, . ^^'- If

A HANDBOOK OF FORECASTING AND CONTROL MEASURE!

Agriculture Handbook No. 194 January 1961

FOREST SERVICE U.S. DEPARTMENT OF AGRICULTURE Ai^ JANUARY 1961 AGRICULTURE HANDBOOK NO. 194

SNOW AVALANCHES J

A Handbook of Forecasting and Control Measures k

FSH2 2332.81 SNOW AVALANCHES

:}o U^»TED STATES DEPARTMENT OF AGRICULTURE FOREST SERVICE SNOW AVALANCHES FSH2 2332.81 Contents

INTRODUCTION 6.1 Snow Study Chart 6.2 Storm Plot and Storm Report Records CHAPTER 1 6.3 Snow Pit Studies 6.4 Time Profile AVALANCHE HAZARD AND PAST STUDIES CHAPTER 7 CHAPTER 2 7 SNOW STABILIZATION 7.1 Test and Protective Skiing 2 PHYSICS OF THE SNOW COVER 7.2 Explosives 2.1 The Solid Phase of the Hydrologie Cycle 7.21 Hand-Placed Charges 2.2 Formation of Snow in the Atmosphere 7.22 Artillery 2.3 Formation and Character of the Snow Cover CHAPTER 8 2.4 Mechanical Properties of Snow 8 SAFETY 2.5 Thermal Properties of the Snow Cover 8.1 Safety Objectives 2.6 Examples of Weather Influence on the Snow Cover 8.2 Safety Principles 8.3 Safety Regulations CHAPTER 3 8.31 Personnel 8.32 Avalanche Test and Protective Skiing 3 AVALANCHE CHARACTERISTICS 8.33 Avalanche Blasting 3.1 Avalanche Classification 8.34 Exceptions to Safety Code 3.11 Loose Snow Avalanches 3.12 Slab Avalanches CHAPTER 9 3.2 Tyi)es 3.3 Size 9 AVALANCHE DEFENSES 3.4 Avalanche Triggers 9.1 Diversion Barriers 9.2 Stabilization Barriers CHAPTER 4 9.3 Barrier Design Factors 4 TERRAIN 9.4 Reforestation 4.1 Slope Angle CHAPTER 10 4.2 Slope Profile 4.3 Ground Cover and Vegetation 10 AVALANCHE RESCUE 4.4 Slope Orientation 10.1 Rescue Equipment 10.2 Immediate Action CHAPTER 5 10.3 FoUowup Action 10.4 Avalanche Accident Report o AVALANCHE HAZARD FORECASTING 10.5 Case Histories 5.1 The Contributory Factors 5.11 Factor Analysis CHAPTER 11 5.2 Delayed-Action and Climax Avalanches 5.3 The Ram Profile 11 AREA SAFETY PLANNING 5.4 Wet Snow Avalanches 11.1 Terrain Analysis CHAPTER 6 11.2 Climate Analysis 11.3 Hazard Classification STANDARD SNOW OBSERVATIONS AND 11.4 Operations Plan TERMINOLOGY BIBLIOGRAPHY

For sale by the Superintendent of Documents, U.S. Government Printing Office, t , Washington, 25, D.C. — Price 60 cents 2332.81 SNOW AVALANCHES Introduction

The first publication by the U.S. Forest Service in 1952, has now become a widely accepted and ef- dealing with snow avalanches, "Alta Avalanche fective means of avalanche control in the Western Studies," appeared in 1948. This booklet reported United States. On the other hand, the basic the results of experiences and studies at Alta, principles of snow mechanics and avalanche for- Utah, where the problem of snow safety for the mation, and the commonsense rules for safe con- skiing public had confronted forest officers since duct in avalanche terrain have not changed, the area had been established 10 years previously. though an understanding of snow mechanics in- The study program at Alta continvied, and ex- creases with passing time and longer experience. perience was rapidly gained in methods of This new handbook provides an opportunity to avalanche control which made possible increasing bring forest officers up-to-date instructions on the use of excellent alpine slopes without compromis- most modern methods of avalanche hazard fore- ing public safety. Study stations similar to that casting and control, and to recast the general text at Alta were also established at Stevens Pass in on avalaiiches in a more informative and concise Washington and at Berthoud Pass in Colorado, pattern. This handbook is intended as a textbook, where the investigations and tests of control as a training aid, and as a field manual for the use methods could be extended to encompass a wider of forest officers and others whose duties involve range of climatic conditions. In 1952 the expe- avalanche hazard. rience of forest officers working in these areas was Directly, or indirectly, many persons and or- compiled and published as the "Forest Service ganizations have contributed to this handbook. Avalanche Handbook." This was designed to be a We acknowledge a particular debt to our European manual of instraction for foresters confronted colleagues for material on the physics of snow and with snow safety problems following the rapid avalanche defense structures. postwar expansion of skiing and development of new ski areas on national forest lands. It also served as a general text on the nature, character- istics, and formation of snow avalanches. The h extensive appendix included a report on avalanche studies and research programs then in progress. The "Forest Service Avalanche Handbook" was ■à M ' the only text in English on this subject widely available in the United States, and it was soon accepted by the public for use far beyond that for which it was designed. The training program of the National Ski Patrol System alone generated a heavy demand for it. Requirements of other groups, such as ski instructor associations, plus mterest of the general skiing and mountaineering public, soon exhausted stocks at the Government Printing Office. Eight years have passed since the first "Forest Service Avalanche Handbook" was published. During this period much practical experience has been gained, both through study programs and through the ever-widening practical application of snow safety measures. Many techniques pro- posed or described as innovations in 1952 are now routine procedures used by Forest Service snow F-495188 FIGURE 1.—105 mm. recoilless rifle mounted on a Forest rangers. For example, the use of artillery (fig. 1 ), Service truck. Tliis is a highly mobile weapon for which was barely out of the experimental stage avalanche control. 2332.81 SNOW AVALANCHES Chapter 1

1 AVALANCHE HAZARD AND PAST location than lack of potential. In the first World STUDIES. The remote mountain areas of War, on the Austro-Italian front in December America are the scene of hundreds, perhaps thou- 1916, thousands of soldiers lost their lives in snow- sands, of avahxnches each winter. No man wit- slides during a single 24-hour period. Total cas- nesses them, and tliey have no measurable effect on ualties from avalanches during the war on that civilization. But in certain locations, which are front were greater than from military action. increasing in number, avalanches and mankind do This was the greatest avalanche disaster in rec- meet. This creates avalanche hazard—a threat to orded history; it was brought about by an ab- life and property from the cascading snow (tig. 2). The avalanche is a great natural force, not in- normal concentration of men stationed in the ferior in ix)wer to the tornado, the flood, and the mountains, abetted by an abnormal winter. It earthquake. It is not so well known as a destroyer demonstrates the difference between avalanches of life and property more because of timing and themselves and avalanche hazard.

r-49B189 FIGURE 2.—A climax, hard slab avalanche ( HS-AE-5—meaning of abbreviations is given in figure 38). This slide nearly buried a lift terminal, but lost its momentum just in time to prevent any damage. Such large slides have not been observed on this slope since it has been stabilized by routine artillery fire during and after every storm During the Gold Bush period, from about 1860 Alta, with its grim history of death and de- to 1910, the mountains in tlie Western States were struction at the hands of the "white death" was miicJi more heavily populated than they have been reborn as a winter recreation area and the site of at any time since then. the first avalanche observation and research cen- From the Sierras and the Cascades to the Rock- ter in the Western Hemisphere. ies, miners swarmed into the high country and In 1960, avalanche hazard threatens more per- built their camps wherever they found silver and sons and more enterprises than ever before, and gold. Telluride and Aspen in the Colorado Rock- the number is increasing. The spectacular in- ies, Atlanta in the Sawtooths, Mineral King in the crease in winter sports activities throughout the Sierras, Alta and Brighton in the Wasatch—to United States in the decade ending 1960 has been name a few—were fainous in their day for fabu- responsible for the greatest increase in hazard. lous riches. They were known also for the sudden, The number of skiers in the national forests of the dramatic, and complete destruction by snowslides Intermountain Region has increased two and one- that often obliterated both miners and their half times since 1950. The number of persons ex- camps. posed to possible danger continues to increase The mining era was the first time that human yearly. beings in large numbers invaded the alpine zones Witli the skier have come businesses, hotels, pri- of the United States. The consequences were vate cabins, and heavier highway traffic, all con- often disastrous. Verifiable records of avalanche tributing to the economy and development of the accidents in the vicinity of Alta, for instance, go West. As new highways are planned and use back as far as 1874, when the mining camp was of the older ones increases, more and more non- practically obliterated by avalanclies; more than skiers are also exposed to possible snow hazards 60 lives were lost. During the next 35 years, ava- in the many mountain passes. The development lanches killed 67 more persons. Three slides in and increasing population in the Western States three diiferent years wiped out more than 10 per- is bringing more and more people in contact wúth sons each. snow avalanches. Much of this contact occurs on Demonetization of silver plus exhaustion of the national forest lands, so the safety problem be- richest claims ended the mining rush, and miners comes an ever-increasing concern of the Forest vanished from their alpine camps almost as sud- Service. denly as they had appeared. Except for a few Large sections of Alaska are first-class ava- trappers and prospectors, the mountains were lanche country, comprising as they do some of deserted again in the winter. The avalanches re- the world's most spectacular alpine terrain cover- mained, but the avalanche hazard Avas sharply re- ing thousands of square miles. Alaska so far is duced because the number of potential victims was very sparsely populated and the hazard in most reduced. places is not yet large, but the high mountains are Other targets quickly developed after the min- a dominant part of the landscape, and even now ing camps were deserted—transcontinental rail- the few roads and railroads connecting the scat- roads and highways. The railroads far excelled tered population centers are endangered by ava- the highway builders in establishing all-year traf- lanches. If Alaska fulfills its promise as a new fic through the western mountains. Even so, it frontier for rapid development, the avalanche was on a railroad that the greatest avalanche hazard will rise rapidly in the next few decades. disaster in United States history took place. In Much of the-accessible mountain terrain near the March 1910, three trains were snowbound at Wel- seaeoast is within the national forests (fig. 4). hn^ton in the Cascade Mountains of Washington. There is no possibility that exposure to ava- A smgle avalanche swept all three into the bottom lanche hazard in the United States will again de- of a canyon at a cost of more than 100 fatalities cline. The pressure of traffic, communications, and a million dollars' property damage (fig. 3). recreation, and industry is too great. Neither is This tragedy forced the railroad underground. hazard the product of climactic winters only. In- Millions of dollars were spent on the tunnel that jury, death, and property damage are an annual bypassed the slide area. toll which merely rise to a peak in years of ab- The succeeding quarter century saw a reversal normal severity. ot the downward trend of avalanclie hazard in nie United States. Industrial and agricultural The oldtime answer to avalanche hazard— development of tlie West led to constniction of •'either stay out of the mountains or take your v^. ;T" "^ I'lg^^^-ays and their maintenance the chances"—is no longer good enough and the winter DusLruii" 1 "'f reopened; logging operations techniques developed by the Forest Service to cope control nr'^T «^«^'^ti«"s ; reclamation 'and flood Avith this hazard are providing a new answer. telenwf ■^!i^'''"g^^t the headwaters; power, _ The study of avalanches is difficult and expen- anjÄn i^P'P'^^'^^^ ^°"^P«ted with highways sive. Necessity is the main reason for undertak- ing it. The Swiss, pioneers in snow research, tC monn^ ? ^' f ' T'^ «f skiing attracted into have lived with avalanche hazard for centuries, and have avalanche defense works as old as the towns they guard. !»"AvV-

y. ■^-'^VK.'_,,J, .;

FIGURE 3 —A.n example of avalanche hazard and protective measures. Here a series of slide paths descend through timber and cross the old railroad grade leading over Stevens Pass, Washington. A long snowshed has been constructed to protect the grade. The arrow points to the site of the Wellington slide disaster of March 1, 1910. A major avalanche overwhelmed 3 trains at this point and carried over 100 persons to their deaths, T^V-"The '•in-rail- road has since been rerouted through a tunnel under the pass.

In 1931 a national commission was established was necessity. This abandoned mining camp with in Switzerland to coordinate all of the work be- its history of death and destruction by avalanche ing done by various groups and organizations. was being developed as the first true alpine winter This resulted in the famous Avalanche Institute sports area in this country. Forest officers recog- with its snow laboratory and staff of technicians nized an ideal combination of terrain and climate and engineers. The Institute engages in all forms near a large city of potential skiers within easy of snow study from basic research in snow driving distance. They further reasoned that safe mechanics to the development of avalanche use by the public was possible only under the barriers. supervision of a snow ranger armed with power Avalanche study in the United States began in to close slopes under avalanche hazard conditions. the winter of 1937-38 with the assignment of the Actually it is surprising that the study of ava- first snow ranger to Alta, Utah. Again the reason lanches did not begin 30 years earlier, after the

577868 0—61 2 F-495191 FIGURE 4.—A striking example of avalanche hazard. Here an important highway and railroad are threatened by a series of slides which fall over unbroken sloi)es from a height of 3,000 feet. Turnagain Arm, on the Chugach National Forest, near Anchorage, Alaska.

disastrous winter of 1909-10. Two official reports trained avalanche observer, their analyses contain were published. One, by a forester, noted the fact some very acute observations. In fact, one state- that forest fires had recently denuded the slopes ment by the meteorologist has become a motto in above the railroad, creating avalanche paths the avalanche profession : wJHcJi had previously been protected by timber cover, i he other, by a meteoro!ofrist, described the "It was not only the quantity of snow that weather patterns whicli led to severe avalanche fell that caused so many avalanches but also conditions. While neither of the writers was a the manner in which it fell." Lacking any means of combating them, the miners endured avalanches as one more hazard in a hazardous occupation. The railroads retired Ko» ^ • into their tunnels and snowsheds and mitil recent years, the law of averages has been a defender of the highways. The odds against a moving car and an avalanche arriving at the same point at the same moment are good. But the odds are getting poorer eveiy year with increasing winter travel by bus, truck, and private automobile. The Forest Service inherited tlie avalanche hazard problem because so much of the desirable alpine skiing terrain was on national forest land. There was strong demand for the development of these areas, but the cost in lives and damage the gold minei-s were willing to pay was not accept- able for public recreation. Thus the Forest Serv- ice in the interest of public safety found itself in ":>.i„T«__»u the business of combating avalanches. P—495192 The Forest Service undertook to obtain maxi- FIGURE 5.—The destructive force of a large avalanche. mum public use of alpine ski areas with adequate This ear was carried several hundred feet off a high- public safety. It was obvious from the start that way and completely demolished by a large soft slab snow and avalanche research in this country would avalanche (SS-N-4). take a course different from that being carried on in the Alps. There was neither the time, the in 1950 with the establishment of the observation funds, nor the trained pei"sonnel to engage in basic station at Berthoud Pass, Colorado. research. Efforts had to be directed toward the Publication of the "Alta Avalanche Studies" rapid development of field techniques for the l)rought about widespread interest in the avalanche recognition and reduction of hazard. hazard problem in this country. This interest was Geography was also responsible for a different further stimulated by a series of severe winters approach. Ski areas scattered throughout the beginning with 1948-19, attended by numerous West were located in widely different climatic avalanche fatalities and much property damage. zones and in varied types of terrain. In contrast No losses occurred at ski areas protected by Forest to the compact areas of the Swiss Alps, the moun- Service snow rangers. This led to requests for tainous zones of tlie Ignited States are spread out assistance on avalanche hazard problems from over a large territory, exposed l)oth from north to many sources. The development of numerous al- south and east to west to different weather pat- pine-type ski areas following World War II led to terns and storm tracks. It was obA-ious that the establishment of observation stations covering local methods would have to be developed in dif- the full range of alpine climate in this country. ferent areas, for no one location could be taken With the assistance of the U.S. Weather Bureau, as representative for all at any given time. The a number of weather instruments were installed— N'orthwest mountains might be experiencing rain, Alta 1937, Berthoud Pass 1950, Stevens Pass 1951, the Intermountain ranges snowfall, and the Colo- Squaw Valley 1956. rado Rockies basking in sunshine. Alta is the central collecting station for ava- Until 1950 the snow safety program was con- lanche data. Office space was added in 1955, pro- fined to winter sports areas under the jurisdiction viding accommodations for Forest Service ava- of the Forest Service. The principal objective lanche files, an avalanche library, and modest re- was the protection of human life. search facilities. The recent study program at this It was soon learned that about 80 percent of the station has included instrumentation development, slides large enough to endanger life and property compilation and review of records, tests of artil- run either during or immediately after storms lery and explosives for avalanche control, and lit- (fig. 5). Therefore, it was decided that efforts erature review and translations. The U.S. Forest should be concentrated on analyzing the weather Service Alta Avalanche School, initiated in 1949, factors which produce these direct-action ava- lanches. But the importance of structural changes is held every other year. It was originally de- in the snowpack which cause delayed-action ava- signed as a training course for forest officers, but lanches was still kept in mind. Delayed-action has since attracted students from a wide number avalanche hazard is an acute problem in the high of private and Government agencies concerned alpine zone. Studies of this hazard were initiated with snow safety problems. 2332.81 SNOW AVALANCHES Chapter 2

2 PHYSICS OF THE SNOW COVER. The scale of geologic time, and their influence is so physical properties of snow and tlie processes of slowly felt that they are obscured to human per- change which it nndergoes are complex subjects ception by short term climatic variations. being explored by research scientists. Tliis chap- A similar withdrawal on a much more acceler- ter can be no more than a short introduction to ated scale takes place with the deposit of a transi- the beliavior of snow. The reader is referred to ent winter snow cover throughout the temperate the bibliography at the end of this liandbook for and arctic zones each year. This annual storage additional sources of information. exercises a strong effect on the yearly amount of The serious student of snow behavior and ava- free water available to arid or semiarid regions in lanches is urged to give careful attention to these these zones. In either case, the storage phenom- fundamentals before proceeding to the practical enon owes its existence to the fact that vast quan- aspects of avalanche hazard forecasting and con- tities of energy ai'e required to return water to trol. Though a purely empirical knowledge of liquid state once it has been deposited as a solid. snow can serve in many instances, a thorough The amount of heat required to melt, without understanding of the basic physical processes in- change in temperature, a given weight of ice is volved is essential to tlie intelligent interpretation seven times that required to melt the same weight of observations. This is particularly true of the of iron, and thirteen times that for the same weight more complex forecasting problems concerned of lead. with delaved-action avalanche hazard. 2.2 Formation of Snow in the Atmosphere. 2.1 The Solid Phase of the Hydrologie Cycle. Wlienever atmospheric conditions favor precipi- The conditions of geology, climate, and life on tation of water vapor at temperatures below the this planet are immensely influenced by the exist- freezing point, precipitation occurs as snow. ence of a temperature range wherein water, one Snow crystals are known to form around nuclei of the most common substances on the earth's sur- of foreign matter in the air, such as microscopic face, exists in all three states of matter—solid, dust particles. The first step is the formation of liquid, and gas. It is difficult to conceive what a small ice crystal around the nucleus. This crys- the face of the earth would look like if the cycle tal grows by the deposition of additional ice from of evaporation, condensation, and runoff were water vapor in the atmosphere. (The transfer of eliminated by temperatures high enough to keep water directly from the vapor to the solid state, all of the water in the form of vapor, or low or vice versa, is known as sublimation.) Recent enough to keep it in the form of ice. A complex investigations suggest that minute water droplets hydrologie cycle exists because the temperatures (diameter around 1 micron) may also contribute that do prevail pennit water to pass readily from to the growth of snow crystals. These crystals in one state to another. The liquid and vapor states general assume a hexagonal pattern, but, beyond of water are the active participants in the hydro- this, the variations in size, shape, and form are logic cycle of nature, while the solid state—snow almost infinite. A general classification according or ice—represents water that has been temporarily to form is given in figure 7. The particular form withdrawn from the cycle and placed in a natural developed in the atmosphere depends on the air storage reservoir. This withdrawal occurs mainly temperature and the amount of water vapor avail- in the form of precipitated snow (fig. 6). The able. Wlien a snow crystal falls through different amount of water contained in this naUiral reser- voir has varied during the history of the earth as air masses with different temperature and vapor great climatic changes of unknown origin have conditions, the more complex or combined types cUmS ^ succession of ice ages and tropical may develop. Crystals formed at the freezing point, or falling through air whose temperature \na^T^T "^ ^^^^ P^-''^'' i'^e ^^Ps i^"d continental is near it, stick together to become aggregates of fitnrÜq'^!. Jiu'l^ ^*^"Ç ^<'™. basis, and water so individual crystals or snowflakes. cvolf/' 7•^^^^r^"'^ from the active hydrol osric When snow crystals fall through air which con- wîhhnïïi ""ï""'^' ^^ thousands of years. This tains water droplets, these droplets freeze to the durW î^ ''* ''■'*'' ^^«"^ the hydrologie cycle crystals as a rime deposit. As the amount of rime oSf wTf' '"f '^^ ^^t^™ during'clim-atic on a crystal increases, the original shape is ob- the Si hf lr°^^"'^^ ^ff^^t^ «" the climate of scured and a rounded ball results, giving rise to the the earth, but these effects are measured on the type of snow commonly known as pellet or tapioca. FALLING SNOW H20 ATMOSPHERIC WATER VAPOR * O ^ / -i-^< 'DROPLETS ^

HOAR FROST REGENERATED SNOW DEPTH HOAR NEW SNOW G = 0.l-0.3 G=0.l-0.3 G = .OI-.25 D=l.2 MM ayi/v">- 0 = 2-10 MM R-=l-IOKG R=5-23 KG R= 1-IOKG

POWDER SNOW H^ WIND-DRIFTED SNOW OLD SNOW 6 = 0.2-0.4 6 = 0.1-0.4 D=0.5-l MM D = 0.5-2MM R = IO-200 KG R = 5-I00KG

6LACIER ICE CONSTRUCTIVE METAMORPHISM 6=0.8-0.9 D = 2MM-30+CM DESTRUCTIVE METAMORPHISM t I I I I—I—I I I—I—1>-

FIGURE 6.—The solid phase of the hydrologie cycle, showing the various paths by which water vapor is deposited as a solid on the earth's surface. The ice so deposited may at any time revert to the liquid phase by melting ; only in cerUin favorable climatic conditions does it reach the final stage of glacier ice. This chart shows the terrestrial part of the cycle only, sea ice having been omitted.

