according to the proposed method, should be preceded by the accumulation of 100 mm of solid precipitation from the date of formation of stable snow cover, etc. When assessing the current situation, the technique can start working from the moment one of the parameters reaches a critical value. For example, for the river basin Kunerma semi-daily precipitation should reach 1 mm.

Method of direct (field) determination of avalanche danger

Regular snow avalanche observations include studying the stratigraphy of the snow mass, measuring the thickness of the snow cover, determining the physical and mechanical properties of snow - density, temporary shear and tear resistance, hardness, tensile strength, etc. Measurements are carried out in the immediate vicinity of avalanche sources in safe areas, having, as far as possible, parameters similar to those of avalanche-prone slopes (steepness, exposure).

The simplest statistical processing of observational data makes it possible to establish empirical relationships that allow, using the measurement results, to determine the possibility of avalanche collapse (Table 5). With the accumulation of materials, typical combined stratigraphic columns and diagrams of the distribution of strength characteristics along the vertical profile are built, by comparison with which the degree of avalanche danger is estimated and the type of expected avalanches is determined.

Table 5

Empirical dependencies for predicting avalanche danger based on sounding data with a cone probe:

Avalanche danger	Probe resistance R, kg	Clutch FROM»1.4R kg / dm 2	Strength ratio of adjacent layers
Severe (an avalanche may occur soon)	Less than 1.5	Less than 2	More than 4
Medium (an avalanche can occur when the snow cover is mechanically disturbed)	1,5-5	2-7	2,5-4
Low (almost no avalanche threat)	5-21	7-30	2,5-1,5
Missing	Over 21	Over 30	Less than 1.5

Avalanche services in many countries have developed systems for testing the stability of the snow mass. During the tests, weakened layers are identified and the force required to shift and subsidence of the snow layer on a specific mountain slope (in the avalanche focus) is estimated. At the same time, both quantitative and qualitative definitions are used for evaluation. The simplest actions using improvised means (shovel, skis) make it possible to determine the degree of avalanche danger on a mountain slope not only for specialists, but also for all those working and relaxing in the mountains. In a number of countries, mastering the tests is included in the mandatory training program for ski and mountaineering instructors. The increased attention to such tests is explained by its focus on ensuring the safety of those categories of people who make up the bulk of the victims of avalanche disasters.

Snow avalanche on the road

Avalanche in the mountains

The so-called "shovel test" (Shovel Shear Test) is performed on a block of snow cut out in the snow mass (Fig. 6.). The force required to tear off a cut block of snow, assessed qualitatively, is a subjective measure of snow stability. Based on the observations, conclusions are drawn about the degree of avalanche danger of the slopes. If the snow is very unstable, then a weak layer comes off immediately, as soon as all four faces of the block are cut out. If lift does not occur, it can be induced by pushing the block down the slope with a shovel.

In recent years, the "sliding block test" (Rutschblock Test), developed by specialists from the Swiss Institute for Snow and Avalanche Research and its modifications, has been used to test snow. Checking the snow cover on the slope is carried out by the skier on blocks cut out in the snow mass (Fig. 7). The skier performs 7 specific actions, positioning himself over a block of snow and moving along it, gradually increasing the load. Tests are carried out until the destruction of the block. The interpretation of the results obtained - the determination of the degree of avalanche danger - is carried out in accordance with the standards developed in a number of countries. In its simplest form, destruction in 1-3 actions means the unstable state of the snow layer on the slope, which will be broken under the action of the skier; at 4-5, a steady state is assumed, but an individual skier can cause an avalanche to collapse; 6-7 - an avalanche collapse by a skier is unlikely. The significant dimensions of the tested block (an order of magnitude closer to the real snow layer on the slope) favorably distinguish this test from most others.

Tests are performed with a certain frequency on different (exposure, steepness) slopes, which makes it possible to identify changes occurring in the snow mass and determine the direction of the metamorphism process.

While such tests often give quite nice results, it is important to understand that a single test cannot determine the stability of the entire slope. The results can vary dramatically depending on which part of the slope the test was performed on. The difficulties of using tests to assess avalanche danger are associated with the lack of consideration of the weight of the test skier, the subjective determination of the efforts made.

Due to their simplicity and rather high reliability, snow cover stability tests are widely used in practice to determine the degree of avalanche danger. The test results are taken into account both in local and background forecasts of avalanches by various methods.

Field observations are the most effective way to determine the possibility of avalanches of long development.

Deterministic method

The measured values ​​of the snow cover characteristics are used to calculate the stability of the snow cover on the slope.

In its simplest form, the stability coefficient for loose snow under the shear mechanism of avalanche formation can be calculated as follows:

f – coefficient of internal friction or friction of snow on the underlying surface,

a — the angle of inclination (steepness) of the slope.

If this ratio is significantly greater than one, there is no avalanche danger; when its value is equal to one, the snow cover is in a state of limit equilibrium i.e. can slide down the slope with a slight increase in load or decrease in restraining forces; if the stability coefficient is less than one, this indicates an unstable state of snow on the slopes.

Empirically, a number of equations have been obtained that make it possible, using field measurement data, to identify the critical values ​​for each layer of the thickness of the overlying snow layer, adhesion at the lower boundary of the layer, and to determine the limiting slope angle for these conditions. The inclusion of meteorological characteristics in the calculation makes it possible to determine the time of the onset of an avalanche danger (assuming that the current weather situation persists).

To speed up the calculation of critical values ​​and make a forecast, nomograms were constructed to assess the state of the snow cover in the field (Fig. 8).

The stability of the snow cover can be estimated from the results of calculating the distribution of mechanical stresses in it. Such a calculation for a snow cover with varying thickness and a significant spatial variation of parameters, lying on a mountain slope of an arbitrary configuration and held by a friction force that depends nonlinearly on the displacement of snow relative to the slope, is a three-dimensional and essentially nonlinear problem and involves a large amount of calculations. By introducing some conditions, the problem is most often reduced to a two-dimensional solution. Mathematical models for calculating the stability of snow on a slope, based on the analysis of the stress state of snow, can be used to predict avalanche danger, but are rarely used in practice. The reasons are the difficulty in obtaining characteristics of the state of snow in avalanche centers, significant errors in their measurement, as well as the impossibility of extrapolating the data obtained at one point to the entire surface of the avalanche source due to the significant variability of the structure and properties of snow.

At present, this direction of the forecast is being developed at the Avalanche Safety Center of JSC Apatit in Khibiny. The calculation based on the developed model determines the probability of exceeding the threshold value of the stress tensor in the snow cover in the avalanche source (Fig. 9) .

The deterministic approach is used to predict avalanches from a specific avalanche source.

The impossibility of carrying out direct measurements of the characteristics of the snow cover in the areas of avalanche separation stimulated the study of physical processes in the snow cover and the construction of models of its structure and evolution. The first such models used statistical relationships and took into account only individual factors - snow accumulation during a snowfall, blizzard snow transfer and wind speed, and the formation of a layer of deep frost. In 1983, the Center for Snow Research (CEN) in France began to develop a new program to study the development of snow cover. The deterministic model estimates the energy and morphological regimes of the snow mass. The simulation calculates the thermal conductivity of snow, moisture seepage, snowmelt, takes into account phase transformations within the snow mass and the most important processes of snow crystal metamorphism. The radiative and turbulent fluxes entering the surface of the snow cover and the geothermal flux from the underlying soil are taken into account. The result of the model operation is the calculated profile of the snow mass with the values ​​of temperature and density distributed over it; unstable layers are revealed. Testing the model in different areas of the French Alps gave satisfactory results, although there is an underestimation of the influence of the wind. . The model does not calculate the formation of surface frost and ice crust on the surface of the snow mass, which are important factors for the occurrence of avalanche danger.

Mathematical modeling of the processes of heat and mass transfer in the snow mass, taking into account its complex layered structure, has also been developed in our country. . At present, it is planned to test the theoretically developed model in the field in different mountainous regions.

Methods for remote monitoring of avalanche danger

Methods for remote monitoring of snow cover for predicting avalanche danger are poorly tested on mountain slopes and exist mainly in the form of theoretical developments. One of such methods is the registration of acoustic emission signals in the snow cover. It has been established that an increase in the average activity of acoustic emission corresponds to a decrease in the stability of the snow cover in the avalanche separation zone.

A method for assessing the stability of the snow cover, using information about the slow sliding of snow supplied by a special sensor, was developed at the High Mountain Geophysical Institute.

Pattern recognition methods

The essence of the pattern recognition method is as follows. An image is a description of any element as a representative of the corresponding class of images, which in turn is defined as a certain category that has a number of properties common to all its elements. With regard to avalanches, the image should be understood as a set of values ​​of a finite number n parameters characterizing the snow-meteorological situation. AT n— in dimensional space, the image is determined by the vector x=( x 1 , x 2 ,…, x n), where x i– parameter values. Obviously, for the purposes of forecasting avalanche danger, two classes of images are distinguished: the class of avalanche and non-avalanche situations. Further, in order to identify the unknown vector x, it is necessary to compare it with some standard of the corresponding class.

The pattern recognition group includes several methods that use the apparatus of mathematical statistics.

Synoptic (standard) method

Methods for the background forecast of avalanche danger using the synoptic method are based on the comparison of statistical information about avalanches with synoptic situations and related weather conditions. Cyclonic processes, invasions air masses cause precipitation, changes in wind direction and speed, air temperature - the leading factors of avalanche formation. Depending on the direction of movement, the depth of the cyclone and the duration of its action, the nature of the influence on different areas of the study area differs - the height of the terrain, the exposure and steepness of the slopes, the orientation and width of the mountain valleys provide a diverse reaction of the snow cover. At the same time, the action of certain processes does not contribute to the formation of avalanches and leads to stabilization of the snow cover on the slopes.

Typification of atmospheric processes for forecasts of avalanche danger is most often carried out in the direction of their movement (Fig. 10 - Typification of cyclones leading to the emergence of avalanches in the central regions of the Magadan region, along the trajectories of movement). When classifying atmospheric processes, it is given complex characteristic meteorological phenomena in the period of their influence.

Daily analysis of the synoptic situation in order to detect and identify various types of atmospheric processes makes it possible to make a background small-scale forecast of avalanche danger with a significant (24 hours or more) lead time.

Participation in the preparation of the forecast by an expert who has current avalanche information and knows the previous situation, makes it possible to refine the forecast (indicate possible places of descent) and achieve results satisfactory for the background regional forecast. The accuracy of forecasts made using the synoptic method reaches 65-70% . When forecasting for the period of avalanche danger, it rises to 80-90%. The quality of the forecast is affected by the fact that, in addition to errors in the identification of an avalanche situation associated with determining the state of snow, such methods also contain errors inherent in the aerosynoptic information itself.

Forecast methods based on the synoptic method are available for the Khibiny mountain range, the central regions of the Magadan region, the Elbrus region, and the Chukotka Peninsula. The synoptic conditions for the occurrence of avalanche danger for the border regions of Russia have been determined.

Consideration of macroprocesses, cyclonic activity, synoptic situations, as well as meteorological conditions of mass descent of especially large (low frequency) avalanches in various mountainous regions of the country made it possible to generalize the patterns and reveal the similarity of the conditions for the formation of especially large avalanches in various climatic and geographical regions of the country:

- in areas with high cyclonic activity (Khibiny, Byrranga, Sikhote-Alin, Sakhalin, Kamchatka), mass gathering is associated with the intensity of cyclonic activity, characterized by the number of days with deep cyclones.