The technical term for these rimed crystals is Graphing snow depths on a time (months and graupel (fig. 8). days) base provides a general record of the snow A relationship has been noted between air cover behavior throughout the course of a winter. temperature and density of newly fallen snow. Periodic observations of density, snow type, hard- In general, new snow density increases with air ness, thickness, and other characteristics of indi- temperatures. The relationship appears to be at vidual layers can be plotted as profiles on the same least partially independent of snow type. The time base. Pits dug from time to time provide percentage of water in new-fallen snow may range information about internal changes taking place, from 1 to as high as 25 percent, with the average and plotting meteorological data on the same value around 10 percent. The lightest snow is graph allows the effects of various weather ele- deposited under cold and very still conditions. ments to be studied. This technique is described The highest new snow densities are associated in detail in chapter 6, items 6.3 and 6.4. with graupel and needle crystals falling at tem- Metamorphism within the snow cover is a con- peratures near freezing. tinuous process which begins when the snow is 2.3 Formation and Character of the Snow deposited and continues until it melts. The Cover. Formation of the snow cover is in some normal path of metamorphism tends to destroy ways analogous to the geology of sedimentary the original crystalline forms of the deposited rocks. Solid precipitation from the atmosphere snow crystals and gradually convert them into (sedimentation) builds up the snow cover layer rounded, isometric grains of ice (destructive by layer, with a resulting stratification that dis- metamorphism). The physical process causing plays the history of the weather variations which these changes is not entirely clear, but transfer of occurred during its buildup. With the passage water vapor by sublimation from the points of of time, structural and crystalline changes within tlie crystal branches to the more central portions the snow cover (metamorphism) erase the original of the crystal appears to play an important role. stratigraphie differentiation and convert the snow The process is strongly influenced by temperature, into new forms. and proceeds at a rapid rate when the temperature o 2 iv ■*ar

5 F-495194 FIGURE 8.—Graupel, or pellet snow, which has been sub- jected to destructive metamorphism within the snow cover. Magnified about 2.5X. The pellet character- 6 istic has been retained, making layers of this snow type identifiable for a long time after incorporation in the snow cover. ( Note : Tliere is no figure 9. ) 7 >r^ i¿ )tfH a»-^ enced by pressure as well as temperature (pressure metamorphism), and the weight of additional snowfalls over a given snow layer causes its meta- morphism to accelerate. 8 Ü O ^^^ ^ ^^ The destnictive metamorphism of the snow crystals results in a reduction of the space a given amount of ice occupies. This causes an increase in density to accompany metamorphism, and the 9 A ©^ snow cover to shrink, or settle. This settlement is a continuous, visible evidence of metamorphism. When the process is accelerated by rising temper- atures, the settlement rate of the snow cover in- O A 5 .3 creases. F-495193 A process known as constructive metamorphism FIGURE 7.—Types of solid precipitation (International Snow Classification) : may also take place in the snow cover. This oc- curs when water substance is transferred from one 1. Plates also combinations of plates with or part of the snow cover to another by the vertical without very short connecting columns. diffusion of water vapor. This vapor is deposited 2. Stellar also parallel stars with very short around new centers of crystallization and forms crystals. connecting columns. ice crystals with quite a different character from 3. Columns and combinations of columns. those of the original snow. These crystals tend 4. Needles and combinations of needles. 5. Spatial spatial combinations of feathery to assume a scroll or cup shape, appear to be dendrites. crystals. layered, and often grow to several millimeters 6. Capped columns with plates on either (or in diameter (figs. 10^ and 11). They form a columns. one) side. very fragile mechanical structure which loses all Î. Irregular irregular compounds of microscopic particles. crystals. cohesion upon collapse and becomes very soft when 8. Graupel pellet snow, or soft hail, with iso- wet. The snow form produced by this type of metric shape. Central crystal metamorphism is known as depth hoar, and is _ cannot be recognized. ». Ice pellets ice shell, inside mostly wet (sleet). popularly referred to as sugar snow. The condi- 0. Hail tions required for its formation are a rapid change of temperature with depth (steep temperature IS near the freezing point. At extremely low gradient) to cause diffusion of water vapor within temperatures metamorphisim is very slow, and the snow cover, and a sufficiently high air per- practically stops below -40°. As this proce.ss meability to permit this diffusion to take place. take.s place with all types of snow crystals, they These conditions are most common early in the tend to approach the uniform condition of rounded winter, when the snow cover is shallow and ice grains, and the snow cover becomes more ho- unconsolidated. mogeneous. The rate of metamorphism is influ- ' There Is no figure 9. 10 FIGURE 10.—Depth hoar crystals from the bottom layer FIGURE 11.—Surface hoar crystals, formed by sublima- of a snow cover 65 cm. deep. This snow has under- tion of water vapor from the atmosphere onto a cold gone partial constructive metamorphism, and only snow surface. These display the same characteristic some of the crystals display the distinct layering and layering and crystal form as depth hoar, but usually cup-shaped form of mature depth hoar. Magnification are flat instead of cup-shaped. Magnification about about 2.5X. 2.5X. Snow is an essentially plastic material. Because foot packing is an important factor in avalanche of its plastic nature, any snow cover situated on a prevention. sloping surface tends to flow downhill under the 2.4 Mechanical Properties of Snow. Snow be- influence! of gravity. This behavior is known as longs to that class of solid materials which are creep, and is highly dependent on temperature, termed "visco-elastic." That is, it exhibits the since snow is very plastic near the freezing point properties of both viscous fluids and elastic solids and becomes increasingly brittle at lower tempera- in varying combinations which depend on density, tures. If the ground surface of a slope is smooth crystal type, and very profoundly on temperature. (grass, for instance), the snow cover will slide at When a force is applied to snow it may deform its base as well as flow plastically within itself. elastically, flow as a viscous liquid, or react as a The creep process is usually very slow, and goes combination of the two (fig. 12). The rate and unnoticed by the casual observer; but it can exert duration of force application will also determine tremendous pressure against solid objects located the relative importance of the elastic behavior. within the snow cover. Even on flat ground, set- Snow which is brittle is able to remain in a stressed tlement alone can exert great force. These forces state for long periods, resulting in persistent ava- can be computed, and must be taken into consid- lanche danger. Snow which behaves more nearly eration when constructing such installations as as a viscous liquid is able to relieve stresses rapidly avalanche barriers, lift towers, and snowsheds. by plastic flow, and will more quickly approach a The tensile and shear stresses produced by uneven stable condition. creep of snow are an important factor in avalanche The degree to which snow behaves as a viscous formation. liquid is described by the quantity known as the Variations in the strength characteristics of coefficient of viscosity ; the higher the value of this snow are among the widest found in nature. The coefficient the stifi^er and less liquidlike is the snow. hardness of wind-packed old snow or frozen firn Figure 13 displays in generalized form the de- may be as much as 50,000 times that of light, fluffy pendence of this coefficient on temperature and powder snow. Strength characteristics continu- density. The very great sensitivity of viscosity ally change as a result of metamorphism, and are to temperature in snows of higher density is an also dependent on temperature for a given snow important factor in slab avalanche formation. type. Wlien snow is disturbed mechanically, then The viscosity of snow also increases with increas- allowed to set, it undergoes a process known as ing grain size. age-hardening. This process reSults in a gradual In contrast, liowever, the elastic properties of hardening of the snow for several hours after it is disturbed. The process is more pronounced snow, described by a quantity known as the elastic at lower temperature. The greatest source of modulus (usually the shear modulus) are only mechanical disturbance in nature is the wind, and slightly dependent on temperature. The elastic an increase in hardness is always associated with modulus does increase exponentially with density wind-drifted snow. Age-hardening by the arti- over the range of densities found in winter snow ficial disturbance of snow slopes such as skiing or covers.

11 /5 -

0 ^ /o- Q/ X ^ >/ / n 0/ Í /

1 1 1— -10 -20 -JO -40

T£MPERATURE. FIGURE 13.—The dependence of snow viscosity on density FIGURE 12.—The plastic character of snow has permitted and temperature. As density increases, changes in it to flow off this roof and develop a large overhang viscosity become very sensitive to temperature. without breaking. pendence on temperature. Accordingly, the man- The strength properties of snow, already ner in which temperature of the snow cover is mentioned, are also determined by density, tem- controlled by heat loss or gain is of primary im- perature, and grain type. The magnitude of such portance to its whole behavior. In addition, the properties as shear or tensile strength determines melting of the snow cover each spring is con- the ability of a given snow mass to remain anchored on a mountainside in opposition to the trolled by the supply of heat. An understand- forces of gravity whicli are trying to pull it down. ing of the relative importance of the various The changes of the strength properties with time meteorological factors which affect the heat supply are important in avalanche formation, and, taken is thus important to forecasting snow behavior and together with time changes of stresses, determine avalanches. when the moment of critical instability may be There are three basic characteristics of snow reached. A generalized picture of the dependence which strongly influence its thermal properties. of tensile strength on the three principal variables First is the high heat of fusion of ice already IS given in figure 1-i. Note that the strength mentioned. For each gram of ice existing at the reaches a maximum for fine-grained old snow, and freezing point, approximately 80 calories are re- then decreases as the grain size increases. This is quired to convert it to liquid water without any one of the dangerous effects of constructive meta- morphism and depth hoar formation already change in temperature. The second is the very mentioned. low heat conductivity of snow which ranges from These mechanical properties of snow will be re- about 0.0001 to 0.0007 calorie/cm.-sec./°C., or up ierred to agam m the discussion of slab avalanche to 10,000 times as great as that of copper. The lormation. third characteristic is the presence of water in the 2.5 Thermal Properties of the Snow Cover. vapor and/or liquid phases as well as the solid I tie common characteristic of the various prop- phase. Vapor and liquid water play an important erties of the snow cover just outlined is their de- part in heat transfer within the snow cover. 12 about 1 cm. of ice each year. In any case, an ap- (!) preciable snow cover serves to insulate the ground surface from external heat exchanges; and the ill internal and stored heat, though it may be rela- tively small, is sufficient to melt small quantities V) of snow throughout the winter. The occurrence of melting means that the temperature must be -4 kept at freezing point. For this reason, the snow- earth interface is almost always at a temperature of 0° C. or 32° F. in snow covers throughout the ííí temperate zones. This is not the case m higher latitudes where very low temperatures, shallow -¿o .4.0 snow cover, and permafrost are the rule. The tem- perature may also remain below freezing for a TEMPäRATURE, C short time at the bottom of temperate zone snow- packs early in fall, when there is frost in the ground and the snow is shallow. Heat is both lost and gained in much larger quantities at the snow surface, and the tempera- ture undergoes wide fluctuations below the freez- ing point. The temperature of the snow surface, which is composed of particles of solid ice, can «O never be higher than 32°F. because water cannot exist in a solid state above that temperature. A thermometer placed on the snow surface will often give a misleading value higher than 32° because of solar radiation. The principal media of heat i5 transfer at the snow surface are turbulence of the air, long and short wave radiation, condensation- O. 3 O. & evaporation-sublimation, rainfall, and internal conduction. The air, radiation, and vapor ex- DEN5ITV, y/ce —- changes, and internal conduction may either add or subtract heat, while rainfall can only add it, since liquid water must always be at a tempera- ture equal to or higher than that of ice. These media are shown diagrammatically in figure 15. The medium of heat exchange with the largest potential is the air. Heat may be conveyed to or «O removed from the snow surface in large quantities by the process shown as eddy conduction, or the turbulent transfer of heat. This process depends on the motion of air over the snow surface, and does not occur with completely calm air. Under this latter condition heat is transferred within air only by molecular conduction and in very small quantities. When the air is warmer than the snow DENOemC FINE COAfiiZE surface, heat is transferred to the snow, and as the GRA/N TYPE difference in temperature increases, so does the amount of heat transferred. With air colder than FIGURE 14.—Dependence of snow tensile strength on tem- perature, density, and grain type (generalized rela- the snow surface, heat flow is in the opposite di- tions). The decrease of tensile strength with the rection. The laws governing the turbulent trans- formation of coarse crystals (constructive metamor- fer of heat are not y^i clearly formulated, but it phism) is an important source of instability. is known that the quantity of heat transferred depends on difi'erence between air and surface Tlie heat supply at the bottom of the snowpack temperatures, the wind velocity, and roughness of Í8 relatively limited. A certain amount of heat is stored in the surface layers of earth each sum- the surface. The amount of snow that can be mer, and this heat will melt some of the snow de- melted by a strong, warm wind, such as a Chinook posited during the winter. If cold weather pre- wind, is much greater than can be melted by any cedes the first snowfall, however, some of this heat other natural source, including solar radiation. may already be dissipated by frost. The internal Large amounts of energy arrive at the earth's heat of the earth also provides a small but steady surface in the form of short wave (visible) radia- heat supply at the surface, and is suiRcient to melt tion from the sun, but only a small portion is avail- 577868 O—61- 13 SKV

rO SPA CE

Ti/ic.aot.eMT rJtÂ/tif£A. Of= LOffGH/Â/£ HADIAT/ON

7K j/votr -Jv/ifAcc rfiANSAt/rreo CO/IDUCTIOH t /APOZ TRANSFER.

FIGURE 15.—The principal media of energy exchange at the snow surface. able to snow because of its high reflectivity. Up to of the snow surface. The net balance in winter 90 percent of the short wave radiation arriving at is usually positive (snow gains heat) in the day the surface of freshly fallen snow is turned back and negative at night. Occasionally under con- by reflection. This figure drops to about 60 per- ditions of clear skies at high altitudes in winter, cent for wet spring snow, and may be even lower if when the sun angle is low, the outgoing long wave there is appreciable dirt on the snow surface. radiation may exceed the short wave radiation In contrast to its behavior under short wave absorbed by the snow even at midday, and the radiation, snow reacts to long wave (infrared) snow will actually be cooled while the sun is shin- radiation as though it were black. In fact, among ing. Maximum positive radiation balance is all substances in nature, snow approaches the achieved, not on clear days, but under conditions nearest to an ideal black body for long wave radia- of thin fog or low broken clouds in spring or sum- tion. This means, for one thing, that snow is an mer, when both long and short wave radiation excellent radiator of infrared, and that it can lose contribute heat strongly to the snow surface. heat through this medium as well as gain it. The behavior of water vapor as a medium of When skies are clear, snow loses heat to space by heat exchange at the snow surface is determined long wave radiation. The amount of heat lost by the high heat of vajwrization of water. The depends on the quantity of water vapor and car- conversion of 1 gram of water from liquid to va- bon dioxide in the air, both of which strongly af- por without change in temperature requires 600 fect long wave radiation. Under overcast con- calories of heat. If water is converted directly ditions, snow will also lose heat by long wave ra- from solid to vapor (sublimation), the heat diation to clouds if the cloud temperature is lower fusion must be added to this figure to give a heat than that of the snow surface. Clouds of any ap- requirement of 680 calories per gram. It is thus preciable thickness also behave as black bodies to seen that, even though the heat of fusion is large, mfrared, and if temperatures of cloud and snow the heat of vaporization or sublimation is very surfaces are known, the amount of heat exchanged much larger. Accordingly, addition or subtrac- can be calculated. Wlien the clouds are warmer tion of material at the snow surface through the than the snow surface, the process is reversed and medium of water vapor must be accompanied by the snow gams heat by long wave radiation. This the exchange of large quantities of heat. This condition occurs most often in spring or summer means, first, that evaporation can take place only when the air and clouds are warm and the snow if enough heat is available to supply the heat of remains at 32°. vaporization. Heat in such quantities is not or- The eíFects of long and short wave radiation add dinarily available to the snow cover, and evapo- aJgebraically to produce the net radiation balance ration losses are generally quite small. Signifi-

14 cant amounts of evaporation can occur only when liquid water through a snow cover transports heat extremely dry air with temperature well above very rapidly, and a sudden thaw or rainfall which that of the snow surface is accompanied by high produces free water at the surface v/ill quickly winds, provided tliat the condition for evapora- warm the whole snow cover up to the freezing tion described below is satisfied. A second point. This process would take a long time by corollary is that condensation of water vapor on conduction alone. Wlien water percolates into a snow surface can be a large source of heat, for snow below the freezing point, the water refreezes when condensation occurs, the heat of vaporiza- and gives up its heat of fusion. Heat is thus trans- tion is given up to the snow. The transport of ported through the snow by the physical penetra- M-ater vapor to or from a snow surface is governed tion of water, circumventing the otherwise low by laws similar to those governing the tui'bulent conductivity of snow. transfer of heat, and increases witli increasing 2.6 Examples of Weather Influence on the wind velocity. Snow Cover. A discussion of actual effects of "\'Maether the condition for evaporation or con- weather on snow as obtained from the records will densation prevails depends on a simple physical serve to illustrate some of the foregoing principles. law. "\^nien the dew point of tlie atmosphere is 1. Radiation. The maximum amount of heat lower than the snow surface temperature, conden- available to the snow from solar radiation can be sation can take place. If the air and snow sur- calculated. The following tabulation gives the face temperatures are below freezing, the ex- amount of heat available to a level snow surface change takes place as sublimation. Conditions exposed to the sun from sunrise to sunset at ap- most frequently favor condensation in spring or proximately the latitude of Alta, Utah. summer when air temperatures (and dew points) Clear day in JANUARY cal/cm.' available become higher, wliile the snow temperature re- (reflectivity of snow 0.9) to snoiB Sunrise to sunset 29 mains at 32° F. Conditions favoring evaporation 1000-1400 18 are more common in midwinter, but very little 1100-1300 9. 4 takes place because the available quantity of heat Clear day in APRIL is small. (reflectivity of snow 0.7) Rainfall ordinarily can deliver only a limited Sunrise to sunset 200 quantity of heat to the snow surface. The spe- 1000-1400 93 1100-1300 48 cific heat of liquid water is 1.0; that is, 1 calorie of heat is involved in changing the temperature Tliese figures actually will be reduced some- of 1 gram of water 1°C. Thus, each gram of rain what by various factors in nature, and the heat which falls on a snow surface and is cooled to the losses due to long wave radiation on clear days freezing point gives up only 1 calorie for each will reduce them further. In January these re- degree centigrade it is cooled. It is seen that ductions leave very little heat available to the large quantities of rain are required to melt any snow. In April the increased sun elevation pro- significant amount of snow (ice) requiring a heat vides much more heat from solar radiation, and of fusion of 80 calories per gram. Heavy snow- this source becomes important to snowmelt. melts sometimes associated with rainstorms can A good example of the effects of long wave usually be attributed to the warm winds and con- radiation loss occurred at Alta in February 1954. densation which accomi^any such storms. Very A weeklong period of dry, clear, and calm weather warm rain in large quantities can cause appreci- brought air temperatures as high as 50° F. dur- able melting, but such conditions are rare in the ing the day and no lower than 30° at night. The high mountains, and mountain snow covers sel- snow remained light, crystalline, and fluffy dur- dom experince much ablation (melting) due to ing this warm spell, and even on south slopes there rainfall. was only very limited melting and metamorphism The flow of heat within tlie snow cover is a com- at the surface. Snow temperature 2 centimeters plex phenomenon due to the presence of all three below the surface was 25° even at the end of this phases of water ; part of the heat is transferred by period. Very strong long wave radiation losses simple molecular conduction and part is trans- almost completely neutralized heat supply to the ferred by circulation of air in the spaces between snow even at midday, and cooled it strongly at the snow grains. The air itself carries some heat, night. but the water vapor with it carries more. It is 2. T em pera f me Gradient. The month of Jan- estimated that up to a third of the heat flow in uary 1955 at Alta was marked by many large ava- snow is due to sublimation and diffusion of water lanches. Some of these removed the snow cover vapor. Because of the high heat of sublimation on certain slopes so that only a thin icy layer re- of water, only small quantities of water vapor are mained next to the ground. Late in January a required to transport appreciable heat in this fall of more than 40 inches of powder snow oc- manner. In any case, the total amount of heat curred, and in many places avalanched again off transferred within the snow cover is quite small, this icy layer. One place that it stuck to the icy due to its overall low conductivity. This situa- layer was on the west face of Greeley Mountain. tion is altered when appreciable amounts of liquid AVliere this heavy snowfall was deposited on a water are present. The downward percolation of deeper snow base which had not previously ava- 15 ]nnched it quickly settled into a very hard and co- of rainwater through the cover caused large hesive layer- This settlement also occurred on changes within a few hours. This rain fell at Greeley Mountain, but because the snow layer temperatures close to freezing, and actually melted rested practically on the ground, it was subjected an insignificant amount of SIIOAV. In fact, later here to a much more severe temperature gradient water-content measurements of the snow cover by cold weather early in February than was the showed that more water had been retained by same layer which rested over a deep snow cover. freezing of the rain in the cold snow than had This gradient was high enough to initiate con- been lost by melting. The effects of this rainstorm structive metamorphism, which gradually weak- on snow temperatures are displayed graphically ened the mechanical structure of the snow. As a in figure 16. result, a large and dangerous avalanche came down 4. Turbulent Transfer of Heat From the Air. the west face of Greeley Mountain more than a As might be expected, heavy melting of snow by week after the storm had ended. Given the same warm air often occurs over glacier surfaces in the temperature conditions at top and bottom of the summer. A clear example of such melting was ob- snow cover, and the same snow layer in question, served in July 1955, on the Lemon Creek Glacier the shallow snow cover became dangerously weak. in Alaska. The ablation (melting) season com- Avhile the deeper one became stable. The only dif- menced early in June, and wastage of the snow ferences were the magnitude of the temperature surface amounted to an average of about 5 cm. gradient and the character of the bond to the un- of snow each day through June and most of July. derlayer. However, on July 21-22 a period of clear weather 3. Heat Transsjer Within the Snow Cover. The arrived, accompanied by a Chinook or foehn wind. effects of liquid water on heat transfer within the Air temperature was near 50° F., higher than the snow cover were clearly demonstrated at Alta in daily means up to this time, and high winds to December 1955. Cold and early storms had built 40 mph. were noted. During this 2-day period, up a fairly heavy snow cover that was at tempera- 45 cm. (11/2 ft.) of snow were melted away. Den- tures well below the freezing point. On the morn- sity of the snow was about 0.5 gm./cc., so this ing of December 22, the snow cover was 157 cm. represented the melting of about 22 cm. of water, deep. Internal temperatures were —1.2° C. at 60 requiring 1,760 calories supplied to each square cm., -2.4° C. at 92 cni., and -3.3° C. at 122 cm. centimeter of the snow surface. As soon as the Three-fourths of an inch of rain fell during the foehn blew itself out, the ablation returned to day, and by the morning of December 23, these temperatures had started to rise. Then during the about 5 cm. of snow each day. Similar high abla- day and evening of December 23, 3I/2 inches of rain tion rates have also been observed in this area fell, and by next morning the entire snow cover when storm conditions are accompanied by high was isothermal at the freezing point. Tempera- winds well above the freezing point. Warm winds tures of such points deep in a snow cover ordinar- such as these in the mountains are often the cause ily change only very slowly, but the percolation of extensive spring wet-snow avalanche cycles.