- in areas with average cyclonic activity (Caucasus), a mass gathering is observed both in winters with an increase in the number of days with cyclonic activity, and in winters with a number of deep cyclones above the norm.

- in the inland regions, the mass gathering is simply associated with an increase in the number of days with cyclonic activity during the cold period.

At the same time, in areas with high and low cyclonic activity, mass gatherings are associated with ordinary synoptic situations, and in areas with average cyclonic activity, synoptic conditions are characterized by an abnormal development and duration.

An analysis of the snow content showed that such events occur in winters with a snow depth coverage of less than 10%.

Graphic method

A series of observations of snow-meteorological characteristics gives in space a certain number of points corresponding to a certain image. In the case of using two signs, the space of images is visually depicted on a plane. When considering more than 2 features, projections of points onto a plane are used. A curve is constructed separating cases with and without avalanches. Graphical regression can be applied without specifying the mathematical form of the relationship between variables. Image recognition is reduced to establishing the position of the point corresponding to the current avalanche situation on the prognostic graph relative to the curve. In this case, a probabilistic approach is allowed, in which a probability field is set in the image space (Fig. 11 - Isolines of the probabilities of avalanches on a plane: the total amount of precipitation for a snowfall - days with cold and warm weather) . The line delimiting the plot areas with and without avalanches is interpreted as an isoline of zero probability of avalanches. When drawing isolines for different frequencies of avalanches, the probability of avalanche formation is determined.

Points can be grouped around some distribution centers, in proximity to which, the location of all other points in space is considered. Thus, several classes of situations can be distinguished. Identification (determining the degree of similarity) can be performed by the distance between points, the angle between vectors, the inclusion of an image inside the area .

Most often, meteorological characteristics are used in the graphical solution, i.e. the current weather conditions are assessed and the moment of reaching critical values ​​is determined (Fig. 12 - the relationship of avalanche formation with the average intensity of precipitation during snowfall (i) and air temperature. Western Tien Shan. 1, 2, 3 - data from various SLS) .

In a number of forecasting methods, specialized observational data are used that directly describe the snow cover and loads on the slope - the intensity of blizzard transport, the density of freshly fallen snow. The graph can reflect the conditions of avalanches of different genetic types.

The presence of long series of observations makes it possible to obtain graphical dependences for estimating the volumes of expected avalanches (Fig. 13 - Relationship between the volume of avalanches (numbers at points) and air temperature and precipitation intensity in the Dukant river basin) .

Graphic links obtained for the forecast of avalanches caused by snowstorms in the Khibiny , avalanches during snowfalls (certain areas of the Magadan region, the Tom river basin), wet avalanches (Tom river basin), dry avalanches during snowfalls and snowstorms (Angarakan river basin).

It is noted that the graphical method can give better results than numerical calculations on the same sample. A freehand line separates avalanche and non-avalanche situations more accurately than a linear function. The accuracy of forecasts and the warning of the phenomenon using the graphical method according to the data of production tests can exceed 90%.

Graphical empirical dependences were also obtained for cases of long-term development of avalanche formation processes. Regular observations in the pits make it possible. The family of straight lines is built based on the results of studying the stratigraphy and structure of the snow mass with layer-by-layer determination of the average crystal diameter and snow density, which indirectly characterize the mechanical strength. It is divided into five structural-density zones, characterized by an interval of critical thicknesses of snow boards that form avalanches of different sizes. This approach is used for preventive avalanches to calculate the time of the most effective impact on the snow cover.

Regression analysis

When predicting the time of avalanches using regression equations, it is assumed that the current conditions or the direction of their change will persist for some time. Periodic updates allow you to make adjustments to the forecast. Empirical formulas for different genetic types of avalanches were obtained for the Main Caucasian Range.

The method of multiple linear regression is also used to calculate the possible number of avalanches in an area with an avalanche forecast, to determine the number of avalanches that block the road (i.e., an estimate of the release distance) and to estimate the maximum volume of avalanches.

Testing methods for predicting the time of avalanches on independent material showed the possibility of using them in operational practice. The average accuracy of forecasts is 80-87%.

Discriminant Analysis

The background forecast of avalanches can be considered as a classification problem in multivariate observations. When separating situations into avalanche and non-avalanche situations, a recognition method is used based on the linear discriminant function algorithm. In the course of the forecast, the belonging of the present image to one of two groups is determined. The decisive prediction rule is the comparison of the discriminant function D with the threshold value R: for Di R, avalanches are expected, for D

The method is convenient for creating an alternative forecast of avalanche danger. Therefore, the use of linear discriminant functions for forecasting avalanche danger has become widespread in operational practice in the USSR.

Most often, linear discriminant analysis is used to separate situations into avalanche and non-avalanche situations during snowfalls and blizzards. The current values ​​of snow and meteorological characteristics are used as predictors.

Discriminant analysis can be used to study synoptic processes and determine their influence on the avalanche danger of vast mountain areas. On the basis of statistical material, the types of synoptic processes that cause avalanches to descend in a certain area are established (described in the section "synoptic method"). When waiting (forecasting) the development of a dangerous process, using a linear discriminant function, the situation is identified as avalanche or non-avalanche. The thermo-hygrometric characteristics of air masses are used as predictors for the forecast. The forecast of avalanche danger is given according to the equations obtained for each type of synoptic situations.

Recently, there have been developments for avalanche forecasting using discriminant analysis for background large-scale avalanche forecasting.

The lead time of forecasts based on methods using discriminant analysis is zero in most cases. The use of predicted values ​​of meteorological elements in calculations increases the lead time of the forecast while reducing its justification - in addition to the error of the method, the error of the meteorological forecast is added. The analysis of published materials showed that the maximum lead time of forecasts, which assess the effect of snow and meteorological factors, reaches 6 hours. Forecast methods using synoptic information have a long lead time - up to 12-20 hours.

The accuracy of avalanche danger forecasts based on discriminant analysis is 65-85%. The degree of warning of the phenomenon is 80-100%. The impossibility of a significant increase in their justification is noted.