6LTA SNOW STUDY PLOT DISTRIBUTION OF ISOTHERMS FROM SNO» SURFÜCe II DECEMBER 1955 TO 31 MARCH 1956 DEGREES CENTIGRADE

FIGURE lb.—Temperature changes in tue snow cover throughout the winter. The gradual temperature changes üue to conduction and the rapid changes due to melt-water percolation are both shown. The rainstorm of late Decem- ber IS discussed in the text. Slopes of the lines marked A and B indicate the rate of melt-water i>ercolation in the spring. The lower layers of this snow cover had been made very hard and dense by refreezing of water from the December rain, and offered greater resistance to the downward flow of spring melt water.

16 2332.81 SNOW AVALANCHES Chapter 3 3 AVALANCHE CHARACTERISTICS. Ava- between the sliding snow and that which has re- lanches which fall over long distances and varied mained in place is often indefinite. Slab ava- terrain may involve more than one type of snow, lanches, on the other hand, are characterized by or more than one method of motion. To permit internal cohesion of the snow involved. This co- a miiform means of description, the primary clas- hesion may vaiy from only a very slight tendency sification of any given avalanche is always based of the snow crystals to stick together up to ex- on conditions prevailing at the point of origin. tremely hard snow which is broken apart only The description may be modified if necessary to with great difficulty. Eegardless of the degree clarify the character of the avalanche over the of hardness of the snow slab, the common charac- length of its path. teristic is that a large area of snow begins to slide 3.1 Avalanche Classification. Two principal at once, instead of starting as a small point as types of avalanches are recognized. These are does the loose snow avalanche. A well-defined loose snow and slab avalanches. The distinction fracture line where the moving snow blanket is based upon the mechanical character of the snow breaks away from the stable snow is the positive at the point of origin. Loose snow has relatively little internal cohesion and tends to move as a identifying characteristic. This fracture line formless mass. Loose snow slides start at a point forms an irregular and often arc-shaped line or over a small area. They grow in size and the across the top of the avalanche path with the face quantity of snow involved increases as they de- of fracture perpendicular to the slope. A definite scend. A clearly distinguishable sliding layer is sliding layer is usually distinguishable (figs. 17, not always present, and the line of demarcation 18,19).

17 P-496198 FIGURE 17.—Skier-released small dry slab avalanche (HS-AS-2). The arrow points to the tracks where the skier entered.

18 FIGURE 18.—Medium-sized dry soft slab avalanche (SS-N-3) released naturally by a cornice fall. The broken cornice is visible at the apex of the fracture line. The slide originated as a shallow slab which triggered a deeper layer of snow some 25 or 30 feet below the cornice ( arrow ).

19 ^»

■ ^ '* - 'A "* .r . kv ^ ■Ä"»» ^ ^-i* -4 ■•-»• .'*»'' #■»- x.-^-; Â "-

" ■» ^«k »

» ■Í'» >

FIGURE 19.—Large, dry, hard slab avalanche (HS-?^). Manner of release unknown. Fracture line extends under the lee side of a ridge, where heavy wind depo-sition has in places created an exceptionally thick slab (upper center of picture). Presence of angular blocks in deposition zone identifies this slide as a hard slab. The ridge running across center of the picture, the upper edge of the stable snow which was overridden by the falling slab, is a characteristic feature of the slab avalanche.

Slab avalanches are subdivided into soft and unstable equilibrium by: (1) deposition of light, hard slabs. Soft slabs slide away with the typical fluffy snow under conditions of little wind; (2) fracture line, but have the general appearance of a reduction of internal cohesion among the snow loose snow avalanche until they actually slide and crystals by metamorphic changes; or (3) lubrica- reveal the soft slab characteristics. The snow tion from percolating melt water. debris deposited by a soft slab avalanche is Any small disturbance is propagated downward usually a formless mass like that from a loose from crystal to crystal, causing the snow to slump snow avalanche. A hard slab is characterized by downhill towards its natural angle of repose, or, snow of sufficient cohesion to retain its form in if any momentum is attained, to slide onto slopes motion. The principal feature identifying the of lesser angle. Disturbance of the unstable hard slab avalanche is tlie presence of angular equilibrium proceeds in the nature of a chain re- blocks or chunks of snow in the debris (figs. 20 action, confined to the downward direction by and21). gravity and laterally by the character of the snow 3.11 Loose Snow Avalanches. T^oose snow ava- and the terrain. More and more snow is col- lanches arise when snow accumulates on slopes of lected in the sliding mass until sufficiently gentle steeper angle than its natural angle of repose. slopes are readied for the moving snow to lose its The snow on a given slope is placed in a state of momentum and come to rest.

20 H^ -v

r n ^4 i

■4 -^

.^ V ^* •'é^^ «f ,*

^jÊ>»^

' r

< t

FIGURE 20.—Typical debris deposition from a hard slab avalanche (HS-AE-3 or 4). Artificial release by cornice blasting. The snow has sufficient internal cohesion to presers-e the large, angular blocks of the original slab.

The quantity of snow involved in a loose snow- slide depends on the depth of unstable snow, the degree of instability, and the length and incli- nation of the slope. The size ranges all the way from tlie veiy common loose or powder snow sluff during or after a snowstorm to the rare, very large loose slide accompanied by destructive wind- blast. In terms of number, loose slides are prob- ably the most common and generally are minor in size and significance. As will be noted in the discussion of avalanche hazard forecasting, these small, loose snow sluffs are actually a stabilizing factor which inhibit development of the more dangerous slabs. Dry loose snowslides (wild snow) of large size are very rare; most of those which might be so classified actually originate as soft slabs. The most dangerous loose snow avalanches are FiGURE 21.—Debris blocks from a large hard slab ava- those which involve wet snow in the sprino- or lanche (HS-AE^). summer. Originating often high on the moun-

577868 O—61- 21 tain as a tnie loose snow avalanche starting at a cohesion among the crystals to stick together as a point, tliese may gatlier large quantities of snow united mass which can be set in motion as a by tlie time they reach the valley floor. The blanket over some definite area. "\'\nien the inter- weight of the wet snow gives them large destruc- nal cohesion is very low, the slide behaves like a tive power even though the velocity may be low. loose snow avalanche but the dividing line is not 3.12 Slab Avalanches. Tlie slab avalanche is always clear. When the internal cohesion is high, of major concern because it forms tlie most im- the snow layer in question cannot be broken loose portant source of winter hazard in the mountains. and stability prevails. It thus might be said that In fact, a review of avalanche accidents recorded a certain amount of strength in the snow bears both in this counti-y and in Europe shows that with it the seeds of hazard. In practice, the very the dry, loose snowslide may for all practical pur- unstable snow types of low mechanical strength poses be dismissed as a source of hazard. Slab actually offer less source of danger, for they slide avalanches, in most cases triggered by the victims off the mountain long before avalanches of danger- themselves, are responsible for the overwhelming ous size can form. majority of accidents. Assuming that a given snow layer has sufficient Slab avalanches may originate in snow types internal cohesion to behave as a slab, the question ranging from soft, fluffy new snow through com- of whether or not it will slide is answered by the pacted, hard old snow to the hea^'y wet snow of mechanical character of its attachment to the spring. Snow crystals may include almost every mountainside. This character is determined by stage in the metamorphic processes from newly the so-called static relations. deposited stars or dendrites to completely meta- 2. Static Relations. The primary anchorage of morphosed old snow, as well as the forms of con- a snow layer on the mountainside is its bond to the structive metamorphism. It must therefore be underlayer. Strength of this bond depends on clearly emphasized that slab avalanches are not the strength properties of the two layers, the na- exclusively associated with any particular snow ture of the underlayer surface—smooth, wind- type, and especially cannot be correlated with some blown, etc.—the temperature at the interface, and particular visible characteristic of snow. True, the age of the bond. Gravity acting on the slab certain types of fresh-fallen snow, noticeably those layer which lies on an inclined plane develops a of higher density, tend to favor slab formation, component of force parallel to the undersurface. but this is as much a property of their water con- This shear stress tends to break the bond between tent as of their crystallography. Wind drifted the slab and underlayer. Opposed is the shear snow very commonly develops unstable slabs, and strength of the bond, which ordinarily increases wind is one of the predominant natural factors with time in the course of destructive metamor- in slab formation. But wind alone is neither a phism and decreases with rising temperature. necessary nor a sufficient factor to cause these The primary condition of instability is reached dangerous slides. Unstable slabs may develop when the shear stress exceeds the shear strength. through processes of internal change alone. This This may be achieved either by a rise in the stress is especially true of climax avalanche develop- (increased snow load) or decrease in strength ment involving several layers of snow, any one (e.g., constructive metamorphism or crust disinte- of wliich may be stable when considered by itself, but which taken together may fall as a large gration). This stress can be calculated rather slab avalanche. simply if the slab thickness, its mean density, and 1. Mechanical State. It is therefore apparent the slope angle are known (fig. 22). The shear that we must regard the slab avalanche condition strength, on the other hand, cannot be readily cal- as a mechanical state of the snow cover. The culated, and is difficult to measure in the field. tendency of a mow layer to slide is dependent not The general relation for primary instability can only on the character of that layer itself, hut on be formulated, but a quantitative determination its relation to adjacent layers and on the size and cannot be readily made. The natural law which shape of the slope. The primary requirement for governs can be recognized, but numerical values slab formation is snow which has sufficient internal can be assigned only with difficulty.

22 T = nig SIM -©. in load or decrease in snow strength causes one of the anchorages to rupture ; suddenly increased WHEdEm.MASS or UNIT stresses are then thrown onto the other supports VOLUME OP SNOW of the slab and they too break.

- y COMft-t SStON

FIGURE 23.—The secondary condition for slab release is reached when stresses at top, bottom, and sides of the slab exceed strength properties of the snow. If some of the anchorages are sufficiently strong, FIGURE 22.—The primary condition for slab avalanche they may sustain the slab even if other supports release is achieved when weight of the slab exerts a are removed. Tliis behavior is clearly seen in the component of force parallel to the sliding surface (shear stress) greater than the shear strength of the common case of a snow slope which cracks but bond. does not slide. In this case the tensile strength of the slab layer has been exceeded, and it breaks, A shear stress which exceeds the shear strength but bonds to the undersurface, or other supports at of the bond is a necessary but not sufficient condi- sides or bottom, hold it in place. tion for a slab to slide. The snow layer is also 3. Dynamic Relations. The discussion of the anchored to the slope at the top, bottom, and sides static relations assumed that the snow slab was an as well as at the undersurface (fig. 23). The inert mass lying on the mountainside. In practice strength of these anchorages must also be less than this is not the case. The slab and the snow layers the stress applied to them if a slide is to be re- beneath constantly experience the slow downhill leased. As with -the bond between layers, the cri- motion of snow creep, which introduces additional tical point may be reached either by a rise in stress complications to the problem of snow stability. or a decrease in strength. Most important of the It is the additional combination of stresses created anchorages is that at the top of tlie slope where by this creep motion that gives to the slab ava- the snowslab is subjected to tensile (stretching) lanche its dangerously unpredictable and unstable stress, for this is where a rupture is most likely character. to be initiated, either naturally or artificially. Under the force of gravity, snow on a slope is The applied tensile stress can be estimated from continually trying to move toward the valley, both measurable properties of snow and slope, though by sliding on the ground and by internal deforma- not quite so simply as the shear stress mentioned tion. On level ground the only expression of this above. The tensile strength cannot be readily cal- motion is the snow settlement. On a slope the culated, and must be measured in the field, again motion results from both creep parallel to the slope with considerable difficulty. and vertical settlement. If all the forces acting to release the slab ex- Zones of stress are formed when one part of a ceed the strength of the anchorages, it obviously snow slope is free to flow by creep while another will not remain in place. In practice only certain part is anchored and cannot move, or one part of a anchorages may be dangerously stressed, or some slope creeps faster than another. The most dan- or all of them may be stressed close to but not ex- gerous of these zones are those where tensile ceeding the breaking point. Any outside trigger- stresses occur, frequently found near the top of ing force, natural or artificial, which cuts or the slope. breaks part of the slab then initiates a sudden re- The visco-elastic character of snow has already distribution of stresses which may break the an- been described in chapter 2. The reaction of a chorages and release an avalanche. The same snow layer to applied creep stresses depends on thing probably happens when a gradual increase this visco-elastic character. As the viscosity of 23 snow increases, it becomes more and more brittle 4. Layering. The formation of unstable slab (the elastic behavior becomes important only for layers is higldy dependent on discontinuities in hifrli rates of stress application). Sucli snow the snow cover. A completely homogeneous snow isîess able to relieve stresses by plastic (viscous) cover with an even gradation of density, crystal deformation or flow, and may remain in a highly types, and other properties from top to bottom stressed state for a long time. A slab so stressed would not likely fonn a slab avalanche. A pos- is said to be subjected to creep tension. sible exception is where the very strong discon- Wlien some triggering agent breaks a slab and tinuity represented by the ground surface causes relieves creep tension, the reaction of the snow the entire snow cover to slide. Each interruption may be very rapid. A whole slope may suddenly of snowfall, each advent of a new storm with dif- shatter into blocks before the slab has a chance to ferent snow conditions, causes a change in snow be set in motion. If the snow is veiy dense and characteristics from one layer to the next. The cold, such as in hard slab, the release of tension interface between two layers usually will have less may be signaled by a sharp cracking sound like a strength tlian that which exists internally within rifle shot. The propagation of a fracture line in a layer. Tliis creates an opportunity for insta- snow which is subject to heavy creeji tension is bility. It is therefore a general rule that an in- extremely rapid, and such fractures may run for liomogeneous snow cover with shaqi discontinui- many hundreds of yards across a slope faster than ties in stratigraphy is more likely to be the source the eye can follow. of slab avalanche clanger than a homogeneous one. If enough time is available, the snow will The formation of discontinuities in layering is eventually relieve or redistribute these stresses quite sensitive to even minor variations in snow by internal deformation. The amount of time and weather conditions. Frequently a short lull required depends on the viscosity of the snow. or cessation of snowfall in the middle of a storm Snow of low viscosity (near the freezing point) will result in tlie snow of that storm being divided can flow quite readily, and dangerous stresses do into two distinguishable layers. Though the not persist for long periods. Thus direct action snow in each of these layers may be for all pur- slab danger from warm snowstorms (25-32° F.) poses identical, they will have a weakened bond is usually of short duration and the snow quickly at the interface which marked the lull. In such stabilizes. As temperature falls, the viscosity in- cases the upper, more recent layer sometimes can creases rapidly and creep tension becomes more be dislodgecl as a shallow slab while the lower persistent (ñg. 13, ch. 2). This explains the basis layer remains stable. for the general rule that the lower the tempera- The major discontinuities from one snowfall to ture, the longer avalanche danger may persist. the next are of course much more strongly de- A quantitative realization of this behavior may veloped. The exposed snow surface in the in- be obtained by referring to the concept of relaxa- terim between stomis can readily be altered and tion time introduced for the general class of visco- aged by the various meteorological factors. elastic materials by the famous English physicist Many slab avalanches, and especially direct ac- James Clerk Maxwell. If a fixed initial stress tion slides that occur during a storm, avalanche on is applied to such a substance, the time required such a surface. for internal deformation to reduce this stress to The surface of discontinuity from which a slab approximately 38 percent (1/e) of the initial value layer may be dislodged is called the sliding sur- is the relaxation time. For the range of densities face. It is the top of tlie older, stable snow which and temperatures commonly found in nature, the remains in place. Hard crusts, such as those relaxation time for snow may vary from less than formed by sun melt or rain, are particularly good 1 second to nearly 2 weeks. Thus it is seen that, sliding surfaces which may be the source of re- on a seasonal basis, creep action will eventually peated avalanche formation. Tliere often exists cause the snow cover to deform in any case, but it between the sliding surface and the slab layer is during the few days or hours when the snow is proper a shallow, weak layer deposited by a light at first unable to accommodate this creep that the snowfall or surface hoar sublimation. " Such a slab avalanche danger may be at its highest. layer, which provides a region of low shear An illustration of the temperature influence on strength, is called the lubricating layer. slabs is the Ijehavior of hard slabs often found 3.2 Types. Both loose snow and slab avalanches at high altit udes in a continental climate. On such are further classified as dry, damp, or wet, ac- slopes a slab may api:)ear to be quite stable dur- cording to the amount of liquid water present in ing a sunny day, but will become very sensitive to the snow involved. Knowledge of snow condi- triggering as soon as it falls into shadow in the tions at tlie time the slide occurred will help de- late afternoon. The very marked drop in snow termine this, as will a study of the snow debris temperature due to radiation cooling when the and pattern of flow of the 'slide. Damp or wet sun goes down is presumed to increase the brittle- slides usually deposit snow in rounded balls or ness of the surface snow layer. This is then easily lumps as distinguished from the angular blocks fractured, and the break propagates throughout of the hard slab avalanches. The wetter slides the entire slab. A fatal accident occurred on just tend to flow in grooves and channels, often leaving such a slope that had been used all day by skiers. snow ridges between the grooves and displaying 24 the feature of "wheeling," or curving off in dif- Classification of avalanches as to size is based ferent directions, when they reach flatter ground on the observer's estimate of threat to life and (fig. 24). property : Another feature of classification, and also a gen- 1. Large or major—highly dangerous to human eral basis of distinguishing wet from dry snow beings and property. A person caught would be avalanches, is the form of motion of the slide. killed or severely injured. Slides in this class Three general types of motion are recognized; destroy buildings, trees, or structural facilities. motion mostly in the air (dust cloud), motion 2. Äledium—dangerous to human beings but not purely on the ground, and mixed motion where likely to cause property damage. both forms are present. Motion through the air 3. Small—harmless to humans or property. alone is rather rare, and usually occurs when dry Abbreviations for the convenient description of snow flows over clift's or very steep terrain. Mixed avalanches in reports of snow and weather records motion—some on the ground and some in the ac- are listed in chapter 6. (These abbreviations are companying dust cloud^—is the most common form used throughout this handbook whenever a spe- encountered with the dry avalanche of midwinter, cific avalanche is described, especially in the photo both loose snow and slab (fig. 25). Motion mostly captions.) on the ground, with little or no accompanying 3.4 Avalanche Triggers. Direct action ava- dust cloud, is peculiar to the damp or wet ava- lanches occur during or immediately after a storm lanche of spring and summer. Wind blast, the as the result of a new fall of snow. Delayed strong wind winch rushes ahead of the sliding action avalanches do not result directly from a snow, is caused by the dust cloud accompanying fresh snowfall, but are caused by wind action, air motion. temperature changes, melting, or internal struc- Avalanches are also classified according to tural changes in the snow. Climax avalanches whether they slide upon an underlying snow layer are caused by the accumulation of snow from a (surface avalanche) or on the ground (gromad number of storms until an unstable state is reached. avalanche). This latter type ordinarily involves a large part Figure 26 illustrates the more important fea- of the snow cover, and occurs infrequently (fig. tures of various types of avalanches. 27). Their occurrence depends on the character 3.3 Size. The dimensions of a slide would seem of weather and snow cover development. The to be the logical method of determining its size. conditions leading to a climax avalanclie situation But to be of any value, the measurements should on a given mountain may occur only once in 5 or be accurate and these are often difficult, danger- more years. ous, even impossible to obtam. The shape of an When snow reaches a critical state of instability, avalanche is irregular. Wind action, sun action, there often is some definite triggering force which and snowfall quickly disguise it. Dimensions do actually sets it in motion as an avalanche. There are many unstable slabs in the mountains each year not always indicate the momentum, one of the most which never fall as avalanches because no trigger important features of a slide. was present to cause their release. Such snow eventually stabilizes and the hazard disappears. In some instances, one period of critical instability may pass without release, only to be followed by another from a different cause—such as thaw in the spring—wliich starts a slide. External triggers are those involving some dis- ruptive force exterior to the slab or loose snow- layer. Common natural triggers in this class are falling cornices, rolling snowballs, lumps of snow falling from trees, icicles, rockfall, or small slides. '-%2r--. Small avalanches or loose snow sluffs which might otherwise be harmless frequently apply enough dynamic stress on the layers over which they slide .^0t»^^^ to dislodge a larger slab avalanche. Large ava- lanches and massive cornice falls are very effective triggering agents, for they usually develop enough force to carry with them any snow layers which have even the remotest inclination to instability. The involuntary artificial release of avalanches is a common cause of accidents. Most avalanche FIGURE 24.—The path of a large wet snow avalanche accidents involve snow slabs released by the arti- (WS-N^), showing the grooving and channeling of the path and the snow boulders in the deposition zone, ficial external trigger furnished by the victims both characteristic of this slide type. themselves. Many such avalanches no doubt never 25 F-496204 FIGURE 25.—Dust cloud accompanying a dry avalanche (HS-AA-4) released by artillery firing in midwinter, The cloud indicates mixed motion that distinguishes this from a wet avalanche. would fall if they had not been artificially trig- portance of this factor to forecasting techniques gered. Artificial external triggers are often delib- is discussed in chapter 5, item 5.1. erately supplied by skiing, artillery fire, or ex- Tradition has long regarded the sound of a hu- plosives for the purpose of avalanche control. man voice as an avalanche trigger. However, it Changes in the stress-strength relationship of a has not been clearly demonstrated that such a weak slab layer, such as temperature changes, overload- force as the sound waves from voices can cause ing by snowfall, or constructive metamorphism, an avalanche. The possibility cannot be entirely are regarded as internal triggers whicli carry some dismissed because stronger sound waves, such as part of the mechanical relationship of the slab the concussion from explosions, have been observed beyond the breaking point and so effect release. to dislodge slides. It is therefore conceivable that The operation of these agents is less obvious than much weaker sounds could trigger a slide when the external triggers, but they nevertheless are re- extreme instability exists. Evidence, on the other sponsible for the natural release of many slides. hand, seems to indicate that many cases where the fall of an avalanche has been attributed to sound The most common one is overloading by fresh may actually have occurred because the collapse or snowfall. In this connection it may be noted that fracture of tlie snow cover was propagated over the rate of overloading or deposition of a slab long distances. It has been observed that such a layer also appears to play an important part in collapse can travel far, even over level ground. avalanche formation and release. This rate is Local mechanical disturbance of unstable snow, expressed practically by the measurable meteoro- instead of sound waves in the air, may thus at logical factor or precipitation intensity. The im- times be responsible for distant avalanche releases. 26 SLA/i LOOSE 5NOW

TYPE.

OF

3N0¥V

IN THE A/J^. ON THE oeO(/Â/D /f/X£D

TYFe.

OF

Afor/oâi

L :^

Fe.B£ DH.y DAMP tV£T UVATES MO fM.ee WATCH. TitAce o^ f»ee wATee. Fx.ee WATet I'/sibLv nttsettr CQ/VT£HT

(SUZFAC£ (GROUND AVALANCHB) AVALANO/B) SLIDING sa/zFAce

i/nOB/LLYIN« Tt4£ <:B.Oí/NP SNOW LAyem.'x

FIGURE 26.—Avalanche classification chart.

27 mL

:j -'. ^«tr^d ,-\^r -. •*- 4äa:: ..'v

A-:.' •■V^:^^

> -i .^^ ■"'•*»■ 'i • r * ^

P-4B5205 FIGURE 27.—Major rliniax slab avalanche which overran a highway located in the bottom of a canyon (HS-N-5). Origin of the slide was far above the slo¡>e shown in this picture. Smaller slides on this path are normally ar- rested on the bench seen near the top of the picture. Argenta Slide, Big Cottonwood Canyon, Utah.

28 2332.81 SNOW AVALANCHES Chapter 4

4 TERRAIN. There are two fundamental re- to recognize that there is a zone of slope angles quirements for snow avalanche fonnation—snow within which dangerous avalanching is most com- and a mountain for it to slide on. Terrain is the mon. Figure 28 has been drawn with this in mind. fixed factor, important but more amenable to description and classification, for it remains un- changed in any given locality. Avalanche barriers are no more than the artificial modification of terrain. Steep gullies and steep open slopes are natural avalanche paths. Ridges, outcrops, and terraces are natural avalanche barriers. Ridges parallel to the fall line set definite boundaries to an ava- lanche path, while those running across the fall line halt, slow down, or divert the moving snow. As a practical matter, ridges are always the safest route to travel in avalanche terrain. On the other hand they modify the orientation of a slope so that one side may be stabilized by sun and wind while the other is an accumulation zone for slab. Rock outcrops make effective diversion barriers or islands of safety whicli attract the eye of the area phmner as likely locations for structures. Terraces provide a zone of transition where the slope angle changes rapidly. They slow an ava- lanche clown and give it a chance to spread out. Many major avalanche paths include one or more FIGURE 28.—The relation of slope angle in the release terraces which retain most of the annual slides and zone to probability of avalanche release. only avalanche to the full length of the slope un- der unusual conditions. The snow ranger takes Avalanches of large size may originate on slopes full advantage of terrain features in preparing steeper or gentler than 2,5° to 60°, but the proba- and carrying out his area safety plan. bility gets smaller as the slope angle approaches Terrain variations are as wide as nature itself. 0° and 90°. Slope angles at the fracture lines of The following are tlie most important. major avalanche paths within the Alta ski area 4.1 Slope Angle. Critical slope angles for ava- range from .35° to 40°. lanclie formation have been the subject of many Practical field experience by the snow ranger and varied pronouncements in handbooks and leads to a certain intuitive "feel" for the steep- manuals. Several sources have set the arbitrary ness of slope above which slides are commonly re- figure of 22° as the critical angle above which leased in the course of test and protective skiing. slides are apt to occur. Avalanches in fact have Actual check of this intuition against an Abney been observed on slopes as moderate as 15°, though level suggests that the common danger zone lies this is an unusual condition. The proljability of above .30-33°. avalanching increases with slope angle up to a 4.2 Slope Profile. Slopes whose profiles are certain steepness, and then decreases as slopes convex in tlie vertical plane definitely favor slab get steeper and approach the vertical. Tlie reason avalanche formation (fig. 29). The reason for for tliis decrease is the failure of snow to adhere this can easily be understood in terms of slab in large quantities to extremely steep slopes; it mechanics discussed in chapter 3, item 3.12 for simply runs off in small, harmless sluffs before any tlie increase of angle downslope on a convex sur- great depth can be accumulated. Occasionally, face results ill an increase in creep velocity down- certain snow conditions will permit the forma- slope. The tensile stress from differential velocity tion of avalanches on very steep slopes, but this is thus added to that between creeping and an- is not the general rule. It does not seem reason- chored snow. On concave slopes, differential able to draw any sharp dividing lines, but rather velocity tends to form compressive stresses which

577868 O—61- 29 oppose the basic tensile stress near the top of the and offers no further stabilizing influence for sur- path This should not, however, be construed to face avalanclies. . . mean that avalanches are unlikely on a concave Heavy stands of timber are usually positive as- slope, for many large avalanche paths are con- surance of stabilized snow. Avalanches do not cave in vertical profile. Actually, the concavity ordinarily originate in a thick stand of trees high (rate of change of slope angle) near the top of enough to remain well above the maximum snow natural concave slopes is frequently low. Since depth, though loose snow slides here are possible such slopes do not differ greatly from a plane under the rare condition of extremely unstable surface, the stress modifications due to slope curva- wild snow deposition. Scattered lar^e trees are ture are small. Creep tension is apt to be more not proof against avalanche formation. Open, strongly developed on convex slopes. This may alpine parklands, even where many large trees are also be true of slopes which are convex in hori- present, are the sites of numerous avalanche paths. zontal profile. An avalanche fatality in Colorado in 1960 oc- 4.3 Ground Cover and Vegetation. Koughness curred in just such terrain which, though the of the ground surface has two important effects. many trees made it appear safe to the casual ob- One is to detennine how much snow is needed to server, was long known as an active avalanche fill in the terrain irregularities and present a path. smooth sliding surface; the other is to control the Heavy timber inhibits the release of avalanches, amount of creep motion of the snow over the but is not necessarily a protection against slides ground and the ability of the snow to avalanche falling from open slopes above. Well-developed on the ground itself. avalanche paths originating above timberline usually have obvious paths cut through timbered slopes below, where forest growth has been pre- vented by repeated slides year after year. Oc- casionally an abnormal situation will bring about large climax slides which break new paths through the heaviest timber. Traces of destruction from such slides can be found in most mountain areas of the Western United States. The catastrophic avalanches in Switzerland during January 1951 furnished several examples of unusual climax slides which fell through heavy timber where none had been observed for hundreds of years. The presence of trees on a slope also has an important influence on wind currents and the dep- osition of snow during storms. This influence may either inhibit or encourage avalanching, de- pending on relation of the trees to a given slope and prevailing winds. Artificial control of snow deposition is possible either by cutting the trees or duplicating their effect with snow fences. 4.4 Slope Orientation. Orientation of a slope in respect to the sun and prevailing storm winds FIGURE 29.—The influence of slope profile on stress dis- tributions within the snow cover. has a prime influence on avalanche development. The amount of heat the snow surface receives from Smooth, grassy slopes favor avalanche forma- the sun is directly dependent on both orientation tion, for very little snow is required to smooth and slope angle. The importance of heat transfer over the irregularities. Wet grass also provides to snow behavior has already been emphasized in an excellent sliding surface which encourages the chapter 2. Slopes w^hich face the northern quar- release of ground avalanches. Snow creep is par- ter receive very little energy from the sim during ticularly rapid over a grass surface. the winter, and those above a certain angle, de- The opposite extreme in ground cover is a field pending on latitude, receive no sunlight at all dur- of large bouldere which may take 6 feet or more ing part or most of the winter. Such snow sur- of snow cover before a smooth sliding surface is faces receive maximum cooling by radiation losses, presented. Creep is low on such a slope, and thus remaining colder and in a less metamorphosed ground avalanches practically unknown. state than snow that can be warmed by the sun. Certain types of buslies—willows, for example— North exposures are the most favored locations provide a stabilizing influence early in the winter. for depth hoar formation early in the winter. In They effectively retain the snow while it is still general, avalanche danger from the dry slides of shallow. Such light vegetation, however, is winter can be expected to be most frequent and quickly flattened by snow creep and settlement. persistent on north exposures, other factors such By midwinter, or earlier, it is usually covered as snow deposition being equal.

30 South exposures, on the other hand, receive the full benefit of solar radiation. Surface of the steeper slopes lies nearly at right angles to the rays from the smi when it is low in the winter sky, so that maximum concentration of radiant energy per unit area is obtained. Destructive nietamorphism and settlement are thus acceler- ated by higher snow tem^ieratures, and snow sta- bility is more easily achieved in the winter snow cover. As the sun rises higher in the sky with the approach of spring, heat supply increases to the point where substantial melting occurs, and then south-facing slopes become the site of in- stability and wet snow avalanching (fig. 30). The lee slopes, those which receive the heaviest deposition of snow by winter storms, are fre- quently the site of avalanche paths. In such loca- tions the snow lies deeper, accumulates rapidly during storms, and is most apt to be deposited as unstable slabs. In addition to this combination of favorable factors of avalanche formation, lee slopes are often overhung with cornices which may fall and provide effective triggers. Those parts of a mountainside which lie to the lee of prevailing winds are thus very likely to be the most dangerous avalanche slopes. This applies to the individual lee aspects of gullies, ridges, and other terrain features subordinate to the principal mountain faces. Such lee slopes may receive snow deposits even during fair weather by wind drifting. Windward slopes ordinarily are much less likely to avalanche, for they receive less snow deposition, '■j^^- and this snow is apt to be more firmly compacted by the wind. Wind pressure no doubt plays a part in this compaction, but the cementing of snow crystals by rime during storms is also considered an important factor in stabilizing windward slopes. The windward exposures should not be considered as completely avalanche proof, because F-495208 under certain conditions of heavy snowfall, slides FIGURE 30.—A small, wet, loose snow avalanche (WLS- N-2). Released on a south exposure by natural sun- may occur without respect to slope orientation. ball activity.

31 2332.81 SNOW AVALANCHES Chapter 5 5 AVALANCHE HAZARD FORECASTING. and wet slides may be just as dangerous when The "art" of forecasting is not sufficiently ad- they fall, but they occur less frequently. vanced to predict the exact time a slide will re- There is absolutely no substitute for field ex- lease on a given a^-alanclle path. It is possible to perience as training in avalanche hazard forecast- estimate the probability that slides will occur on a ing. A thorough practical knowledge of snow be- slope. Such an estimate is a hazard evaluation. havior and an understanding of the basic princi- When the effects of snow and weather develop- ples outlined in this handbook are essential to suc- ments are projected into the future, the evaluation cessful hazard evaluation. Moreover, each region is a hazard forecast. and locality has its own peculiarities of terrain and Under many conditions, especially those for di- climate which must be learned by actual observa- rect action slab avalanches discussed below, the tion. The principles of snow behavior are the same period of high hazard (high probability of slide everywhere, but the individual variations are occurrence in an area) can be evaluated with rea- endless. sonable accuracy. These periods can also be fore- 5.1 The Contributory Factors. The analysis cast with good accuracy if the weatlier trends— of contributory factors is a method of estimating particularly siorm behavior—can ])e foreseen. the development of avalanche hazard. This The practicing avalanche forecaster will soon method applies primarily to direct action dry slab learn that this "if" is the biggest stumbling block. avalanches which fall during or immediately after The application of fundamental knowledge of a snowstorm. Delayed action climax avalanches snow and avalanche behavior, qualified by local ex- and the wet slides of spring are not readily pre- perience, often makes it possible for the fore- dictable from a contributory factor analysis; caster to say with confidence, "If certain weather danger from these can be foreseen only by the conditions persist, large avalanches will almost careful study of structural, crystalline, and tem- certainly fall": or "If a tliaw does not set in with- perature conditions within the snow cover. in the next 24 liours, the snow may be considered In the Western United States, weather and stable." Xeither he nor the professional meteorol- climate produce three alpine zones, each with its ogist can say with an equal degree of assurance, dominant type of avalanche : "Precipitation intensity will be 0.15 inch per hour The High Alpine Zone, such as the Colorado tonight," or "Heat balance of the snow cover will Rockies, is characterized by comparatively low not exceed a given number of calories per square temperatures, light and very dry snowfall, very centimeter per hour this afternoon." Unfor- strong wind action, and hard slab avalanches tunately the art of mountain weather forecasting, associated with depth hoar. especially wliere precipitation quantities are con- The Middle Alpine Zone, such as the Wasatch cerned, lags behind that af avalanclie forecasting. Range in Utah, is characterized by fairly cold The degree to which a snow ranger is able to fore- temperatures, hea\'7 and dry snowfall, strong cast avalanche hazard largely depends on his wind action, and dry, soft slab avalanches. shrewdness in estimating mountain weather. In The Coastal Alpine Zone, such as the Cascades actuality, his forecast is more often an evaluation in Oregon and Washington, is characterized by of present conditions. Mindful of the adage that moderate temperatures, heavy and damp snow- "only fools and amateurs ti-y to predict mountain fall, strong wmd action, and damp soft slab ava- weather," the avalanche forecaster is in an uncom- lanches. fortable position. Nevertheless, "hazard forecast- The zones overlap and may produce slides hav- ing" will continue to be referred to in this chapter ing similar characteristics, but the dominating wliile recognizing the limitations of this types of avalanches are distinct enough to warrant expression. identification of these zones. Understanding of the various contributory fac- The forecasting factors which follow were de- tors used to forecast direct action dry slab ava- veloped primarily in the Middle and Coastal Al- lanches has improved. Experience has demon- pine Zones. strated that direct action dry slabs represent the Experience at the Berthoud Pass avalanche greatest source of hazard in ski areas. Tliis is the station in the High Alpine Zone has shown that type of avalanche which is encountered through- the critical values of some of the factors listed be- out the winter, storm after storm. Delayed action low have to be modified for the conditions prevail- 32 ing at high ahitudes and low temperatures. more likely to contribute to avalanche formation Storms in this zone generally bring higher winds when fresh snow falls, because the intervening and lighter snowfalls than those in the other zones. layer of loose snow provides a lubricant between The result is a tendency for localized formation the new snow and the crust. New snow may some- of hard slabs instead of the more widespread soft times make a firm bond to even a very hard crust ; slab development encountered when snowfalls are whether or not this liappens is controlled prima- heavier and the areas in question are nearer to or rily by new and old snow temperature. If the below timberline. new snow falls at temperatures near the freezing The ten factors which are used in this method point, or if the old snow surface layers are warm, of hazard forecasting have been selected empiri- then a good bond is possible. A surface of very cally on the basis of experience and snow, storm, loose or fresh snow may sometimes form a poor and avalanche records over the years. There may bond, while firm or settled old snow is more apt to well be other factors, or aspects of the present offer a good bonding surface. Extremely rough ones, which have not yet been clearly recognized. windblown snow (sklaver) offers sufficient ir- For instance, the exact type and percentages of the regularities to provide bond for the new snowfall, different kinds of snow crystals in a fresh snow- so that slide formation is hindered. fall are believed to have some bearing on the for- The character of the old snow also determines mation of snow slabs, but this relationship has not the amount of support it can give to subsequent yet been exactly determined. At present the prac- fresh layers, and whether or not it will itself be- tice is to give all ten factors somewhat equal come involved in slides originating in the new weight in the analysis. Actually, they do not all snow. Unstable slabs in the old snow may in have equal effect, "but the relative importance of themselves be the primary source of an avalanche, the different factors has not been measured. for the weight of the new snow may provide a Experience has shown that this method of haz- trigger for their release, even if the new snow is not ard forecasting is reasonably accurate. A general inclined to slide. Weak or ¡looi'ly bonded layers, direct action avalanche cycle almost always coin- old crusts, or air spaces within the old snow may cides with a predominant number of factors favor- all be considered favorable to avalanche forma- ing avalanche formation; and, conversely, when tion. A particular type of weak layer which is most are unfavorable, a general slide cycle seldom especially disposed to cause avalanching is depth is observed. In practice, the observer's analysis hoar (chapter 2, item 2.3). This type of crystal should always be checked by actually testing the forms a very weak and fragile mechanical struc- snow on steep slopes, either by explosives or by ture and offers little support to successive layers skiing, to see if unstable conditions actually are of snow. Even a very shallow layer of old snow developing. This is particularly true in border- which consists mainly of depth hoar may induce line situations Avhen the contributory factor ana- avalanching. lysis does not give a clear indication one way or 3. Neio Snoiö Depth. The single most impor- another. tant cause of direct action avalanches is a large 1. Old Snow Dejyth. The primary contribu- quantity of new snow. Deep falls may not neces- tion of the old snow, without reference to its char- sarily produce avalanching, and even small quan- acter, is to fill in the terrain irregularities and tities of new snow can cause large slides under provide a smooth, even surface upon which an the right conditions; but as a general rule, deep avalanche may start. The amount of snow it takes new snow is a warning of possible hazard. The to cover up the irregularities—rocks, stumps, deeper the ne^y snow, the more likely that unstable bushes, etc.—obviously will depend oh the nature conditions exist. New snow of a foot or more of the terrain, so the exact value of old snow depth may cause avalanches of significant size. New which begins to contribute toward avalanche for- snow of less than a foot does not usually build mation will depend on the locality. On smooth, up a general hazard, except in the High Alpine open grass slopes as little as 6 inches might be a Zone where 4 or .5 inches of new snow with high favorable factor, while in a field of large boulders winds may produce a general hazard. A few it may take 6 feet before everything is covered inches accompanied by high winds may produce and a smooth snow surface is present. In the localized danger from slabs on lee exposures in usual mountain terrain a depth of 2 to 3 feet might the other zones. be taken as an average value. The general rule in -t. New Snoio Type. The type of snow crystals evaluating this factor is that the deeper the old which make up new snow considerably influence snow, the more likely avalanches are to form. the type of avalanche which may develop or However, given the proper type of old snow, even whether any develop at all (chapter 6). Pellet a few inches might be a dangerous indication. snow, or graupel, for instance, is peculiarly dis- 2. Condition of the Base. The nature of the old posed to form slab avalanches, both because it is snow surface determines in part whether a new usually denser than normal snow types and be- fall of snow will form a good bond. A crust on cause It easily forms a stiff and slablike layer A the old snow may provide a smooth sliding sur- snowfall composed entirely of needles also leads face and favor avalanches. A crust covered with to a dense slab formation, while granular snow a few inches of unconsolidated light powder is (fine, rime-coated crystals) is another type asso- 33 ciated with this kind of nvalanche. The fluffy, inch per hour, accompanied by winds above the feltlike snow built up primarilj' from stai'crystals critical level for an extended period, has a high and spatial dendrites does not usually leací to a correlation with avalanche occurrence. Short pe- slab formation unless subjected to heavy wind riods of high precipitation intensity do not neces- driftinij. However, there appears to be rehation- sarily lead to avalanches, but the deposition of 1 ship between this snow type and soft slab forma- inch or more of water (as snow) at a high rate is tion when there is some rime coated on the crystals. a favorable factor. Extensive direct action slab Stellar or dendritic snow, if deposited in appre¿:: 4^mS^ in a given situation. The liiffher the "liazard factor," the stronger the indication is tliat avalanches may be expected. '>. A hazard factor falling in the midrange;—40 to 60—would indicate a borderline situation, or limited slide activity ; while a high number—75 or more—would be a definite warning of impending hazard. The analysis of contributory factors is a useful aid in estimating avalanche danger; but it does 0- not give the whole answer nor a positive one. Tliere is no substitute for actual testing of snow conditions in the field; and decisions involving public safety should, whenever possible, be based on such tests as well as on general hazard fore- casting procedures. 5.2 Delayed Action and Climax Avalanches. These are the avalanches which fall as the result of cumulative factors working over longer time intervals than those associated with single storms. The most common type of climax avalanche prob- , ■ i. ably is that which occurs because successive snow accumulations finally overcome the support of some weak layer within the snow cover (fig. 31).