Methods based on linear discriminant analysis have been developed: to predict snowstorm-type avalanches in the Khibiny, snowfall avalanches for several sections of the Tenkinskaya highway (Magadan region), freshly fallen and snowstorm snow avalanches for the basins of the Kunerma, Goudzhekit and Angarakan rivers (Baikal and Severo- Muya Ranges), wet snow avalanches for the SLS Pass area. The discriminant analysis method is not used to predict long-term avalanches, the collapse of which is not associated with the current meteorological and synoptic conditions. Obtaining reliable statistical estimates of the influence of factors is hindered, as a rule, by a limited number of data on the descents of such avalanches.

Nearest Neighbor Method

The presence of a database that includes information about avalanches and the values ​​of snow and meteorological characteristics makes it possible to use the possibility of searching in the past for situations similar to the current one for the purposes of forecasting.

The theoretical development of the method was carried out in the early 70s in the USSR. The database includes accumulated arrays "Meteo" (classifier of weather types and meteorological data for each day of the avalanche period), "Avalanche" (passports of avalanches), and fixed data in the array "Slope" (parameters of avalanche sources). Newly incoming avalanche and meteorological data are compared with records in the database - a study is made of the weather conditions preceding the event for any number of days before the avalanche, which can provide a certain lead time for the forecast. Nearest neighbors (Nearest Neighbors - a term adopted abroad) - days with similar weather conditions, snow conditions and avalanches or no avalanches. An automatic classification of weather types and recognition of avalanche situations are carried out according to the values ​​of the main avalanche-forming factors for different sources. An indication of a possible avalanche coming down from a separate avalanche source is the fall of values ​​beyond the critical threshold, which is determined for each parameter by its coefficient of variation. In addition to the time of descent, with the accumulation of regime information, it was assumed to predict other characteristics of avalanches - the sliding surface, the type of snow, the type of path, the height of the avalanche separation.

The nearest neighbor method requires significant computational resources and therefore has not been used in the USSR, but is widely used to predict avalanche danger abroad (Fig. 14 is an example of searching in a database for days with similar meteorological characteristics) . The main area of ​​application is the background forecast. At the same time, forecasting methods were created not for specific foci, but for territories. The disadvantage of the method is the impossibility of determining the degree of avalanche danger, as is customary in the avalanche services of foreign countries. It is not possible to estimate the number and size of avalanches. The method does not cover all the causes leading to avalanche formation, and is applicable to predicting avalanches of only certain genetic types, for example, avalanches from fresh snow.

Point system

To predict avalanche danger, the influence of certain factors and their combination on the probability of avalanches is considered. Analysis can be performed in one of the following ways:

each factor is assigned the sign "+", "-" or "0", depending on the direction of its influence on avalanche formation at a given time. The excess of negative signs suggests the absence or low degree of avalanche danger, the predominance of positive signs indicates the presence of avalanche danger, the greater, the greater their predominance. This technique, which does not take into account the specific weight of each factor in the formation of avalanches, is recommended for use in forecasting in the absence of sufficient series of snow avalanche observations.

1. quantization of predictors is carried out - each factor is assigned a certain number of points according to the degree of danger caused by it. In this case, 2 options can be applied:

1) predictor values ​​are quantized into equal intervals and each interval is assigned an increasing number of points with a constant step;

2) non-uniform quantization - non-uniform partitioning of predictor values ​​into intervals or non-uniform scoring of intervals.

Such quantization is carried out by specialists based on their own experience and its quality is highly dependent on their qualifications.

The result of the summation of points can be compared with one threshold value that divides situations into avalanche and non-avalanche situations (alternative forecast) or several - the degree of avalanche danger is determined.

The correct determination of points allows you to make a forecast (background and local) with the same accuracy as using equations.

The point system can be effective in assessing the spatial distribution of the degree of avalanche danger. Such an approach (Lawiprogmodel) using GIS technologies is proposed for the creation of the Swiss Avalanche Bulletin. The overlay function is the superimposition of several layers on top of each other, which makes it possible to obtain summary estimates of avalanche danger for different parts of the earth's surface. The degree of avalanche danger of the site is estimated by the product of the points assigned to the acting factors. These include: snow cover stability determined by test results (Rutschblock) - from 2 to 10 points, mountain slope exposure, absolute height of the site and slope steepness - each from 1 to 5 points. The weights of the first two factors change depending on the snow-meteorological situation, the values ​​for assessing the influence of other factors in this method are unchanged (Fig. 15 - weight factors of slope steepness and altitude level) .

The degrees of danger according to the European scale of avalanche danger correspond to certain values ​​of the products of points:

5 – 1250, 4 — 1000, 3 -750, 2 — 500, 1 – 250

The simulation result is a generated avalanche hazard forecast map.

The weight of the factors of the Lawiprog model is set by experts, but, as the authors note, further production verification is required to clarify the values.

Expert systems

In the presence of a variety of methods, the final determination of the wording of the avalanche danger forecast remains with the specialist. Education, experience, intuition, the ability to evaluate factors not taken into account by predictive technologies, to identify the leading one at the moment allow an expert to make quick and correct decisions. The automated expert systems that have become widespread in the practice of forecasting avalanche danger in the last decade are based on modeling the process of making a decision by an expert.

The work of expert systems is carried out in accordance with the rules formulated by specialists, while using a scoring system for assessing the influence of factors. Expert systems are often used in combination with other methods (statistical and deterministic models are used). Parallel and sequential use of various methods allows obtaining optimal results of the avalanche danger forecast.

However, the expert is not always able to explain his actions with clear rules. In this case, it is proposed to use artificial neural networks that mimic the work of the human brain (human associative memory). For example, a self-organizing Kohonen feature map (SOM) is used with an unsupervised learning algorithm in which neurons compete with each other for the right to best fit with the input signal vector and win the neuron whose weight vector is closest to the input signal vector . The weights of the winning neuron and its neighbors are adjusted taking into account the input vector, i.e. the assignment of points to the factors of avalanche formation is carried out by the computer and their value is corrected as new information arrives.