Successful assessment of approaching danger from F-495207 such slides depends on the recognition of mistable FIGURE 31.—Major climax avalanche (HS-N-5) which configurations of snow stratigraphy and the proc- crossed a small canyon, covered a highway, and as- esses wliich weaken the snow, especially construc- cended 300 feet up the opposite slope. Channeling in tive metamorphism. the deposition cone indicates damp snow was involved in this slide. Scale is given by the tractor in the Dangerous stratigraphie weaknesses cannot be foreground, working directly over the highway. Coal- recognized by casual inspection. Excavation of a pit Slide in Little Cottonwood Canyon, Utah. pit is one means of gaining insight into the snow structure, but to be most useful, such a pit should depth hoar formed by periods of clear weather be dug near the release zones on the avalanche early in the winter when the snow is shallow and paths. The ram penetrometer discussed in chapter easily subjected to steep temperature gradients. 6, item 6.3, can be used to obtain part of the neces- This condition is most common on northerly ex- sary structural information without digging a pit, posures which receive little sunlight. The mechan- but evidence gathered by this method is far from ically weak layer of depth hoar near the ground conclusive. The day-to-day record of snow, is unable to support a very large overburden, so weather, and avalanche conditions throughout the that the first heavy snowfall is likely to trigger winter in a given area is another means of deter- widespread avalanching. If the depth hoar is mining whether an unstable condition may be only partially formed (incomplete constructive forming. The most satisfactory information metamorphism), the layer may not fail mechani- comes from a combination of all thi-ee methods, cally until several storms have deposited their estimates from the daily observations being loads. Once an avalanche occurs in such a condi- checked by periodic pit excavations and by ram tion, exposure of the remaining depth hoar next profiles on the actual avalanche paths. This ideal to the ground sometimes results in a hard crust is achieved only in certain areas where trained ob- formation which provides an excellent sliding servers are on the groimd all winter to observe and surface for subsequent avalanches. record snow and weather developments. Else- Formation of a smooth crust or ice layer next where, climax avalanche danger can only be to the ground early in the winter by rain or thaw crudely estimated, especially if the history of snow is another frequent source of climax avalanches. and weather conditions is missing. This layer provides a good sliding surface, but A knowledge of early winter weather conditions unlike depth hoar, has high strength. The criti- is essential to anticipate many climax slides. The cal factor then becomes the character of the bond character of the lowest layers of the snow cover between the ice layer and subsequent snowfall. and their subsequent evolution largely determine Such a bond, though weak, is generally stronger the stability of the snow cover as a whole. The than the support offered by depth hoar alone, and most frequent cause of instability is a layer of accumulation of much snow may be required be-

36 fore critical instability is reached. An ice crust A series of light snowfalls, especially at low tem- exposed by avalanching may also serve as a sliding peratures, will form a much more unstable stratig- layer once again as more snow accumulates. Like raphy than will heavy snowfalls. In the former depth hoar, it can be the cause of repeated ava- case, the bond between snowfalls is often poor, lanche action, either direct or climax in character. and tlie individual layers never develop very high By late winter or early spring rising temperatures mechanical strength unless subjected to rework- eventually cause a bond to form between new snow ing by wind drift. Very deep falls of dry, loose and such layers, so that the hazard from this powder snow usually attain high mechanical source diminishes as the season progresses. strength once they have undergone settlement and The existence of marked inhomogeneities, espe- incor]5oration within the snow cover. cially those caused by weak layei-s or crusts, is apt 5.3 The Ram Profile. Snow and weather ob- to introduce the possibility of climax avalanches servations, snowpit techniques, and operation of at almost any stage of snow cover evolution. the ram penetrometer—all necessary for the esti- Tliose of early winter have been emphasized be- mation of climax avalanche hazard—are described cause they are most frequently associated with the in detail in chapter 6. The following paragraphs large and destructive avalanches. illustrate some of the principles used in interpret- A reduction of tensile or shear strength within ing ram profiles in terms of potential avalanche certain snow layers can lead to eventual avalanch- danger. These interpretations are based on work ing of layei-s which may at one time have been conducted at Berthoud Pass. entirely stable. The most common form of this At Berthoud Pass the studies have been focused weakening is that resulting from melt water in- on identifying the characteristics which indicate trusion, details of which are discussed under item a snow slab that will fracture and become an ava- 5.4. At this point we are concerned with the more lanche. The principal test area is "The Koll," a subtle influences of constructive metamorphism slope prone to the formation of slab in snow trans- and crust disintegration. ported from the surface by wind. The internal diffusion of water vapor and re- During one observed 36-hour period, wind aver- crj'stallization which characterizes constructive aging 27 m.p.h. deposited 6 feet of snow on "The metamorphism leads to reduction in strength Roll." This high rate of deposition indicates that properties of snow, whether or not the process is a delayed action slab avalanche condition can de- carried through to the end product of fully de- velop in a very short time in areas similar to the veloped depth hoar. A reduction in tensile or test area. In fact, the development is so rapid shear strength so effected can create dangerous in- that stabilization by skiing could not be effective stability in a snow slab; hence this weakening unless done day and night. Incidentally, the test process must always be watched as a possible area is an area of concentrated public use. source of hazard. As pointed out in chapter 2, a Tlie following profiles and inteq)rotations illus- steep temperature gradient is the prime cause of trate the methods used to reach certain prelim- constructive metamorphism, so the temperature inary conclusions regarding the stability of slab influences on the snow cover must be accurately as revealed by the penetrometer. The profiles estimated if hazard development is to be foreseen. were all taken directly in the slide path either Long periods of cold weather without precipita- just before or just after the occurrence of an ava- tion should be a warning sign, as should periods lanche. In two cases the avalanches were trig- of clear weather which permit strong radiation gered by explosives. In one case blasting pro- cooling on shaded slopes. If the snow cover is duced extensive fracturing with stabilization in deep, strong surface cooling will usually affect place. In two cases avalanches were produced by only the surface layers. These might either be- protective skiing. come dangerous in themselves, or else serve as a 1. Interpretation : Chart No. 3 {fig. 32). Pro- weak layer, inadequate to support a subsequent file was taken in the Berthoud Pass test area. The snowfall. When the snow cover is shallow, steep sections of particular interest are marked J., 5, C, temperature gradients throughout the snow cover and D. A was slab deposited during a storm when are possible. Even when winter is well advanced, there was wind action on snowfall and on snow the latter conditions may be found on avalanche transported from the surface. B was a weak layer paths whicli have lost part of their snow cover by of older snow. 0 was an old comparatively thin sliding. The surrounding undisturbed snow may slab of high resistance. D was a very weak layer be deep enough to forestall development of strong of cup crystals. gradients, but this will not be the case on the Under explosives, the low-resistance slab layer partially denuded avalanche path (chapter 2, item A fractured. Tlie break penetrated weak layer B 2.6, example 2). Significantly, high-resistance layer C withstood Icy crusts tend to disintegrate under these same the shock of tlie explosion and supported the vi- influences. The individual crystals lose their bration and weight of the avalanclie until it had strong cohesion to one another and become loose attamed full momentum. Layer C eventually col- grains which may provide a lubricating layer for lapsed and the fracture penetrated weak layer D a slab resting on them. to the ground.

577868 O—61- 37 CHART NO S CHART NO 5 PENETROMETER TEST PENETROMETER TEST ON "ROLL" ABOVE SUOE OF DEC 15, 1951, BERTHOUO PASS TOP OF CLIFF - 200 YDS NORTH OF TOWER NO 3, BERTHOUO PASS UECEMBER l6,1951 JANUARY 5,1952

RESISTANCE TO PENETRATION (KILOGRAMS) RESISTANCE TO PENETRATION (KILOGRAMS) 60 0 1 C 20 30 40 50 0 i:^ 1

1 1 20 Snow deplh = 76 inches ■10" Woter equivalent = 24 Averoqe deosily 3 1

Steeper portions slid lo ground w

'zo" "L, Í ^ ■30"

1 -40" 1 1 1 i , |I20 ^ -50"

VI cL H

-60- s

I6Ü

1 j

-TO'

Í 1 Measurements bf Sttllman, Cram Bor lona f^ »round s^ SVMIIIII'y'/V<<\\\MllllfoV/X 'yVWSJ Mil Y///^/K FIGURE 33.—Chart No. 5, penetrometer test. FIGURE 32.—Chart No. 3, penetrometer test. This profile illustrates several important points. upper layer avalanched. The observer concludes The snowpack contained two slab layers, each that the lower layer stabilized in place due to its lying above a zone of minimum support. The strong anchorage at the toe of the slope. It was upper slab layer had low resistance, the other high buttressed by the debris from several earlier resistance. The low-resistance layer was 33 cm. slides. thick, the high-resistance layer only 13 cm. thick. Chart No. 8 indicates that only the upper layer Layer C was subjected to a blast and a moving avalanched at first. Eventually the lower, high- load which amounted to 108 pounds per square resistance layer did break and carry with it the foot, deadweight. Nevertheless it was able to remainder of the snowpack. 5.4 Wet Snow Avalanches. Of the various maintain its position on a 40-degree slope until the types of slides, those originating with wet snow avalanche from layer ^4 attained full speed. sometimes may be predicted with the greatest pre- 2. Interpretation: Chart No. 5 (fig. 33). This cision as to time and place, because their formation profile was taken in an area similar to the test area, depends on measurable formation of melt water and located near it. To the observer it indicated and its penetration into the snow cover. The a situation comparable to that on Chart No. 3— absence of wet slide hazard can be rather firmly a slab layer of low resistance over a zone of little established by the observation of snow tempera- support. Under explosives, the area avalanched tures below the freezing point. The presence of to the ground. liquid water in snow means that its temperature 3. Interpretation: Charts Nos. 7 and 8 {ßgs. 34 has been raised to 32°F. Cold (subfreezing) snow and 35). These profiles illustrate similar condi- will not contain water in the liquid phase and does tions in two areas of similar aspect but a number not participate in the origin of wet slides. Snow of miles apart. Again we have two slab layers, temperature readings provide a useful guide to each lying over a zone of weakness. In both cases an avalanche was triggered by wet snow avalanche hazard, but the measurements protective skiing. Chart No. 7 shows that the should be made at the fracture zone to be effective. fracture penetrated both slab layers but only the The positive heat balance of the snow cover which 38 CHART NO 7 CHART NO e PENETROMETER TEST PENETROMETER TEST UPPER SLOPE OF ROLL, 8ERTH0U0 PASS LOVELANO PASS, SEVEN SISTERS,- 7rh SISTER AT TIMBERLINE MARCH I7,|9SI MARCH 7, 1951

RESISTANCE TO PENETRATION (KILOGRAMS) RESISTANCE To PENETRATION (KILOGRAMS)

SNOW SURFACE 20 40 60 80 100 120 -0 40 80 120 160 1 L_ , ? -10- c i= 1 ^ Tills portion Gild March 7,19^2 Snow slid to here caused by observer / Caused by skier /^ digging out sno« sloke (Slide occurr- -30- ^1 ed Oder )5 minutes of digging! £' .J

-L "1 ^,^Shd "1

-60' Í Uj .^ r J=' -TO'

L 1 -80' L —1 t

'90*

■|00' J

Ueosurements by Taylor, Weich, Borland

[unabi to defer mm« SnoM Dap h e OCtly. Is t>y g"" I ——/ 1 II »a about t5 feet } Povio Ck, Bronsc ", BorlOfitJ ■ 1. FIGURE 35.—Chart No. 8, penetrometer test. FIGURE 34.—Chart No. 7, penetrometer test. leads to melting may differ greatly between valley should be noted. In particular, observe that the heat balance at the snow surface may be more floor and ridge top, between north exposure and strongly positive on a hazy day than on a clear south. Moreover, penetration of percolating melt water can alter snow temperatures very rapidly. day, and that warm winds are an especially eflFec- tive source of lieat. These two conditions some- A slope which appears safe because of cold snow one hour may be ready to avalanche the next. times combine to form very marked thaw Aveather which may precipitate slide action. Remember Wet snow avalanches are directly associated that wet snow action on south exposures is often with a source of liquid water at the snow surface absent during clear, cold weather, even though which can wann up the snow and then lubricate it sunlight is intense, but a slide cycle can follow to reduce internal cohesion, or the adhesion be- when the weather turns warm, windy, and cloudy. tween layers. An obvious source of liquid water is rain. Rainstorms in winter or spring should al- South exposures receive the full benefit of direct ways be regarded with suspicion until the .snow sunlight and are the ones most commonly asso- has either slid or proved itself stable. Warm rains ciated with wet slide activity. North slopes are in midwinter immediately following a heavy fall not immune, because a warm Avind or rainstorm of dry powder snow are a common cause of major will affect all slopes alike without regard to orien- avalanche cycles in the Coastal Alpine Zone of the tation. Solar radiation has one additional prop- Cascades and Sierras. Heavy melt at the surface erty which makes it especially effective in warm- is the other common source of liquid water which ing and melting snow. The sun's rays can pene- may precipitate slide action. A review of the trate the snow surface to a depth of 4 to 8 inches discussion on thermal properties of the snow cover and cause subsurface melting, even at times when in chapter 2 is recommended at this point. The the snow surface may be maintained at subfreez- major sources of heat which might cause melt ing temperatures by cold air.

39 Dangerous wet slide activity occurs most com- even if new snow falls on settled, stable old snow. monly when a sudden thaw or rainstorm affects Some of the largest and most destructive wet a cold, unsettled snow cover. The very rapid avalanches are the gromid avalanches which re- metamorphic changes and melt water penetration sult from prolonged thaw or rain. The entire in dry loose snow create serious instability. The same amount of rain or thaw that would generate snow cover is thawed and permeated with melt large slides in soft, new snow would be less likely water, which lubricates the ground that serves to cause trouble in settled, old snow which had as a sliding surface. The total depth of the win- already been warmed to the melting point by pre- ter snow cover is then released as a wet slab ava- vious thaws. The first heavy thaw of late winter lanche. Such slides are most common on smooth, or early spring is likely to be the most dangerous grassy slopes or smooth rock which provide a one. Tlie same conditions might be repeated later good sliding surface when properly lubricated. in the season after a fresh fall of new snow. Wet In some situations a heavy ice layer located deep surface slides are likely to occur when a late spring in the snow cover may serve the same function as snowstorm is followed by a bright, sunny day, a smooth gromid surface.

40 2332.81 SNOW AVALANCHES Chapter 6

6 STANDARD SNOW OBSERVATIONS meteorological shelter may be used. Tliis instru- AND TERMINOLOGY. The importance of ment does not give the accuracy of thermometers. standardized, systematic observations of snow and Maximum and minimum temperature readings weather phenomena cannot be too strongly em- are taken once daily and at the same time every phasized. Estimates and forecasts of snow and day. The recommended time is in the late after- avalanche conditions are based on the total experi- noon, but early morning readings are acceptable. ence of the observer and of others wlio have pre- The maximum and minimum temperatures are en- ceded him. To permit comparison of present con- tered in degrees Fahrenlieit each day on the Snow ditions with this accumulation of knowledge and Studies Chart. past experience and drawing valid conclusions, 2. Wind. The preferred information is the observations and records must all be expressed in number of miles of wind past the station each 24 the same terms and the same units of measure- hours, to be entered in the 24-hour total column. ment, and to the extent feasible, measured and re- Either a dial anemometer which displays miles of corded in the same manner. One observer's ex- wind directly, or a continually operated wind re- perience and knowledge must be recorded in a corder is required to obtain this information. The manner that permits the next man to compare the dial is the easiest to operate but requires that the results with his own observations if the informa- observer climb up the anemometer mast to get the tion is to be useful to future workers in the field. reading. A remote recorder can be read inside of Systematic terminology and methods of observ- a building but is more complicated and requires ing and recording snow data liave gradually been continuous power. evolved by Forest Service snow rangers working When an indicating or recording anemometer is every day in the field. The records they are ac- not available, the observer must enter his estimates cumulating are gradually being brought to a for the different periods of the day under the 6- single standard. It is the purpose of this chapter hour average velocity column, and indicate the to instruct Forest Service officers in these uniform direction. Velocities are given in miles per hour methods of observation. and estimates may encompass a range of velocities, 6.1 Snow Studies Chart (fig. 36). The funda- such as "3-5 mph." or "10-15 mph." mental record is that of the daily observations of "\ATienever terrain, building location, and other weather, snow, and avalanche conditions entered considerations permit, the wmd record should be in the Snow Studies Chart ( fig. 36 ). The methods obtained from an anemometer mounted on a high, and terms for each item on this chart are dis- exposed ridge or peak. Tlie wind behavior at cussed below. higher elevations is of primary concern in ava- 1. Temperature. Accurate measurement of free lanche hazard forecasting. air temperature, which requires that the thermom- Snow movement refers to the amount of snow eter be unaffected by radiation or local air varia- observed drifting each day. Drifting snow can tions, is extremely difficult. The best approxima- develop avalanche conditions on lee slopes just as tion is achieved with standard thermometers a storm can. An estimate of wind-transported mounted in a properly designed and installed snow is an essential part of hazard forecasting. meteorological shelter. A proper shelter is the A^Tiere the drifting is not directly observed, evi- most important item of the temperature equip- dence such as fresh deposits in lee areas can serve ment ; much more accurate and consistent records as the basis for estimating what may have hap- can be obtained witli a dime store thermometer pened behind clouds or during hours of darkness. correctly mounted than with the finest instrument Widely different conditions between ridges and the casually himg on tlie wall of a building. When- valley floor may be noticed. A separate record ever possible, a standard Weather Bureau meteor- should be kept for each area. ological shelter should be used, located, and 3. Precipitation. Regular, accurate, daily mounted according to Weather Bureau instruc- measurements of snow depths are the backbone of tions. Standard maximum and minimum ther- these records. In addition, more frequent meas- mometers, used according to instructions, are pre- urements of new snowfall—as often as once every ferred. Where a station cannot be attended hour in some cases—are an important part of regularly every day, a thermograph mounted in a storm records and hazard forecasting. 41 42 A minimum of four snow stakes are customarily to collect a snow sample from the 24-hour stake used for standard observations. One is a fixed platform. The can is held upside down and ver- stake used to measure total snow depth on the tically over the platform adjacent to the yardstick ground. It is usually a length of clear 2x2 inch and pushed down through the snow until it rests lumber, painted white, with zero at the ground against the platform. Both can and platform, level and each inch of height clearly marked with carefully held together, are turned over. The dark paint. This stake is bolted to an angle iron platform may then be removed and the core of driven deep enough into the ground to provide snow collected allowed to slide down into the 8- permanent support, and is long enough to record inch can. The can is carried inside a building and total snow depth in the area. This stake should the snow melted. This may be done by pouring be labeled with total inches from bottom to top, onto the snow a measured quantity of hot water, rather than in feet and divisions thereof. The which is later subtracted from the measured pre- othei-s are short, portable stakes made of a yard- cipitation ; or the can may be set over a low source stick firmly bolted in an upright position to the of heat until the snow is melted. Holding the can center of a 2 x 2 foot platform of i/4-inch plywood. in a sink so hot water runs over the outside is a The plywood bases must be painted white, both quick way to melt the snow. The water obtained to protect the wood and to reduce radiation melt- in the can is then poured into the narrow collect- ing of the snow. ing tube of the rain gage and the amount of water The snow observation site where the stakes are measured with the reguhir 10:1 rain gage dipstick. to be located must be as free as possible from wind pjxactly the same procedure is used to collect drifting and snow creep. Tlie best location is in and measure water content samples from the in- a level opening in heavy timber. "VAHiere such a terval stake for use in computing precipitation site cannot be found, a compromise must be made, intensity (see under item 6.2). Instead of collect- but the stakes should be kept as far as possible ing the 24-hour core from the 24-hour stake plat- from buildings and other large objects which in- form, a separate plywood sheet may be put out on fluence the deposition pattern of snow. The site the snow surface and reset each day like the 24- should not be located so far away from the ob- hour stake. Both new snow depth and new snow server's dwelling that daily observations are apt water content are entered each day in the appro- to be neglected because of inconvenience. priate columns on the snow studies chart. Snow observations should always be made in the The specific gravity of the new snow meas- early morning. Erroneous values may be obtained ured is obtained by dividing the water content, if snow depths are read in the late afternoon or in inches, by the snow depth, also expres.sed in evening, because the new snow depth will be re- inches. Tliis figure is also known as the water ra- duced by settlement, or possibly melted away if tio, which may be expressed as a decimal figure the the sun has come out during the day following a same as specific gravity, or as percent of water snowfall the previous night. (multiply specific gravity by 100). For instance, a The first of the three short stakes is known as 1.5-inch fall of new snow which contained 1.20 tlie storm stake. This is set out on the surface of inches of water would have a specific gravity of the old snow and is left undisturbed for the dura- 1.20/15 = 0.08, which is the same as a density tion of any one storm period. "Wlien the storm is of 0.08 gm./cc. or a water content of 8 percent. over it is dug up and reset on the surface to be Though it is actually specific gravity that is deter- ready for the next storm. (The total depth of mined when snow depth and water equivalent are snow deposited on the ground from a single storm measured, this figure is usually referred to as the lasting more than 24 hours cannot be measured density for the sake of convenience. The correct without the storm stake. Summing the readings symbol for density is the Greek letter Rho, p. of the 24-hour stake gives a different value because The form, or type of snow, in a fresh snowfall of settlement.) The second yardstick is used as is entered each day on the snow studies chart. A the 24-hour stake. At each morning's observation general description such as "powder," "pellet," or the depth of new snow which has fallen in the past "granular" is given. In addition, the exact type 24 hours is read from this stake ; it is tlien cleared or types of snow crystals should be recorded ac- of snow and reset on the surface. The third stake cording to the system of the International Snow serves as the interval stake. It is used to obtain Classification, Types of Solid Precipitation (fig. snowfall measurements over intervals shorter than 7, ch. 2). The number or numbers corresponding 24 hours or during a storm, being reset on the to the type of crystals observed are noted immedi- surface after each observation. The use of obser- ately following the written general description. vations with the interA'al stake is discussed in more Special care should be taken to note whether or detail under item 6.2. not the individual crystals show a coat of rime. Water content of the new snow which has fallen The presence of any rime should be noted on the during the past 24 hours is measured at each chart. Most crystals can easily be identified with morning's observation. The ovei-flow can of the the naked eye, but occasionally a low-power mag- standard 8-inch Weather Bureau rain gage is used nifying glass may be required. 43 The following explanations of crystal types as- ard in the morning. There may often be a wide sociated with the broad snow classifications will difference between the two entries for the same aid the observer. day, resulting from changes in snow and weather Powder snow. Fluffy, feltlike new snow con- conditions or forecasting errors. Every effort sisting mostly of stars and spatial dendrites, with should be made to keep the "1700 actual" record occasional admixture of some plates or needles. as accurate as possible. Gramdar snow. A collection of small crystals The time and location of each avalanche ob- made up of stars or plates which are partly coated served in the area are entered in the appropriate with rmie. It may also consist of irregular column. A\niere a large avalanche cycle has oc- crystalline aggregates. Snow that is almost all curred and there is not enough space on the chart needle crystals will likewise appear granular. for all the slides, enter a reference to the storm re- Pellet snow. Also known as graupel, or soft port where the avalanches are described in detail. hail, is made up of crystals (most often stars) I)escriptions of the avalanches are entered accord- which have become completely and heavily coated ing to the "Avalanche Abbreviations for Snow with rime so that they form individual round Study Cliart" ( fig. 38 ). When these abbreviations balls. are used, an avalanclie is assumed to originate in Snow ßahes. Aggi'egates of individual crystals dry snow unless the lettei'S for damp or wet condi- which stick together at warmer temperatures. tions are included in the description. Examine carefully to see what kinds of crystals Avalanche types and classification are discussed are actually present. in detail in chapter 3. Windblown snow. Either during or after a It is important to record all snow movement. snowfall, may often consist of fragments of such This includes minor activities such as sunballs on crystalline forms as stars or spatial dendrites. south exposures and small loose snow sluffs. The original crystal type is difficult to identify. These are indications of changes taking place in The settlement column under "Pi^ecipitation" the snow and are significant in evaluating future on the chart refers to old snow settlement. Old avalanche hazard. snow in this context means snow which as been 6.2 Storm Plot and Storm Report Records. on the ground for more than 24 liours. This The methods of observing and recording snow con- settlement is computed for each day according to ditions described above apply in many cases to the the following expression : storm plot and storm report forms (figs. 39 and K Total snow \ 40). These more detailecl records of snowfall will new snow in total depth "1 depth of pre- I -I- { at time of I be kept whenever avalanche liazard is a problem vious day / V past 24 hours )]-[ observation J and forecasting activity must be undertaken. Certain refinements of observations are required 4. iSfate of Weather. Report the daily weather conditions concisely and accurately. The term when reporting on storms. They are discussed as follows : "clear" should be reserved for days with cloudless 1. Neto Snow Density. Snow density determi- skies; a sunny day with higli scattered clouds should be reported as such. To give more accurate nations have already been described in connection information, report cloud cover as scattered, with the daily snow observations. Density values broken, or overcast. Scattered clouds cover less determined from the 24-hour records, however, are than half of the sky, broken clouds cover more an average for the total amount of snow that has than lialf but not all, while overcast means that no fallen in that period. In the course of 24 hours blue sky is visible. ^^Hien changes in weather snow type and density can change widely. Even occur during the course of a day, note them. if they do not, settlement will make the average Record times and duration of precipitation on the value for the total quantity of snow different from storm records (see item 6.2). Wlien Weather that of the snow most recently deposited on the Bureau forecasts are available, they should be surface. For tliis reason, accurate new-snow den- entered in the column provided. sities must be measured at the surface almost as 5. Snoio Surface. This record is important both soon as the snow falls. One way to get an undis- for evaluating avalanclie conditions and for col- turbed sample is to put a shallow container such lecting data on skiing conditions. Follow the as a 1-pound coffee can on the snow surface and terminology of this liandbook and be careful to allow it to fill full of snow. Tlie snow is then distinguish between the different conditions on carefully scraped smooth over the top of the can nortli and south exposures. This record sliould be and weighed. If the weight and volume of the can have been previously determined, the snow the source of accurate information to the public density can be obtained by dividing the volume of on the actual state of skiing conditions (fig. 37). the snow so collected into its weight. For ex- 6. Avalanche Conditions. The degrees of haz- ample, a can with a volume of 1,000 cc. and a ard should be entered in tlie appropriate column. weight of 100 gm. which, when filled with snow, The "1700 actual" column is the record of ava- Aveighed 170 gm., would show that the snow den- lanche conditions that actually occurred, to be sity was: entered at the end of the day. The "0800 esti- 170-100 ^_ mated" column is for entering the anticipated haz- ^^ööö-=0.07gm./cc. 44 Snow Conditions Reports

Where reports of snow conditions are made to the public, particularly for winter recreational purposes, only major classifications would be used.

The following items are of interest to the general public:

1. Total depth of snow. 2. Depth of snow fallen in previous 2k hours. 3. Type of new snow: Dry, damp, or wet. k. Condition of the surface; new snow, old snow, ski-packed, crushed, etc. 5- Condition of the undersurface as it affects skiing conditions. 6. Weather conditions and forecast. 7- Avalanche hazards (if any). Closures. 8. Condition of highways. 9' Classification of skiing conditions, which may vary on different parts of the area.

Skiing conditions are classified as follows:

I. Excellent

1. New powder 3. Skipack dry or damp 2. Old powder h. Com snow

II. Good

1. Damp new snow h. Unbreakable crusts 2. Damp old snow 5. Windpack dry or damp 3. Skipack wet

III. Fair

1. Variable crusts k. Wet new snow 2. Icy crust unbreakable 5. Wet old snow 3. Windpack wet

IV. Poor

1. Breakable crusts 2. Slush

FIGURE 37.—Snow conditions reports.

45 Avalanche Abbreviations for Snow Study Chart

HS —- hard slat N - natural

SS —- soft slab A - artificial

WS --- wet slab AA - artificial, artillery

L - — loose snow AE - artificial, explosives

l'JL --- wet loose snow AS - artificial, ski

D --- damp (prefixed to 0 - surface avalanche HS, SS, or L) G - ground avalanche B - — sunballs J - windblast

1 - ■ - sluffs T| temperature rising

2 - • - small T I temperature falling

3 - ■- medium

1+ - •- large

5 - -- major

Example : A large, dry, soft slab avalanche released by artillery fire would be entered on the Snow Study Chart as SS-AE-^l-.

FIGURE 38.—Avalanche abbreviations for snow study chart. If available, a standard 500-cc. snow sample tube snowfall is the sum of these increments. The may be used to collect a core of snow from near more frequently the interval stake is reset, the the surface and the density determined by weigh- more nearly will the sum of the increments ap- ing as above. The tube should be chilled and very proach the true Avalué of the amount of snow carefully inserted into the snow to get an accurate which actually fell, free of settlement. sample of light, fluffy new snow. The settlement in the new snow during the time 2. Gross and Net Snoivfall and New Snoio of its deposition is the difference between the gross Settlement. The amount of snow measured on and net snowfall. This difference divided by the the storm stake is the net snowfall for the storm. gross snowfall and multiplied by 100 gives the This is less than the amount of snow which ac- percent settlement of the new snow during deposi- tually fell because the new snow is continuously tion. Once snowfall has ended, further settle- settling. The amount which actually fell, with- ment of the new snow can be observed directly by out settlement, is called the gross snowfall, a quan- watching the drop of the surface level on the storm tity determined with the aid of the interval stake. stake. (An example of these computations is When only 2 or 3 inches of snow are allowed to given in figure 40, sheet 1.) build up on the interval stake each time before it is 3. Snowfall Intensity. Snowfall intensity reset, there is little chance for any settlement to (S.I.) is the rate at which snow is deposited, ex- take place. If the interval stake is reset frequently pressed in inches per hour. Divide the amount of during a snowfall, a number of small increments snow which has fallen in a given interval of time of accumulation will be obtained. The ffross by the length of that interval. To be correct, this 46 storm Plot No. Date 12/25/59 Ai-ea AItu

Time Snow Depth New Snow Old P. I. S. I. WinA 2k Den Snow 31- rotal ätorm Hr fi #2 Type sity Sett Sett T.B. Gatie Core Stake Instr. Speed rec. Temp. Notes

2^ Decerabe r

0800 17 6 6/c 6/ Qnp. 0.12 0.10 1.1 15 W 27°

1215 __ 12 5 5/' 1 Qnp. 0.10 0 0.12 1.2 15-20 W 2S°

__ a sha rp tei pel at ire drop aroui d 12G 0, wir d aji L stom inte nsity pickin ! up.

1520 28 16 9 V' ) 0 0 1-3 9 w 26 Decembi r It. It. & Density from 2l4--hour ü6oo 37 25 19 ^^ pdx, 0.0Í 0 0 0.6 variai le 8" core, clouds now breaking, snowfall light.

FIGURE 39.—Storm plot form.

iunouiit of snowfall should be the gross snowfall evei', are highly significant for avalanche forma- during the interval. Measurements sliould be tion. P.I. values for shorter intervals should be made over sliort intervals whenever possible so determined whenever possible. that variations in the snowfall intensity during a 5. Precipitation Intensity Factor. The Precipi- storm can be recorded. If this cannot be done, tation Intensity Factor is an index of possible then the average S.I. for each day can be obtained hazard due to direct action soft slab avalanches. by dividing new snow depth of the 24-hour stake The first criterion is that winds must be above the by 24. critical level of velocity. When this is so, the 4. Precipitation Intensity. Precipitation inten- P.I. Factor is the amount of solid precipitation, sity (P.I.) is the rate at which water (which may (expressed in inches of water) which fell during be in the fomi of snow) is deposited, and is ex- any given interval when the precipitation inten- pressed in inches per hour. Divide the water sity was continuously above 0.1 inch of water per equivalent of the snow that has fallen in a given hour. Two inches of precipitation at a rate of time interval by the length of that interv^al to get 0.0.') inch per hour obviously would not fit this precipitation intensity. The customary method of definition, and for such a situation the factor obtaining the water content of the snow that falls would be nonexistent and inapplicable. If 2 inches over short intervals is by collecting sam^jles from of i)recipitation fell continuously at 0.15 inch per the interval stake platform with the 8-inch rain hour while winds were high, this factor would gage can, as previously described. A recording exist as an index of hazard, and the "P.I. Factor'" precipitation gage may also be used. The slope of would be 2.00. the precipitation curve on the chart is the P.I. 6. Hno\i^ Settlement. Settlement of old snow Such a gage should always be equipped with a during all or parts of a storm period may be de- scoop for obtaining accurate sampling of snow- termined just as it is for 24-hour periods. fall ; core samples with the rain gage must also be 6.3 Snowpit Studies. Periodic study and re- collected to check the catch in the gage. cording of suljsurface snow conditions to show The average value of P.I. for a 24-hour period snow cover evolution through the winter has be- can be obtained from the regular morning snow come standard procedure at Forest Service ava- observations, by dividing the total amount of lanche stations. This section gives instructions for water collected from the 24-hour stake platform collecting, plotting, and interpreting snow pit by 24. Variations in P.I. during a storm, how- data. 47 Storm Report Instructions

Storm No.: Number the storms of each season consecutively. 1, Record the hour of beginning and ending of snowfall as closely as possible. Record the duration, or durations, of the intense storm period in hours. 4. The snowfall according to measurements taken at the shortest intervals in use at your station. Record this interval: 2 hrs., 4 hrs., 24 hrs., or whatever. 5. The amount of snowfall at the end of the storm. Should be taken from a platform stake set on the old snow surface and not reset during the storm. 7. Average S.I. for the major storm period. Maximum S.I.: Record peri- ods of S.I. at l"/hr. or more. 8. Amount of water per inch of snow. Computed from the snowfall of each 24 hr. period. 9. Average P.I.: for the major storm period. Maximum P. I.: Record all periods when P.I. was . 10"/hr. or more. The gage record must be rea- sonably close to the precipitation according to the snow core method. If a recording gage is not in use, P.I. can be computed directly from the snow core, but the snow should be sampled not less often than every 4 hrs. 10. Record any period when P.I. was continuously at or near . 10"/hr. with strong wind action. Example; 6 hrs. at . 18: P.I. Factor 1.08. 11. Average wind speed for major storm period. Maximum: Record periods when wind speed averaged 15 m.p.h. or more. 12. Sample computation assuming that you have a total snow depth stake, a storm stake, and an interval stake: Old snow depth------12" (total stake) Gross snowfall ______12" (from interval stake summation) Net snowfall- - ______10" (storm stake) Total depth at end of storm - 21" (total stake) New snow settlement was 2 inches and old snow settlement was 1 inch. 13. Percent of settlement is computed by dividing gross snowfall into the amount of settlement. 14. Record general trend and any marked changes. 15. Record time, type, size, and location. Give best estimate of the trigger. 16. As above; also record method of release: Ski, explosives, artillery fire, etc. Record stabilization in place if it occurs. It is important for the observer to record his personal opinions, also any unusual or unfamiliar developments. 17. Sufficient copies of the Storm Report should be made so that one is avail- able to forward to the avalanche study center at Alta, Utah. Reports should be made promptly following the storm.

FiGUKE 40.—.Storni rei>ort (sheet 1). 48 STORM REPORT Havens and Storm No. 1 Date 12/2^/59 Area Alta Observer LaChapelle

TECHNICAL DATA

1. Began 0100 2$ December 19$9 Ended 0600 26 December 19$9 Duration Ivlajor Storm Period 29 hrs. 2. Old Snow Depth 8"

3. Old Snow Surface depth hoar on N. slo-pes. bare ground on S. slopes k. Gross Snow Fall 2¿]

5. Net Snowfall g^

6. New Snow Type 13" damp new snow foUnwPd by 12" dry powder 7. Snowfall Intensity Average 0.9 in/hr. Maximum 1.3 in/hr. ~

8. Water Ratio 0.09 Total Precipitation 2-1¿^" 9. Precipitation Intensity Average 0.07$ in/hr. Maximum 0.1$ in/hr.

10. P.I. Factor never reached 1.0

11. Wind Speed Average 10 mph Max. 18-20 moh (short i^erindRl Direction W & SW

12. Old Snow Settlement negligible

13. New Snow Settlement Inches

14. Temperature began 30°, falling to 10° at end of storm 15. Avalanche Occurrence - Natxiral General cycle of small, medium and large soft slabs of northerly exposures.

16. Avalanche Occurrence - Artificial Extensive ski and explosive releases--see gun report and analysis.

Observer's Analysis Trouble had been expected on north exposures with the first storm to deposit much snow over the existing shallow depth hoar. The danger was further heightened by the existence of a crust layer in the depth hoar resulting from a November salt storm. The early part of this storm brought damp, heavy snow which favored slab formation, and avalanche activity began early, when little more than 6 to fi" of snow had fallen

FIGURE 40.—Storm reiwrt (sheet 2). 49 Test skiing on the morning of the 25th revealed highly unstable conditions, and the skiers were restricted to minimum hazard areas on the Collins lift for the rest of the day. A test release in one of the chutes of the Peruvian Ridge resulted in release of a soft slab extending all the way from Peruvian Bowl to Westward Ho. Very extensive cracking of the snow was observed even early in this storm, and persisted for a day or two afterwards.

Artillery fire on the morning of the 26th brought down several large slides, the most extensive being those on Baldy Shoulder and Simspot-Race-Course. Many small patches of slab existed throughout the ski area, and test skiing and blasting in the Peruvian Ridge-V/ildcat areas proved these to be sufficiently dangerous that they were kept closed throughout the day. A small part of the ridge around the top of the Wildcat Lift was left open to permit use of this lift, but even here the snow was un- stable, and two small slabs were dislodged by skiers, one of them burying another skier who had the misfort\ine to fall just as the slab patch was released. The latter was quickly dug out and was uninjiored. A heavy influx of skiers out to take advantage of the first good skiing of the winter made the area control problem difficult, and there was considerable complaining about the restrictions until it became obvious to even the most careless skiers that a serious hazard situation existed.

It is interesting to note that the slopes of Stonecrusher and lone Pine, which had been previously foot-packed, did not slide, though adjacent slopes slide vigorously following artillery release. Increasing evidence confirms that foot-packing is an effective countermeasure against the hazards of depth hoar. Peruvian Bowl also did not slide, for it had been thoroughly side-stepped with skis several days prior to the storm.

Blasting and artillery fire continued for two days after the storm, and danger in the outlying ski runs was eventually reduced. A serious condition of instability continues on north slopes, however, and more trouble may be expected from future storms. At the date of this writing (3 January), occasional patches of unstable slab have continued to be uncovered in the ski area, and a general hazard condition is presiimed to exist in the back country where no control measures have been taken.

FIGURE 40.—Storm report (sheet 3).

50 1. Preliminary Requirements. The prerequisites snow investigator with clues to the various condi- of a site for snow profile investigations are few, tions existing in a given snow cover. but important. For practical use, the study plot The penetrometer consists of an aluminum alloy must be accessible to the observer throughout the tube approximately '?> cm. in diameter, 1 m. long, winter. It must be an area free from avalanche and weighing 1 kg. The 4-cm. diameter point danger, disturbance by unnatural deposition of is a 60° cone tipped with steel for wearing quality. snow (such as the discharge from rotary snow- A scale in centimetei-s is engraved on the tube, plows), and foot or ski traffic. If disturbance by reading upward from the point. Depths over 1 m. passing skiers is likely, the area must be roped may be probed by the addition of successive 1 m. off. The study plot should be reasonably level lock-connected lengths of tube. and free from exposure to snow creep pressure A sketch of the ram penetrometer, together with from adjacent slopes. It should be located away tlie fornnila used to compute ram resistance, ap- from ridges, cliffs, or other terrain irregularities pears in figure 41. that would tend to modify the normal conditions The force of penetration is developed by the fall of snow deposition and metamorphism, and should of a weight along a guide rod supported on the be protected from the scouring action of high upper end of the penetrometer tube. The force winds. The ideal location is a level opening in a may be varied by varying the height of fall and stand of heavy timber. It should be such size that size of weight. The calculated value of ram re- the distance from the edge of the study plot to the sistance, given in kilograms, is an arbitrary figure nearest trees is at least as great as the height of depending on tlie dimensions of the penetrometer. the trees. Normally, the study plot should be Headings taken under different conditions and by located close to the total snow stake where accu- different observers can usually be compared. Op- mulation and snow depth records are obtained. eration of the ram penetrometer is as follows : If tliis is not feasible, a separate total depth stake a. Hold the penetrometer tube vertically with must be provided at the study plot. A number of point at the snow surface a)ul allow it to sink pits will be required in the course of a winter; by its own weight into the snow. Read and re- therefore the test area must be staked out and the cord penetration. sites of future pits planned in advance. Each pit b. The 1 kg. ram weight and guide rod are then is refilled with snow after the studies have been carefully set on the tube and any further pene- made, to keep the snow cover in the test area as tration due to the increased weight is recorded. homogeneous as possible. The limits of the area c. Raise the ram weight and let it fall from a which has been dug up for study must be clearly height sufficient to produce at least 1 cm. pene- marked, so that fi-esh undisturbed snow can be tration. Note scale reading at the snow surface excavated from each succeeding pit. Excavated and the fall height of the weight. As the proc- snow must be thrown away from any area which ess is continued and firmer snow layers are may subsequently be sîtudied. reached, tlie distance of fall for the weight will In sliallow snow covers, a square pit dug through have to l)e increased. "When maximum possible to the ground is usually the quickest and most con- fall height is insufficient to produce at least 1 venient. For deep snow, a long narrow pit in- cm. penetration, then two or more falls must be creasing in depth by steps from one end to the used between successive readings. A greater other provides easier access ; or a vertical shaft can sensitivity to small variations in ram resistance be dug with the snow hoisted up from the lower and to thin snow layers is usually obtained if levels by bucket. A D-handle shovel with a square short fall heights and a number of falls between blade is the best for all-around use. A short penetration readings are used as regular prac- shovel with a pointed blade may be needed to tice. If heavier ram weights are available, they break up very hard snow. may be substituted for use in very hard snow. In practice the ram profile is taken first, and the d. When the first tube has penetrated to ap- penetrometer left in place to guide the pit excava- proximately 80 cm. a second tube is added, as is tion and to serve as a scale of reference. The sub- a third tube when the depth has reached about sequent tests are then made in more or less the 180 cm., and so forth. Addition of the extra order given below. tubes must be noted in the record. Changing 2. The Ram Penetrometer. The ram penetrom- the total weight of the penetrometer changes the eter (Haefeli cone penetrometer), or "ramm- calculated value of ram resistance for given sonde,"' was developed in Switzerland over 20 conditions. When the point has penetrated years ago. It has Ijecome a very useful tool in through the snow cover and reached the ground, various kinds of snow investigations. This instru- a sudden increase in ram resistance is usually ment measures the resistance snow offers to pene- tration of a cone-shaped point, and indicates the observed, and the ram weight makes a duller relative degrees of hardness and cohesion in vari- sound when it falls. ous snow layers. The method of use provides a e. From the quantities recorded (fall height, graph of ram resistance against depth without ram weight, penetrations, etc.) the ram resist- digging a pit. The stratification of the snow is ance for each snow layer can be calculated with readily apparent from the graph and provides the the formula given in figure 41.