The neural network approach is especially effective in peer review tasks because it combines the computer's ability to process numbers and the brain's ability to generalize and recognize.

The functional diagram of the expert system consists of the following blocks:

1. knowledge base, including data and formulated rules;
2. a block for substituting actual data into rules and obtaining machine output with the required result;
3. block of results interpretation;
4. a conversation manager who broadcasts or presents the results;
5. a data collection unit that integrates successful results into the system to improve its further work.

Currently, several expert systems have been created and are being applied in practice or are undergoing production tests in various mountain regions and several expert systems are being improved.

Avalanche

The first attempt to formalize the experience of an expert in avalanche forecasting was carried out for avalanches associated with snowfalls in the Elbrus region. In the process of interviewing a specialist with many years of experience in the study area, using the methodology of "diagnostic games", signs were identified (the final number was 6) used by the specialist in making a forecast, their gradation and rules were determined (the order of assessment, the critical importance of factors in certain situations and the degree of their influence), which made it possible to draw up a formal prognostic scheme. In the course of the forecast, the presence or absence of avalanche danger, the places of descent and the size of avalanches were determined. The justification of the technique on independent material was from 55 to 93% for snowfalls of different intensity.

The mechanism of compilation and operation of a modern expert forecast system is clearly illustrated by the example of the DAVOS and MODUL models created at the Swiss Institute for Snow Avalanche Research.

Both models use generic COGENSYS™ inductive decision making software.

At the initial stage, the expert "trains" the program by introducing examples and interpreting the situations caused by them. Based on observation of the mentor's decision, the program computes a Boolean value for each input parameter. The logical value in this case is a measure of the influence of the parameter on the quality of the model, calculated taking into account how many situations would be indistinguishable if the parameter were excluded from consideration. Depending on the degree of influence, the parameters are assigned a value from 1 to 100. This value is continuously modified in the process of receiving new information. When faced with a new (undescribed) situation, the program searches the database for similar situations.

Each set of data corresponding to the current snow and meteorological situation is determined by the degree of avalanche danger caused by it. As a result, the program issues a judgment on the degree of avalanche danger in accordance with the European scale of avalanche danger.

Additionally, the significance level of the forecast is determined - an indicator of the program's confidence in the correctness of the result.

The difference between the models is that DAVOS uses only measured values ​​(up to 13 parameters), while MODUL estimates 30 parameters that are sequentially (step by step) calculated by the program in 11 subtasks. These include the interpretation of the Rutschblock test.

For the latest modifications of the DAVOS model, the accuracy of forecasts and the warning of events exceeded 60%. The justification of the MODUL model reached 75%.

The database of the NivoLog expert forecasting system contains numerical information on weather, snow cover, slope topography, geographic features and observed avalanches. This information is structured according to the relational data model. In addition to numerical information, NivoLog can process images such as maps, photographs or orthophotos. The combination of the expert system and the nearest neighbor method makes it possible to evaluate the stability index of the snow cover and determine the corresponding degree of avalanche danger.

The SAFRAN-CROCUS-MEPRA model package developed by French specialists has gained great fame. Only the data of daily meteorological observations are entered into the package. In this case, the main assumption is the spatial homogeneity of the data array, which determines the working scale of the package.

The output of the 1st block of SAFRAN, working according to the nearest neighbor method (thermo-hygrometric characteristics of air masses are used as factors), is a model of the fields of the most important meteorological characteristics (their surface values), cloudiness, solar radiation and average snow cover thickness at various heights and slopes different exposures at one hour time step. The model works in analysis mode or forecast mode (range 1 and 2 days).

The SAFRAN findings are then used by the deterministic CROCUS evolution model to calculate the structure of the snowpack. At the third step, the MEPRA expert system diagnoses the stability of the snow mass at different altitude levels and slopes of different exposures, taking into account its internal state, modeled in the CROCUS block. The final conclusion of the model is the forecast of the degree of avalanche danger for individual (up to 400 km 2 in area) mountain ranges with a lead time of up to 2 days.

Long-term forecast of avalanche danger

The possibility of developing a long-term forecast appeared with the creation of numerical models of climate change. The problem is solved by switching from the climate characteristics predicted by the model to the avalanche-indicating ones. The basis is the established analytical relationships between climatic characteristics (air temperature, precipitation), calculated by the model and avalanche indicators (snow cover thickness, duration of its occurrence, amount of solid precipitation, number of days with heavy snowfalls and with thaw). Further, using certain dependencies, a change in the boundaries of avalanche-prone territories is revealed, the duration of the avalanche-prone period and the number of avalanche-prone situations are calculated - a conclusion is issued on the avalanche activity of the territory in the future.

This approach was used in the work, in which the global circulation model of climate change GFDL was used.

Another method used for long-term forecasting of avalanche activity is to find in space or time a situation analogous to the predicted climate change. In this case, the data of an analogous situation are taken as avalanche-indicating characteristics and, using the established relationships, the parameters of avalanche activity of the study area are calculated for the predicted period of time.

Conclusion

The combination of numerical methods, taking into account the experience of specialists in the practical activities of the avalanche divisions of the State Committee for Hydrometeorology, made it possible to make avalanche forecasts with at least 90-95% accuracy. At the same time, extreme situations (mass avalanches, avalanches in the area of ​​activity of the population, a direct threat to objects) were predicted on the basis of intuitive thinking with almost 100% justification. However, validated and validated techniques existed for avalanche forecasts of only certain genetic types.

The progressive development of expert systems that make it possible to predict the development of avalanches caused by various factors does not yet improve the quality of avalanche forecasts. Also, deterministic models did not give a significant gain in the quality of the forecast, whose application was constrained by the impossibility of obtaining data from the avalanche origin zones. Only in recent years have models of the evolution of the state of snow cover on mountain slopes come into practice.