51 JZÂM IV£^/Gß/r

outoE e.oD —». f^ÂlL //£/G//r

SMO^ SUK^ACe Q - HVe/GHT Of" TUBE (s) IN t«j. /Z.= We/OHT or liA^ C-ENT;M£T£íC SCAL£ n - No/^isEe. OF »LOWS P£/iFritATIOM TUßE h = FALL ÑefGMT OF eAM

A = P£N£Te.àTtOM FOK. h BLOWS

1 I I PEA/areATJOH I A

FIGURE 41.—The ram penetrometer.

o. Tem-perature. The simplest way to deter- face are being made. If glass thermometers are mine temperature profile is to insert thermometers used, a metal punch should be kept handy to make at regularly spaced intervals in the wall of a holes for the thermometers in all but the softest freshly dug pit. Eeliable temperature records snow. Tlie Weston dial thermometer is more cannot be obtained in this manner after more than rugged, and is i-ecommended where great precision an hour or so of exposure of the pit wall. A us- is not required. A punch should also be used for able curve of temperature against depth can be it in very hard snow. Care must be taken to avoid obtained if a few precautions are taken. The best parallax when reading the dial thermometer. method is to measure the temperature from the After the thermometer is inserted in the snow, it pit wall as the pit is being cíug. This insures sliould be allowed to remain there for a few min- minimum disturbance of the natural temperature utes to approach equilibrium. Then it should be gradient. Where more accurate records are re- shifted to a fresh spot of snow at the same level quired, resistance thermometers, thermistors, or and once more allowed to approach equilibrium thermocouples must be allowed to accumulate with before being read. This is necessary because the the snow cover during tlie winter so that they are tliermometer initially inserted in the snow is usu- surrounded by undisturbed snow. ally at a different temperature than the snow, and The thermometers may be inserted in the pit the first equilibrium temperature reached will be wall at 20-cm. intervals, unless more detailed appreciably different than the surrounding snow studies of snow temperature changes near the sur- temperature. 52 An accurate temperature of the snow surface ting a time profile of snow cover evolution (see I is very difficult to obtain. An approximation may item 6.4). The threads become incorporated in be obtained by sliding the thermometer just under the snow cover and plainly mark layer boundaries the snow surface, and shielding it from direct solar wliich might otherwise be difficult to distinguish. radiation. 5. Dem'ify. Snow density is determined by the 4. St7'atigraphy. The stratigraphie, or layer, familiar method of weighing a sample of known profile ig derived from detailed observations of vohune. The snow specimen is collected by insert- grain size, grain type, layer boundaries, hardness, ing a standard 5()0-cc. sampling tube horizontally and other special features in the snow cover being in the i)it wall. Tlie snow is cut away around the investigated. The results as plotted show the tube, and the snow at the ends of the tube is care- various layei-s of snow and their characteristics. fully cut or scraped flush with the tube, and it is See the pit profile plot in figure 42. weighed. The tube weight is deducted from the The snow crystals observed are classified ac- total weight to get the snow weight, and the den- cording to the outline given in the Types of De- sity is computed from this known weight and posited Snow chart, and the appropriate graphic volume. If a large number of sam^ile tubes are symbols are used for plotting the results. available, the specimens from an entire profile may A few general remarks atjout visual observa- be collected at once and weighed at a convenient tions will aid the field worker. One of the quick- indoor location. If the tubes will be exposed to est ways to locate the various layei-s, where large melting temperatures before being weighed, they differences in hardness are present, is to draw the must be capped with iiibber caps to prevent loss finger or a pencil tip down the pit wall at an even of any meltwater formed. rate from top to bottom. The variations in resist- Density samples should be collected from each ance offered by the difièrent layers will be quite separate snow layer in the profile large enough to apparent. Wliere the dift'erence is slight, careful admit the sample tubes. Several samples should horizontal stroking of the pit wall with a soft be obtained from the thicker layers. The height brush will model out the layers. AVhen applied of each sample collected in the snow cover must to new snow near tlie surface, this teclmique will be recorded. The density values are plotted even reveal the variations in wind velocity which against snow depth to give the snow density curve occurred during deposition. Care should be taken, for the profile, as illustrated in figure 42. however, to make a close visual inspection for the B. Plotfing the ProfJex. The general technique layer boundaries, because a change in grain type in plotting a complete snow profile from pit data is or size will sometimes occur with very little change suggested in the accompanying example (fig. 42). in hardness. The direction of the axis on which any given char- In the absence of a hardness gage or penetrom- acteristic is plotted is merely a matter of conven- eter, a rougli estimation of hardness may be ob- tion: the example shows the method of plotting tained by notjng the size of common articles which foiuul convenient after considerable use. Milli- may be easily pushed into the snow. Tlie follow- meter graph paper used with a scale of 1: 10 (one ing table will serve as an example : millimeter for each centimeter of snow depth) Approx, ram Hardness resistance gives a convenient and easily manageable profile. Mittened fist (sidewise mo- very soft Ü-3 kg. 6.4 Time Profile. The end product of snow pro- tion possible). file determinations is the time profile, from which Flat hand pushed endwise._ soft 3-1.5 kg. the investigator may observe the overall character Single outstretched finger. _ med.-hard 15-50 kg. Sharpened pencil hard _. 50-150 kg. of snow coAer evolution throughout the winter. Knife blade very hard over 150 kg. This time profile is prepared by plotting (on a time base) a series of pit profiles obtained from a Special features such as sun or rain crusts, thin sequence of pits dug at the same study plot layers of surface hoar crystals, the presence of throughout the winter. Daily snow depths, settle- depth hoar, ice laminations, and ice lenses, should ment data, and meteorological records are also Ije noted. incorporated in the time profile. If care is taken Individual snow layers may be marked by to identify individual layers in each pit profile, stretcliing colored thread on the surface after each tlie behavior of these layers can be followed snowfall when profiles are to be taken throughout throughout the winter. The desirable interval for the winter at a single site for the purpose of plot- collecting pit data is every 2 weeks.

53 TBMP£R. A Tt/fi. £. 7? -^ -I 1 u- -iz -JO -a -6 -4 -z

fiÀM R.ESI STANCE. 0£NS/rY KG •* h- H ^ cc or oz 0.3 osr S^/i JLOO

New jHow D£jvs/ry

/SO - rtME G It Af ME O OLP SA/ûiy

OLD CgA(/P£L. too -

COAKse axA/A/eo OLD SNO^

SO - COAIST/i(/CTIY£

META M ofi,PNisM

D£PTN HOAK.

/^/^^/r//s:f/ ^3^ 3^?^^ TTT, dtOt/f^D

FIGURE 42.—Typical plot of pit profile data, showing symbols for various deix>sited snow types.

54 2332.81 SNOW AVALANCHES Chapter 7

7 SNOW STABILIZATION. Snow may be High explosives represent another elïective stabilized and avalanche hazard consequently re- means of inti'oducing mechanical disturbance in duced or eliminated by the following methods. the snow. P^xplosives are safer when large ava- These represent measures of avalanche control and lanche paths are involved, because they may be prevention as applied to the snow itself, in con- delivered from a distance either by hand-throwing trast to such defense measures as retention bar- or by an artillery shell. The shock of an explo- riers which permanently alter the nature of the sion is extremely effective in relieving creep ten- terrain. sion, even in the hardest slab. Intense age-hard- Mechanical disturbance of the snow with one ening is also initiated in the immediate vicinity exception results in a trend toward greater sta- of the blast crater, but this effect is conñned to bility. This works in two ways. One is the proc- the snow whicli is physically displaced by the ess of age-hardening already mentioned in chap- explosion. î^or this reason it is not efficient to ter 2, item 2:2. "VMienever snow is disturbed and depend on explosives to create widespread age- then allowed to set, its ñnal hardness is greater liardening over a slope. Rather, they are used than if left undisturbed. This eft'ect is accentuated primarily to initiate avalanche release and/or to at lower temperatures. The other method of relieve creep tension. The use of explosives is stabilization is by relief of creep tension or inhibi- discussed in detail in item 7.2. tion of its development. Existing creep tension An avalanche release also stabilizes the snow. is relieved if the homogeneous layer of a snow An avalanche in motion represents a major me- slab is broken. Once the slab is broken into seg- chanical disturbance of both the sliding snow and ments, there is much less likelihood that such ten- tliat nearby which falls under its influence. The sion will again develop. Slabs which have been consequent age-hardening tlius eliminates large broken without an avalanche release tend to be- quantities of snow from the probability of being come a well-attached part of the snow cover. responsible for a later release. Instead of stabiliz- The one exception to the stabilizing effect on ing snow in place on an avalanche path, a slide snow of any mechanical disturbance is wind, for removes a large quantity of unstable snow and it transports the snow from place to place as well carries it to more gentle slopes below where it can as disturbs it. It is a potent natural stabilizer cause no harm. The immediate hazard on a given when it forms windpack and windcrust. When it path is eliminated, and the future hazard is re- constructs a slab or transports an overload of duced, for there is less snow available for subse- snow from one part of the area to another, the quent slides. Avalanching is frequently a less de- reverse is true. sirable protective measure than stabilization in Skiing is the most obvious means of mechanical place. In a ski area good skiing snow is removed disturbance available in a ski area. It is a well- from attractive slopes, and on a highway a slide established general observation that slopes which on the roadbed means extra labor and cost of receive heavy ski use are much less likely to ava- clearing. Nevertheless, avalanches are frequently lanche than those which do not. Light skiing— precipitated to insure public safety, and the bene- a few tracks criss-crossing a hill—also has a sta- ñcial effects should be recognized. bilizing influence. First, it may relieve creep ten- 7.1 Test and Protective Skiing. Protective ski- sion in a soft slab. Second, each ski track and mg is the deliberate, day-to-day disturbance of each area of snow displaced in a turn, represents snow on avalanche slopes in order to encourao-e an island of harder snow which will impart stabilization as described above. Test skiino- is an strength to the upper snow layers. In short, a attempt to artificially release avalanches on little skiing is better than none, and a lot of skiing selected small slopes by skiing. It serves as a field is even better. Breaking up the snow will im- test of snow stability to check the conclusions of prove its stability; compacting it will assure a hazard forecasting. The two functions of test and solid, well-anchored layer. protective skiing are so often combined that they When snow reaches a certain degree of hardness, are treated here together. The safety precautions caused by high winds and low temperatures, the for both are identical. ski is no longer able to penetrate the surface. In The reaction of snow under the snow ranger's such cases more violent means of breaking the skis constantly provides him with information on snow are required, such as foot-packing or the use the physical processes which are taking place in of vehicles. the snow cover. This technique provides warn- 55 inff of snow which is tending to form slabs and slope and open it to use. Closures are a temporary create an inipendinc; hazard, particuhxrly during measure intended to protect the public during pe- storms. An important key to successful test skiing riods of high avalanche danger, especially durnig is the selection of test slopes which accurately storms. They should be lifted as soon as possible represent the snow conditions found on the larger after the critical period passes, so that full benefit and more dangerous avalanche paths. In most of stabilization by use may be achieved. It should areas it is possible to locate some of these small, be the objective of a snow safety program in any short avalanche patlis where slides may safely be ski area to keep all the slopes open just as much as dislodged by skiing. It is very desirable to find possible without exposing the public to hazard. such paths which have the same orientation and Safety precautions to be observed by men engaged approximately the same slope angle as the larger in test and protective skiing are outlined in chap- paths. Configuration, relation to timber, and al- ter 8, item 8.3. titude must also be considered. Safe routes of Depth hoar requires special attention for proper approach, access, and exits are essential. stabilization. When a shallow, early-winter snow Cracks which form in the snow with the cover has been converted to depth hoar, the base passage of skis are the best indicators of develop- for subsequent snowfalls is dangerously unstable. ing slab avalanche danger. Breaks in the snow I^nlike fi-eshly fallen snow, depth hoar is not read- immediately adjacent to the skies are not neces- ily stabilized by skiing. It is not easily compacted sarily a danger sign ; the weight of the skier may into a hard layer, and tends to reform if steep simply be cracking the snow locally. Breaks or temperature gradients persist. Repeated side- cracks wiiich run ahead of the skies are more stepping with skis tends to compact it sufficiently clearly a warning of hazard. A crack which may to harden the layer, but breaking down the struc- appear for just a few feet ahead suggests that slab ture by skiing alone is not sufficient. The same conditions are forming and caution is indicated. slope nnist be tramped over several times by a Cracks which run far ahead of the «ki (e.g., 50 number of skiers to insure adequate stabilization. feet or more) indicate serious instability. Under Foot packing is a more reliable method. The these conditions avalanches on steep open slopes higher specific pressure of a boot appears to pro- are almost certain. Occasionally conditions will vide satisfactory compaction and hardening. Ex- be found in which cracks appear everywhere in perience has demonstrated that foot packing will the snow at the slightest disturbance—on gentle satisfactorily stabilize early season depth hoar, slopes and among trees—as well as on obvious while adjacent slopes not so treated persist as a avalanche paths. Snow in such a condition is very source of avalanche hazard. Increasect safety and "touchy"' and avalanches are often released un- ease of control warrant the work involved to foot expectedly on slopes where they normally do not pack slide paths which threaten a busy ski area. occur. The initial formation of depth hoar is inhibited If snow actually avalanches on a test slope fol- by thoroughly packing each fresh fall of snow as lowing the passage of a skier, the character of the it occurs. This will not prevent depth hoar from slide will furnish additional information on the developing, Ijut a reduction in the porosity and degree of instability. Considerable external force increase in density will restrain free circulation may be required to initiate an avalanche, if only of water vapor within the snow cover and reduce the snow dislodged by the skier slides away. If the adverse effects of steep temperature gradients. the fracture line propagates for some distance, and 7^ Explosives. The use of explosives to stabil- the slide gains mass and momentum as it descends, ize snow—either hand placed or as projectiles— the triggering of large slides ol)viously will l)e has been routine in the United States for a number much easier. "\^nien relating these observations to of years. Techniques and equipment have steadily larger avalanche paths, remember that differences improved. in location, slope, angle, etc., will cause ditferent Before the explosives techniques were devel- paths to behave in different fashions. The test oped, the only avalanche defense the snow ranger skiing procedure offers an additional check on liad was the closed area sign. The signs did noth- snow conditions, but it does not funiish all the ing to deter avalanches and very little to deter the answers. When done by an experienced observer, people. however, test skiing does often furnish accurate Explosives have numerous advantages as a evidence on the approach of dangerous conditions. meaiis of avalanche hazard control. They are a The experienced snow ranger will enlist the aid logical extension of the test and protective skiing of the skiing j^ublic in his program of snow stabi- techniques which seek a definite answer to the lization and avalanche prevention. Once a slope question "Is this slope stable?'' A heavier shock has been adequately tested (usually with explo- can be delivered to the snow, and it can be done sives) and proved safe, skiers should be encour- quickly and safely. Explosives are the only means aged to use it and get the snow broken up and of getting a definite answer where hai-d slab con- packed into a solid layer. Closure of ski runs is ditions prevail. not a satisfactory avalanche defense measure. Where projectiles are available, large areas can Neither is it an acceptable substitute for approved 1)6 controlled from a few firing positions. The control measures whicli can be used to safeguard a savings in time and labor are notable. At Squaw

56 Valley, during the 1960 Winter Olympic Games, 30 different and Avidely scattered avalanche paths were controlled as a matter of routine in less than '2 houi-s. Where a highway crosses the area, the rifle can be hauled along it and deliver projectiles to avalanche patlis otherwise inaccessible. Contrary to the belief of the general public, the objective of using explosives in avalanche liazard control is not to start avalanches. The objective is to determine if the snow is stable or unstable and eliminate hazard. The latter does not necessarily require that an avalanche fall. Education of the public is important because it leads to the cooperation and acceptance by the public, without wliich no snow safety program can be successful. Accordingly, avalanche con- trol opex-ations are often performed in full view of spectators, and all personnel engaged in ava- lanche control work are encouraged to explain the objectives. When an explosive charge is detonated in an avalanche area, one of three things happens: 1. An avalanche occui-s. This proves that the snow is unstable and further control action is needed on other slopes. 2. The snow cracks and either slides only a short distance or stabilizes in place. This indicates a moderate degree of instability. It calls for fur- ther control action, since some other slope may be ready to slide. Stabilization in or nearly in place F-462503 is a better outcome than an avalanche, for there FIGURE 43.—Removal of cornice by a heavy charge of is no debris to roughen a ski slope or fill a highway. tetrytol. ?). Nothing happens except the formation of a crater in the snow. This suggests stability. If For a soft slab, a 2y2-pound block of teti-ytol confirmed by several more shots on key avalanche (military demolition) is generally a sufficient paths, it may be taken as evidence that the area charge. If the slope is wide it is better to fire as a whole is stal)le. This is the best answer of all. a nudtiple charge strung out along the fracture Any one of tlie three answers is satisfactory, line than to increase the size of a single shot. If for it is definite. If there was a hazard, it has the slide path is unusually large it may be neces- been removed; if there was no hazard, this has sary to blast at several different trigger points. been demonstrated. For the hard slab-depth hoar combination, the size Continuity of avalanche liazard control opera- of the shot may be increased to 5 or even 10 pounds. tions is also important. Tlie routine and frequent The ignition type detonator (blasting cap) saves stabilization of avalanche paths by all methods— time and is approved for avalanche blasting. (For from protective skiing to explosives to skiing winter work such as this, the fuse is most easily lit use—is the best guarantee tliat no deepseated, with a pullwire igniter.) With this type of climax-type hazard can develop. In the educa- detonator slide paths under a lift line have been tional aspect, people become accustomed to the successfully blasted by dropping the charge from operation, miderstand what it is all about, and a moving cliair. even identify tliemselves witli the program if they For cornice removal, the standard procedure is are allowed to help out in small ways. to drill shotholes along the expected sheer plane 7.21 Hand-Placed Charges. For slope ava- at 6- to 8-foot intervals and to a depth that will lanches, exi)losives with a liigh rate of detonation ])lace the explosives in tiie center of the snowmass. such as military demolitions and seismograph The holes are loaded with 3 or 4 sticks of dynamite powder are the most effective. For coniice re- and well tamped with snow. The charges are moval, the slower explosives such as 40 percent gelatin dynamite are moi'e effective. The faster linked together with primacord and fired as a explosives deliver a maximum blow to the surface multiple blast (figs. 44 and 45). The shotholes and also propagate a shock wave which counters can be dug with a snow sampler. A handier tool the marked absorptive powers of snow. Cornice is a sectional avalanche probe with a casehardened removal, on the other hand, is like a quarrying cone-shaped head, the same diameter as a dyna- opei-ation which i-equires a strong push rather than mite stick, which can be screwed on to the bottom a sharp blow (fig. 43). section of the probe. 57 FIGURE 44.—Dynamite releaise of massive cornice at Twin Laiies Pass, Alta, Utah. This cornice is about 75 feet long and 2G feet high at the face. Ten shot-holes were used, spaced 8 feet apart along a straight line and drilled 6 feet deep. Each hole was loaded with two sticks of 40 percent gelatin.

A cornice wliich forms every year or several 7.22 Artillery. Fundamentally, artillery repre- times durino; the winter may be préchargée!. Ex- sents an extension of the snow ranger's throwing plosives are placed on tlie ground as near as pos- arm. A projectile launcher must be accurate, sible to the shear plane, or ¡placed after the cornice mobile, reliable, simple to operate and maintain, is partially formed. Tlie charge can then be fired and have adequate range. whenever the overliang becomes dangerous and Of tlie numerous weapons listed, the 75 and the j)rocedure rei)eated as often as necessary. The 105 mm. recoilless rifles and the 75 mm. pack precliarging technique has advantages: it effects howitzer have proved to be best adapted for ava- more complete removal and permits placing the lanche control operations. Weapons larger than charge in favorable weather. Moistureproof and 105 mm. ai'e not I'ecommended. frostproof explosives must be used. Military The 75 mm. recoillesx rife is especially suitable. demolitions, seismograph powder, and bangalore It weighs l(í7 pounds, is readily mobile, and has torpedoes are suitable. Such cliarges must be very a range of 7,000 yards with pinpoint accuracy up securely anchored against snow creep, and should to 3,000 yards. This rifle is easily operated and be armed with 2)rimacord—never with blasting maintained and can be fired from any platform caps. Where tlie cornice is inaccessible, projectiles whicli will support its weight (fig. 46). Its only have been used effectively. The technique is to drawback is the "backblast" common to all recoil- get "on target" by tlie laclder method with point- less rifles. Gun crews should use air-valve type detonation fuse; tlien to fire delay-fuse rounds ear plugs because of its sharp concussion. into the cornice to get the burst inside the snow- The 75 mm. pack honntzer is similar to the re- mass. coilless rifle of the same calibre in range, reliabil- 58 FiGUKE 4').- -Sefiuel to figure 44, after coruice has fallen. Arrows point to location of shot-holes, Note the hard, smooth surface of an older cornice underneath the one removed. ity, iuid shock power. However, it is consideriibly Firing ponifion. The selection of firing posi- more complicated. It weighs approximately 500 tions requires careful study. They must be acces- pounds and is thei'efore much less mol)ile. Unlike sible under all conditions of weather and hazard the recoilless which hres fixed ammunition, the and protected from avalanclies. The safety of pack howitzer lias a variable powder cliarge. Un- habitations, parking areas, and other facilities less fired with Charge No. 4 (full propellant must be taken into accoiuit. With all these limita- charge) there may he a i^roblem with duds or deep tions in mind, the snow ranger chooses firing posi- bursts in soft snow. Since there is no backblast, tions from which the maximum number of taro-ets it can be fired from restricted areas such as the can be reached. They may be strung out along a doorway of a shelter. It is well suited for use in highway or concentrated in the vicinity of the a ski area where it can be fired from a fixed posi- ranger station, or located on a ridge if a lift pro- tion, or along a highway where it can be towed. vides access. With proper care, rifles and other The 105 mm. recoiUe^K rife (fig. 47) is similar weapons can be left in the open all winter. to the 75 mm. in accuracy and reliability. It has greater range and shock power. It weighs 700 In a ski area, gun platforms raised above the pounds and is not mobile over snow, but it can be maxinnmi snow level are essential. They also mounted in and fired from a wheeled vehicle. make it possible to fire on short, notice without Concussion and backblast are heavier than with delay. They provide the fixed position for blind the 75 mm. recoilless but not so sharp. liar plugs firing, "\\niere the rifles are to move along a hio-h- are required. The 105 mm. recoilless is recom- way, fixed positions can be provided by markîng mended particularly for use on liighways where the wheel positions of a mounted vehicle or a ranges may be extreme. towed weapon.

59 F-495210 FiGuiíE 46.—Typical fixed position for a 75 mm. recoilless rifle. Here the rifle is on a pedestal mount, which is interchangeable with the normal tripod mount. An aiming stake for blind firing is visible in front of the platform. The elevated position gives ample clearance for the backblast with all snow depths. F-495211 Targeting. The estimated range should l^e de- FIGURE 47.—10.5 mm. recoilless rifle permanently mounted termined from a topograpliical map when firing on a tower for avalanche control in a major ski area. A winter's supply of ammunition is stored in the base at target for the first time. Tlie first shot sliould of the tower. With such an installation, accurate be visually aimed so that it will land well down blind firing data can be acquired so that control meas- on the slope ; then proceed upward, ladderwise, un- ures can be taken during storms. Alta, Utah. Note til the burst is on target. At all times, be sure that lightning grounds as safety measure. there is enough dead space behind the target so tliat an "over" will not endanger life or property. pose, it is extremely etfective in shortening closure A range caid should be prepared for each gun time, preventing the development of major ava- position showing both the range in yards to each lanche conditions, and reducing the volume of target as determined by visual firing, and the ele- snow and amount of avalanche debris in a ski area vation of the barrel as measured by the gunner's or piled upon a highway. Blind firing can be quadrant. done safely during blizzards when it would be un- Firing projectiles into or over areas occupied by safe to use hand-placed explosives. With proper people is not permitted. For this reason most fir- blind firing data, the gunner can hit any target he ing in ski areas is done in the early morning. On can see during clear weather. highways a blockade system is essential. At each fixed gun position, the snow ranger sets Projectiles should not be used where the target up aiming stakes which determine the line gun- is close to structures, because of fragmentation. target. Elevation is set by the gunner's quadrant Shrapnel dispersion from tlie lOf) mm. shell is par- as taken from the range card. ticularly dangerous. Training. Personnel assigned to use artillery BVtnd ßring often starts a premature avalanche or hand-placed explosives must receive special cycle or stabilizes the snow in place. Whether in training. (See Forest Service Safety Regula- a ski area, along a highway, or for any other pur- tions, chapter 8, item 8.3.)