Often it is not possible to evaluate the advantages of one method over another, since parallel testing of several methods on the same source material is not done.

Improving the quality of the forecast can be facilitated by the introduction of GIS technologies, which are already actively used in calculating the dynamic characteristics of avalanches and in assessing the avalanche hazard of the relief. The functionality of modern GIS allows you to continuously accumulate data, perform various calculations, and spatially reference their results. The most important applied task of the developed GIS is the prediction of the time of avalanches.

Literature

1.	Abdushelishvili K.L., Kartashova M.P., Salukvadze M.E. Methods for predicting avalanches of different genetic types. Tr. 2nd All-Union. owls. on avalanches, Leningrad: Gidrometeoizdat, 1987. p. 83-87.
2.	Akifyeva K.V. Avalanche mapping in Europe. Tr. 2nd All-Union. meeting by avalanches. L., Gidrometeoizdat, 1987, pp. 214-219.
3.	Akkuratov V.N. Forecast of the onset of avalanche danger based on the values ​​of blizzard transport and thermal compression of snow. In: Questions of the use of snow and the fight against snow drifts and avalanches. M., Publishing House of the Academy of Sciences of the USSR, 1956, pp. 167-183.
4.	Berry B.L. Methods for operational forecasting of avalanches based on the use of information about the initial stages of destruction and snow movement. Tr. 3 All-Union. meeting by avalanches. L., Gidrometeoizdat, 1989, pp. 94-99.
5.	Blagoveshchensky V.P. Definition of avalanche loads. Alma-Ata. "Gylym". 1991. 116 p.
6.	Bozhinsky A.N., Losev K.S. Fundamentals of avalanche science. L.: Gidrometeoizdat, 1987, 280 p.
7.	Bolov V.R. Formation, forecast and artificial collapse of avalanches caused by snowfalls, snowstorms and sublimation recrystallization of snow. Abstract diss. for the competition uch. step. cand. geogr. Sciences. Nalchik, 1981, 26 p.
8.	Vetrov N.A., Grakovich V.F., Trutko T.V. Synoptic-climatic analysis of avalanche situations in the Elbrus region. Tr. VGI, 1984, issue 52, pp. 16-32.
9.	Gelfand I.M., Rosenfeld B.I., Urumbaev N.A. Predicting avalanches using rules that formalize the experience of a specialist. M., Scientific Council on the complex problem of "Cybernetics". 1985. Preprint.
10.	Geography of avalanches. Ed. Myagkova S.M., Kanaeva L.A. Publishing House of Moscow State University, 1992, 331 p.
11.	Glazovskaya T.G. Assessment of avalanche-prone areas of the world: methodology and results. Abstract for the competition uch. step. cand. geogr. Sciences. M., 1987, 24 p.
12.	Glazyrin G.E., Kondrashov I.V. On the methodological basis of avalanche forecasts. Tr. 3rd All-Union. owls. on avalanches, Leningrad: Gidrometeoizdat, 1989. p. 155-164.
13.	Glaciological dictionary. L.: Gidrometeoizdat, 1984. 526 p.
14.	Grakovich V.F. Information system for organizing an avalanche warning service. Abstract diss. for the competition uch. degree cand. geogr. Sciences. Moscow. 1975.
15.	Grishchenko V.F. Physical and geographical conditions of snow accumulation and avalanche formation in the Ukrainian Carpathians. Abstract diss. for the competition uch. degree cand. geogr. Sciences. Tbilisi. 1981.
16.	Grishchenko V.F., Dushkin V.S., Zyuzin V.A., Kanaev L.A., Khristoev Yu.V., Chernous P.A. Forecast of snowstorm avalanches in the USSR. Proceedings of the 2nd All-Union Conference on Avalanches. L.: Gidrometeoizdat, 1987. pp. 46-57.
17.	Dziuba V.V. Geographical principles for the development of methods for forecasting avalanche periods for little-studied areas. Abstract diss. for the competition uch. step. cand. geogr. Sciences.
18.	Dzyuba V.V., Sokolov V.M., Shnyparkov A.L. Synoptic conditions of avalanche meteorological phenomena in the coastal regions of the Chukotka Peninsula. Tr. 2 All-Union. meeting by avalanches. L., Gidrometeoizdat, 1987, pp. 94-99.
19.	Drozdovskaya N.F., Kharitonov G.G. New methods for avalanche forecasting. Tr. 3 All-Union. meeting by avalanches. L., Gidrometeoizdat, 1989, pp. 164-171.
20.	Epifanov V.P., Kuzmenko V.P. Investigation of the conditions of avalanche formation using acoustic methods. Tr. 3 All-Union. meeting by avalanches. L., Gidrometeoizdat, 1989, pp. 94-99.
21.	Izhboldina V.A. Aerosynoptic conditions for the formation and descent of snow avalanches on the Kola Peninsula. Sat. Research of snow and avalanches in Khibiny. L., Gidrometeoizdat, 1975, pp. 51-63.
22.	Isaev A.A. Experience in detailing specialized avalanche hazard forecasts for the Kamchik Pass. Tr. SANIGMI, 1998, issue 157 (238), pp. 14-19.
23.	Cadastre of avalanches of the USSR. Volume 1-20. - L .: Gidrometeoizdat, 1984-1991.
24.	Kanaev L.A. Scientific and methodological foundations for ensuring avalanche safety. Abstract diss. for the competition uch. degree of doctor of geogr. Sciences. Tashkent. 1992.
25.	Kanaev L.A. On the variability of snow cover properties. Tr. SANIGMI, 1969, no. 44(59). pp.25-42.
26.	Kanaev L.A. The main results and objectives of research on forecasting avalanche danger in the USSR (review). Tr. 2nd All-Union. owls. on avalanches, Leningrad: Gidrometeoizdat, 1987. p. 28-36.
27.	Kanaev L.A., Sezin V.M., Tsarev B.K. Principles of forecasting avalanche danger in the USSR. Proceedings of the 2nd All-Union Conference on Avalanches. L .: Gidrometeoizdat, 1987. pp. 37-46.
28.	Kanaev L.A., Tupaeva N.K. Background forecast of Western Tien Shan avalanches during cold air mass intrusions and cyclonic processes. Tr. 2nd All-Union. owls. on avalanches, Leningrad: Gidrometeoizdat, 1987. p. 69-77.
29.	Kanaev L.A., Kharitonov G.G. Evaluation of the information content of avalanche formation factors. Proceedings of the 3rd All-Union Conference on Avalanches. L.: Gidrometeoizdat, 1989. pp. 135-145.
30.	Kondrashov I.V. Conditions of formation, methods of forecasting avalanches and protection against them in the mountains of Kazakhstan. Abstract diss. for the competition uch.step. geogr. Sciences, Almaty, 1995, 40 p.
31.	Avalanche regions of the Soviet Union. Ed. Moscow State University, 1970. 200 p.
32.	Avalanches near the BAM route. Moscow: Gidrometeoizdat, 1984, 174 p.
33.	Losev K.S. On the methods of forecasting avalanches. Tr. SANIGMI, 1970, no. 51 (66), p. 100-104.
34.	Losev K.S. Fundamentals of the doctrine of the genesis of avalanches and its application to solve applied problems of avalanche science. Abstract diss. for the competition uch. step. geogr. Sciences. M., 1982. 44 p.
35.	Masyagin G.P. Calculation Methods for Forecasting Some Hydrometeorological Elements and Particularly Hazardous Weather Phenomena on Sakhalin. Tr. DVNIGMI, issue 97. 1981.
36.	Methodological recommendations for the forecast of snow avalanches in the USSR. M. Gidrometeoizdat. 1990. 128 p.
37.	Guidelines for snow avalanche support of the national economy. Tashkent. 1987. 48 p.
38.	Moskalev Yu.D. Avalanches and avalanche loads. Tr. SANII, issue 109 (190). 1986. 156 p.
39.	Okolov V.F., Myagkov S.M. Methodology for long-term forecasting of climatically conditioned hazardous phenomena (on the example of avalanches). In: Assessment and long-term forecast of changes in the nature of mountains. M.: Ed. Moscow State University, 1987, pp. 104-120.
40.	Waterwater M. Avalanche hunters. M.: Mir, 1972. 269 p.
41.	A practical guide to predicting avalanche danger. L.: Gidrometeoizdat, 1979. 200 p.
42.	Problems of the effectiveness of protection against avalanches. Ed. Bozhinsky A.N., Myagkova S.M. Dep. in VINITI N 3967-B91. M., 1991. 285 p.
43.	Guidelines for the preventive descent of snow avalanches using artillery systems KS-19. Moscow: Gidrometeoizdat, 1984. 108 p.
44.	Snow avalanche guide (temporary). L.: Gidrometeoizdat, 1965. 397 p.
45.	Seversky I.V., Blagoveshchensky V.P. Assessment of the avalanche danger of the mountainous territory. Alma-Ata. 1983. 220 p.
46.	Sezin V.M. Classification of situations into avalanche and non-avalanche situations when southern cyclones enter Central Asia. Tr. SANII, 1983, issue 99 (180), pp. 112-118.
47.	Seliverstov Yu.G. Methodology for calculating the economic damage from avalanche blockages on highways (on the example of Kyrgyzstan). In: Overview mapping of natural hazards and natural disasters. M.: MSU, 1992. S.233-242. Dep. at VINITI 24.04.1992. 1389.B.92.
48.	Snow and snow falls in the Khibiny. M., L.: Gidrometeoizdat, 1938, 100 p.
49.	Sokolov V.M., Troshkina E.S., Shnyparkov A.L. Handbook for forecasting avalanches in the border areas of the USSR. M.: GU PV KGB USSR, PLSLS MGU, 1991, 129 p.
50.	Troshkina E.S. Avalanche regime of the mountain territories of the USSR. M., VINITI Publishing House, 1992, 196 p.
51.	Troshkina E.S., Voitkovsky K.F. Predictive assessment of the effectiveness of anti-avalanche measures. In: Snow cover in the mountains and avalanches. M.: Nauka, 1987. S. 137-143.
52.	Tushinsky G.K. Glaciers, snowfields, avalanches of the Soviet Union. M., 1963. 312 p.
53.	Guidelines for the calculation of avalanche loads in the design of structures VSN 02-73. M. Gidrometeoizdat, 1973. 20 p.
54.	Kharitonov G.G. Method for forecasting avalanches in the basin of the river. Kunerma (Baikal Range). Tr. 2nd All-Union. owls. on avalanches, Leningrad: Gidrometeoizdat, 1987. p. 87-94.
55.	Chernous P.A., Fedorenko Yu.V. Probabilistic assessment of snowboard stability on slopes. Mat. glossy iss. 2000, issue 88. pp. 87-91.
56.	Shnyparkov A.L. Particularly large avalanches and conditions for their mass descent. Abstract diss. for the competition uch. degree cand. geogr. Sciences. Moscow. 1990.
57.	Shubin V.S. To the forecast of avalanche danger along the Tenkinskaya highway in the area of ​​the Dondychan avalanche post. Tr. 2nd All-Union. owls. on avalanches, Leningrad: Gidrometeoizdat, 1987. p. 100-107.
58.	Shubin V.S. Forecast of avalanche danger for the inland regions of the Magadan region. Inf. letter from the Magadan State Medical Center. Magadan, 1987.
59.	Ammann W., Buser O., Vollenwyder U. Lawinen. Basel: Birkhauser V., 1997, 170 S.
60.	Avalanche classification. Hydrological Science Bulletin. 1973, 1b, No. 4, p.391-402.
61.	Birkeland, Karl W.; Johnson, Ron; Herzberg, Diane. 1996. The stuffblock snow stability test. Tech. Rep. 9623-2836-MTDC. Missoula, MT: U.S. Department of Agriculture, Forest Service, Missoula Technology and Development Center. 20p.
62.	Bolognesi R. NivoLog: An Avalanche Forecasting Support System. ISSW'98. URL: http://www.issw.noaa.gov/hourly%20agenda.htm
63.	Bolognesi R., Buser O., Good W. Local avalanche forecasting in Switzerland: strategy and tools. A new approach… ISSW'98. URL: h
64.	Bolognesi R.,Denuelle M. , Dexter L. Avalanche Forecasting with GIS. URL: http://www.avalanche.org/~issw/96
65.	Brun E., Martin E., Simon V., Gendre C., Coleou C. An energy and mass model of snow cover suitable for operational and avalanche forecasting. J. Glaciol., 35 (121), 1989, 333-342.
66.	Buser O., Föhn, P., Gubler W., Salm B. Different methods for the assessment of avalanche danger. Cold. Reg. sci. Technol., 1985, 10(3), 199-218.
67.	Buser, O., Butler, M. and Good, W. 1987. Avalanche forecast by the nearest neighbors method. IAHS Publ. 162.557-569.
68.	Durand Y., Brun E., Merindol L., Guyomarc'h, Lesaffre B., Martin E. A meteorological estimation of relevant parameters for snow models. Ann. Glaciol., 18, 1993, 65-71.
69.	Föhn P., Haechler P. Prevision de grosses avalanches au moyen d'un modele deterministe-statistique. In Deuxieme Rencontre Internationale sur La Neige et les Avalanches. 1978 Competes Rendus. Grenoble, Assotiation Nationale pour l'Etude de la neige et les Avalanches, 151-165.
70.	Föhn, P. 1987. The Rutschblock as a practical tool for slope stability evaluation. IAHS Publication, 162, 223-228.
71.	Föhn P. An overview of avalanche forecasting models and methods. Oslo, NGI, Pub.N 203, 1998, 19-27.
72.	Giraud O., Brun E., Durand Y., Martin E. Safran/Crocus/Mepra modelsas a helping tool for avalanche forecasters. Oslo, NGI, Pub.N 203, 1998, 108-112.
73.	Glazovskaya T. Global distribution of snow avalanches and possible change of avalanche activity in the Northern Hemisphere due to the climatic change. Annals of Glaciology. Cambridge, UK, 1998. Vol. 26, p. 337-342.
74.	Houdek J., Vrba M. Zimni nebezpeči v horbch. Praha: Statni Tĕlovechovni Nakladatelstvi, 1956. 205 p.
75.	Judson A., Leaf C.F., Brink G.E. A process-oriented model for simulating avalanche danger. J. Glaciol., 26 (94), 53-63.
76.	Klinkenberg P. Avalanche hazard modeling using GIS. URL: http://www.csac.org
77.	LaChapelle E. Avalanche forecasting – a modern synthesis. Publ. Assoc. Intern. Hydrol. Sci., 1966, No. 69, p.350-356.
78.	Leuthold H., Allgöwer B., Meister R. Visualization and analysis of the Swiss avalanche bulletin using GIS. ISSW'98. URL: h
79.	Leuthold, H., Allgower, B. and R. Meister. 1997. Visualization and analysis of the Swiss avalanche bulletin using GIS. Proceedings of the International Snow Science Workshop 1996, Banff, Canada. 35-40.
80.	McClung, D.M. and P. Schaerer. 1993. The Avalanche Handbook. The Mountaineers, Seattle, Washington, U.S.A., 271 pp.
81.	Meister R. Country wide avalanche warning in Switzerland. ISSW'98. URL: h http://www.issw.noaa.gov/hourly%20agenda.htm.
82.	Observation Guidelines and Recording Standards for Weather, Snowpack and Avalanches Prepared by the Canadian Avalanche Association. 1995, ISBN 0-9699758-0-5
83.	Perla R.I. On contributory factors in avalanche hazard evaluation. Can. Geotech. J., 7(4), 1970, 414-419.
84.	Schweizer J., Föhn P. Two expert systems to forecast the avalanche hazard for a given region. ISSW'98. URL: http://www.issw.noaa.gov/hourly%20agenda.htm.
85.	Schweizer J., Jamieson J.B., Skjonsberg D. Avalanche Forecasting for Transportation Corridor and Backcountry in Glacier National Park (BC, Canada). Oslo, NGI, Pub.N 203, 1998, 238-244.
86.	Schweizer, M., Fohn, P.M.B. and Schweizer, J. 1994. Integrating neural networks and rule based systems to build an avalanche forecasting system. Proc. IASTED Int. Conf.: Artificial Intelligence, Expert Systems and Neuronal Networks, 4-6 July 1994, Zurich, Switzerland.
87.	Seliverstov Yu., Glazovskaya T. Forecast of avalanche danger for the intracontinental regions of Northeast of Eurasia. Oslo, NGI, Pub.N 203, 1998, 245-248.
88.	Stephens J., Adams E., Huo X., Dent J., Hicks J., McCarty D. Use of neural networks in avalanche hazard forecasting. ISSW'98. URL: h http://www.issw.noaa.gov/hourly%20agenda.htm.
89.	Tschirky F. Lawinenunfallstatistik der Schweiz 1985 – 1998. URL: http://www.slf.ch.
90.	URL: http://www.avalanche.org
91.	URL: http://www.neuroproject.ru.
93.	URL: http://www.csac.org
94	Ward R.G.W. Avalanche prediction in Scotland. Applied Geography, 1984, vol. 4, p. 91-133.

The use of this material on other resources is prohibited!