60 2332.81 SNOW AVALANCHES Chapter 8

8 SAFETY. By now it should be obvious to the of caution to men engaged in control work and reader of this handbook that avahvnches are a to anyone traveling in avalanche terrain. Almost dangerous and sometimes unpredictable force of all accidents involve slides which are started hy nature. It cannot be emphasized too strongly that the victims themselves. Accidents which result the artiñcial release of this natural force involves from naturally released slides falling on an unsus- considerable danger unless carried out in strict pecting victim are rare. The time it takes for even conformance with established safety procedures. a large slide to fall is measured usually in seconds. 8.1 Safety Objectives. The three principal ob- The chance that this time will coincide with the jectives of avalanche control safety regulations passage of a traveling party across an avalanche are to : slope is very small, unless an artificial trigger is 1. Carry out avalanche control without endan- supplied. The moral of this lesson is very plain: gering forest officers or others participating in the don't cross avalanche slopes and provide the trig- control operation. ger for a slide whicli may overtake you. If neces- 2. Assure that control operations do not endan- sary, go below the slide path—this is safer than ger skiei-s or others who may be in the vicinity. being on it. Better yet, go alrove the slide, even 3. Prevent damage to lifts, buildings, automo- if this means a long detoui". biles, powerlines, and other facilities. o. It should become a matter of habit among To attain these objectives, the Avalanche Con- men engaged in avalanche control work to avoid trol Safety Regidations in this chapter have been using wrist loops on ski poles. Once a man gets promulgated. These have the official status of caught in a slide, his chances of survival, or escap- a supplement to the Forest Service Safety Code, ing injury, will be greatly enhanced if his hands and should be so respected. are free to protect his face, excavate breathing 82 Safety Principles. The following general space, or dig himself out. He may be in trouble safety principles are based on accumulated if his arms are pinned by buried ski poles. experience. 4. Small avalanclies, which ordinarily would be 1. Avalanches nnist never be artificially released harmless to a man caught in them, can be danger- unless it is absolutely certain there are no humans ous if they fall over cliffs or into gullies or ravines. on tlie slopes below. If these slopes are not visible Burial and suffocation are usually thought to be to the man effecting the release, a system of guards the common fate of those caught in avalanches, and signals must be arranged. Normally, release but practical experience has shown that mechan- should not be attempted during poor visibility ical injury from striking trees, stumps, or rocks (fog, snowstorm, etc.) except when blind firing are as responsible for death or injury as burial. with artillery and then only if rigorous safety Early winter slides running on a depth hoar layer precautions are taken. next to the ground are particulai'ly dangerous in Merely walking or skiing along a ridge or near this respect. Their paths are often studded with an avalanche path may effect release during con- I'ocks exposed by the sliding snow. ditions of exceptional instability. With such 5. Safety rules for handling explosives are conditions, artillery fire or blasting may cause treated in detail in the Forest Service Safety Code the sympathetic release of large slides at great and handl)ooks prepared by explosive manufac- distances from the actual target point. Fracture turers. Strict observance of these blasting regula- lines or the collapse of an unstable snow structure tions must be enforced in avalanche control work may be propagated for a very long distance. In the same as in any other blasting operation. Mod- one recorded case an artillery shellburst started em explosives and blasting equipment are manu- sympathetic release of one adjacent slide path factured to a high standard, and misfires due to after another until a large avalanche fell on the defective manufacture are very rare. Most mis- opposite side of the mountain from the target fires can be traced to improper handling, storage, point. When such conditions are suspected, great or preparation of the charge. Since there is more care must be exercised to see that thei'e are no than normal chance for these to occur under the people in possible danger zones. often adverse field conditions of avalanche blast- 2. Avalanche accident statistics illuminate one ing, special care must be taken. Do not allow important point which should be a constant source frozen toes, cold fingers, or an unpleasant blizzard 61 to encourao-e short cuts or Cixreless preparation of a b. Must be repeated often so that : (•liar

62 cl. Detonators: (5) Pei-sonally arm, place, and fire the (1) Type: electrical, ignition, or prima- charges; aim and fire the gun. cor d. (6) Provide for party safety: routes, pro- (2) Size: for military demolitions—No. 8 tection from blast, secondary ava- or larger. lanches. For commercial explosives—as recom- (7) Provide for area safety: closures, mended by manufacturer. guards, connnunications, and signals. For primacord—No. 8 or two No. 6 taped (8) Prepare report on each operation as together. When used as a detonator, required by forest supervisor or other primacord will be attached to the ex- autliority: i.e.. Army Gun Book. plosive as prescribed in Dupont c. For training purposes, the blaster may Blaster's Handbook or Army Field supervise while an assistant performs any Manual FM5-25. of the above. 2. Storage. d. Assistants must : a. ^Magazines for explosives and detonators (1) Be competent ski mountaineers. will meet Forest Service Safety Code (2) Have proper equipment as directed by standards. They may be raised above blaster. ground to normal snow depth. (3) Obey blaster's instructions. b. Field caches : during operations season not e. Spectators. It is recognized that ava- more than 50 pounds of explosives—not lanche blasting operations have educa- detonators—may be stored in the field tional value. Spectators are permitted. convenient to the blasting area. Require- They must be : ments : (1) Briefed on safety requirements. (1) In a substantial, locked box. (2) Controlled, by extra guards if neces- (2) Plainly marked. sary. (3) Isolated location. f. Publicity: c. Primacord is classed as an explosive and (1) Will be avoided. will be so stored. Distinguish carefully (2) Scheduled demonstrations may be held from safety fuse. for press and public with prior ap- d. Ignition type detonators require special proval of forest supervisor. protection from moisture. Use water- . Technique. proof packaging in storage and ring or a. Preparing cliarges: cutthroat type crimping tool in the field. (1) Use standard procedures outlined in Taping joint between cap and fuse is Blaster's Handbook or Army Field recommended. Manual. 3. Transportation. (2) Make sure all materials and equipment a. According to Safety Code except over are in working order. snow. (3) Premature detonation of electrical caps b. Over snow, vehicle or backpack, require- by snow static is possible. Ignition ments : caps are recommended for avalanche (1) Redflagging. blasting. (2) Physical separation of detonators and (4) When using electrical detonators, explosive. check firing circuit with galvanom- (3) Detonators in box individually eter at each stage. Keep circuits wrapped or padded. shorted until ready to fire. If blow- (4) Projectiles in individual containers. hig snow is present, prepare complete c. Via lifts and tows : firing circuit before inserting detona- (1) Normally at times when not in use by tor in explosive. the public. (5) When using ignition detonators, test (2) lOO-foot minimum separation between at least 3 feet of each new lot of explosives and persons on the lift safety fuse for burning time. Fuse other than blasting party. may be cut and cap crimped on be- 4. Party Management. fore proceeding to the field. Cut a. Blasting party consists of the blaster and fuse for minimum 3-minute delay, at least one assistant. longer if necessaiy for protection of b. Blaster organizes and conducts all opera- blaster. tions: Pull-wire fuse igniters are recom- (1) Brief assistants and notify authorities. mended. (2) Assemble, check equipment and ma- Note the time when fuse is ignited. terials. Be sure fuse is burning before leavino- (3) Choose the targets. the charge. (4) Transport the detonators. Replace igniter if fuse fails to ignite.

63 If blaster lenves the charge and it fails d. Preparing cornice shotholes : to explode, it must be treated as a (1) Use coring tool or drill such as snow misfire, even if he thinks the cause sampler. is the failure of the fuse to ignite. (2) Place holes so as to split off overhang. (6) Primacord: (3) Stem the holes with snow. Is recommended for detonating multi- (4) Holes should be 6 to 8 inches apart, ple blasts. depth so that charge is in center of Is required for 6-shot blasts or over. snowmass. Attach ¡irimacord pigtails to main (5) Man working on the cornice must be line with 2 clove hitches, pulled tight. belayed. An ignition cap crimped on the end of e. Preparing preplanted cornice charges : each primacord pigtail and inserted (1) Primacord firing circuit required. in the explosive is recommended. (2) Charges should be securely anchored Primacord mainlines over 50 feet long againt snowcreep. should be doubled. (0) Protect firing circuit from snowcreep. (7) Amount of explosive to use : (4) Fire the blast well before cornice might Determined by blaster. fall naturally. Four pounds demolition or seismo- f. Misfires: graph powder generally sufficient for (1) Electrical detonator: check the circuit slope avalanche. For wider slopes and attempt to fire; rearm with a use multiple shots at 100-foot inter- new detonator and fire. vals. Do not use less than 2 pounds (2) Ignition detonator: wait 30 minutes; per shot. without disturbing the charge, place Where hard slab is prevalent, with or a new charge alongside and fire. without depth hoar, increase size of (3) If unexploded powder goes down the charges, up to 10 pounds per shot. slope in an avalanche, treat as a mis- Three to four sticks of dynamite per fire. hole is generally sufficient for cornice g. Projectiles : removal. (1) Follow Field Manual instructions. Use too mucli explosi\e rather than too (2) Avoid firing over buildings, other little. Get the answer. sti'uctures. b. Placing slope avalanche charges : (3) If necessary to fire over a structure, (1) May be lowei-ed or tossed into position. the distance gun-to-structure must The following precautions should be be not more than two-thirds the dis- taken : tance, gun-to-target. Detonator securely taped or fastened. (4) If necessary to fire alongside a lift, Charge securely packaged. the line of fire must diverge from No strain on lead-in wires of electrical the lift line. caps. Lower by blasting wire or (5) Area gun-to-target, to be clear of all separate line. personnel. Belay charge if snow is hard and it Exception—Personnel may be inside might slide. a fragmentation-proof structure, not If blaster goes onto avalanche slope to less than 300 yards from the target ])lace the charge, he will be belayed. area, not within the avalanche path, Belay j^ersonnel working on cornices. and not on the gun-to-target line. (2) May be di'opped from ski lift, but lift (6) Safety limit, gun crew to target: must carry no other passengei"S be- 600 yards. sides the blaster. (7) Beware of damage from dispersion Pull wire igniters required. Throw the of fragments—especially with 105 charge to one side and behind the mm. weapons. blaster. (8) Provide traffic control if firing on or c. Firing : near a highway. ( 1 ) Use a])proved warning signals. (9) Vehicle clearance from gun position, (2) Establish a minimum party clearance other than gun carriage, should be of .50 feet plus substantial terrain 100 yards for 75 mm. weapons and barrier. 200 yards for 105 mm. (3) Party location must be protected from (10) Prepare one round at a time. Keep secondary avalanches. other ammunition in container's. (4) Blaster must have visual observation (11) Conventional cannon : Provide clear- of the slide path and adjacent dan- ance for muzzle blast. Keep per- ger areas, or fire upon signal from a sonnel behind and to either side of guard. the breech.

64 (12) Eecoilless cannon: Provide clear- (17) Misfires: Proceed as instructed in ance for back and muzzle blast. Army Field Manual. Dispose of Keep personnel on a line at right defective ammunition by demoli- angles to the barrel. tion. (13) Visual firing: Check sighting equip- ment by boresighting before each (18) Duds: Note point of impact. Make operation. Do not fii-e wliere an search for dud as soon as possible. "over" endangers life or property. When found, dispose of in place by If range is unknown, use ladder demolition. Be extremely careful method of getting on target. Place not to disturb the dud. Secure first sighting shot below target or military assistance if necessaiy. on an adjacent, higher slope. Make (19) Care and maintenance: i'ollow Army a range chart. Field Manual instructions. If ( 14) Blind firing : Target data must be pre- rifles are left in the open remove the determined by visual firing. Double breechblocks wlien not in use. In- check the lay of the giui. Provide a spect the rifle before using for de- fixed gun position and aiming stakes. Recheck target data fre- fective parts, especially sight mount quently by visual firing or aiming. and obstructions in the bore. Check ( 15) Firing on slope avalanches : Use point all moving parts for malfmictions detonation fuse. Exception—If before loading first round. an unstable layer within the snow- 8.34 Exceptions to Safety Code. Due to the pack is suspected, fire first with special conditions of avalanche blasting—absence point detonation fuse, then with of flying debris, danger from snow static, for 0.05-second delay fuse. example—certain exceptions to tlie Forest Service (16) "When firing on cornices: Point tar- Safety Code are authorized; summarized as get weapons such as the recoilless follows: rifle may be used effectively against 1. Hard hats are not required. cornices. Do not employ unless 2. ]\Ietal cutters are permitted for opening there is a backslope or sufficient boxes with steel strapping. dead space behind the target to take care of an "over." Use the ladder .'5. Ignition type detonators are authorized and and burst-on-target method of get- recommended, ting the exact range, with super- ■i. For electrical detonation. Army combat wire quick fuse. "WHien a direct hit can or eciuivalent is authorized as lead wire. be obtained on the cornice, use delay Maximum length is 200 feet. Damaged fuse. wire should not be repaired or spliced.

65 2332.81 SNOW AVALANCHES Chapter 9

9 AVALANCHE DEFENSES. Avalanche bar- Although long lasting, masonry deteriorates from riers of all types are much moi'e extensively used the effects of frost and Avater seepage, and repairs in Europe tlian in the U.S.A. In the Alps, towns, are a major project in themselves. Walls and lines of supply and communication, farms, and similar continuous barriers have other disadvan- factories are exposed to avalanche hazards (fig. tages. They are very costly and frequently in- 48). crease the load on a slidepath or cause the develop- The need for avalanche defense structures in the ment of cornices. United States is increasing rapidly. Motor traffic European engineers have studied the forces ex- on mountain highways is growing heavier and erted by both avalanches and snowcreep. This more passes are being kept open each winter. Dam has led to the development of fence-type struc- builders, loggers, miners, and winter sports areas tures which are cheaper to install and maintain, are moving farther and higher into the deep snow and are more efficient than masonry. country. 1. ArJherg Fence. This is the prototype for In addition to some snowsheds and tunnels in all desigiis, so-named because it was invented to the Western States, only a few barriers have been protect the railroad in the Arlberg Pass (fig. 51). constructed by highway departments and lift A typical installation utilizes railroad steel posts operators in recent years. For the latest develop- set in concrete 8 to 10 feet apart. Cables which ments in this field a pei-son must turn to Austria support vertical poles spaced 4 to 6 inches apart and Switzerland. Both countries maintain re- are fastened to the posts, which are braced uphill search establishments to develop and test new with guy wires. The height seldom exceeds 12 desigjis. Significant progress has been made feet. recently. This makes a very rugged structure that has The modei-n trend is away from avalanche walls proved itself over the years. Modern engineering and towards fencelike retaining structures which has added many innovations in designs, materials, do the same job as well or better at less cost. The and layout. One of the most important advances trend is also away from solid or continuous retain- has been to determine the optimum angle of in- ing structures. clination of the fence with respect to the slope. 9.1 Diversion Barriers. Diversion barriers give Originally, avalanche fences were built perpen- protection by diverting avalanches—over high- dicular, regardless of the grade. Modern practice ways or railroads, away from villages, around is to set the fence at an angle slightly more than electric transmission pylons (fig. 49). Walls, tun- 90 degrees from the slope on the uphill side. This nels, snowsheds, and bulkheads are examples of arrangement utilizes the full height of the fence diversion-type barriers. Although they have the and snow pressure drives it against, not away effect of slowing down or halting an avalanche, from its anchorage. It is also possible to brace they do nothing to prevent it from starting and the fence from the downhill side so that ava- little to limit either its size or penetration. Some- lanche and snowcreep do not affect the supports times they are the only solution. The Flexen- (fig. 52). ... strasse, access highway to tlie famous ski centers Avalanche engineers are experimenting with of Zurs and Lech in Austria, would be blocked or treated and untreated wood, steel, aluminum, ny- hazardous most of the winter without its elaborate lon, prefabricated concrete, and combinations of system of snowsheds and tunnels. Stabilization these. They are also testing various fence designs, of these slidepaths would be too expensive (fig. including the spacing and arrangement of the bars (horizontal, verticalj lattice, or net type). The 9.2 Stabilization Barriers. Barriers of this objective is to obtain maximum efficiency with type attempt to control the avalanche hazard by a minimum amount of material. Kegardless of holding the snow in place or reducing its move- how the bars are arranged, it has been found that ment to minor proportions. To be effective, they if airgaps in the fence exceed a width of 6 inches, must go to the top of the avalanche path. In very dry and very wet snows flow through them. former years, design was limited to walls and The design of interrupted fence patterns com- terraces. Such structures are difficult and costly ing into use takes into account that it is seldom to build where the avalanche release point may possible to keep avalanches from starting under be several thousand feet above the valley floor. all conditions. The interrupted pattern requires

66 .^*!*^:'*'-

t '

0"

'■uK ,t ■ j^ I.

V-- > if-

r^ r •••.•V-.* ''/>.>-'' 'Mgy- -^ -^

-« j»*^'^*

F—49'»212 FIGURE 48.—A century of avalanche defense at Davos, Switzerland. Ancient masonry walls and terraces with Arlbere fences added later, and flanked by modern fences. The six dots near the center of the picture are wind hiffli^ c-ontroUing an adjacent slide path. ocimei. less fence to cover a given area and is less likely to middle, are hung with considerable slack and act as a snow catcher (fig. 53). anchored upliill. Among the fillers under test 2. Nets., Hedgehoga, Trianghs. In their search are cable mesh, wire mesh, and horizontal wooden for effectiveness at least cost in labor and material, European avalanche engineers have developed the poles. Results to date favor the wood poles avalanclie net. In the Austrian version the net spaced 4 inches apart. The poles counteract the is hung on two posts of ingenious design, like nefs tendency to fold together when struck by an very large ski poles. Eight to ten feet apart, these avalanche. The virtues of this net are simplicity, posts are driven into the ground and guyed above ease of construction, and flexibility. It can roll and below at an angle to the fall line (fig. ,54). witli the puncli of an avalanche. An installation The net cables, located on each side and in the using nylon nets has recently been made in France. 67 F-40Ô213 FlGIMíE 4!l.—('(Hiérete l)ulliliead guarding an electric transmissiiHi iiyhm. Note diversion wall protecting houses to right, Arlberp fences above and snowshed at ui>i>er left. Stuben, Austria. FlOUKE 51.—Old style Arlberg avalanche fence. Installed for protection of pipeline in foreground.

HRLREHG FEhlCE iO 'HISH 70% EF^tCTIVt

MODBaN FENCE 10'HIGH /ooyo EffecnvE

FIGURE 52.—The original Arlberg vs. modern avalanche F—I fence, .showing angle of inclination and effect of snow FiGUKE 50.—The Flexeustrasse snowsheds. l)ressure (broken arrow).

68 NCT CAbLt ANCUOR

STEEP

GUY^,

y TRANSITION

5TEEP

QbLIQU^

FIGURE 53.—Avalanche fence layout, interrupted pattern.

3. Windbaßes, Drift Fences. The windbaffle FIGURE 34.—Austrian avalanche net. (Kolktafel) is a panel set at right angles to the prevailing wind in a slab area. It was based on Emergency airfield mats have proved successful the supposition that a system of these panels would for fence material. They come in perforated steel break up the current of windborne snow and panels 8 feet long and 18 inches wide, convenient create islands of strength and discontinuity in any to transport and erect. U-bolt fasteners make it slab that did form and, perhaps, inhibit the forma- possible to slide the panels up and down the posts tion of depth hoar. Baffles do some good but are for adjustments in height and airgap. With not an absolute control. In Europe they are used proper bracing they are strong enough to use as to reinforce or supplement other defense measures defense stiiictures (fig. 57). (fig. 55). 4. Avalanche Mounds. Momids are a type of Drift fences have long been used along high- braking or diversion barrier. They are effective ways to alter the fallout pattern of windborne when installed on a slidepath that includes a tran- snow. Strategically located fences can reduc« the sition zone. The crosscurrents, internal friction, amount of snow deposited on an avalanche path by and age-hardening process set up by the mounds in causing premature fallout. In a similar manner, a falling avalanche reinforce the natural slowing a fence can modify the windflow over a ridge in down effect of the transition. such a way as to reduce or prevent the formation Avalanche mounds originated in Europe, where of overhanging comices (fig. 56). they have proved their value (figs. 58 and 59). Conventional drift fences, made of wooden slats A system of mounds has been constructed by the woven together with wire, are not strong enough highway department on tlie Willow Swamp slide- for the winds that blow over the peaks. At the path in southern Colorado that has proved to be present time, the location, height, and spacing successful. Mounds have effectively protected two are a matter of trial and error on the ground. lift towers and two lift terminals in Squaw Val- Every installation must be a tailormade job de- ley. In Alaska mounds have been installed to pending upon terrain, prevailing wind direction protect highways near Anchorage. These have and force, amount of snowfall, vegetation, and proved very successful, including those constructed nature of the hazard. on a slope above Turnagain Arm where the 69 inouiitíiiiiside falls steeply for about 3,000 feet witliout any transition slope. They have been ])articnlarly eflfei-tive in arresting wet slides on these steep slopes (ñg. 60).

^\ X f<

F-495218 FIGURE 57.—Snow fences made of steel landing mats. These are used to prevent cornice formation near the summit of Squaw Peak, California (see fig. 56).

FIGURE 55.—Recently designed Austrian windbaffle, omni- directional, tripod suspension, nose of baffle anchored to ground. The material is aluminum strips woven together with wire.

\^:^

F-495219 FIGURE 58.—Arzler Aim and the original system of ava- lanche breakers. Arrow indicated the mound of waste material which was the model for present-day design. Note natural reforestation behind the breakers.

Where terrain favor's the construction of mounds, they can be Iniilt economically and quickly witli bulldozers out of the material at the site. Fill should always be taken from the up- hill side of the mound. Spacing should be as close together as possible, leaving enough room between mounds so tliat equipment can work and get suf- l'-405217 ficient material. To be most effective they should FiGUUE 56.—Arrow ¡wints to cornice on Squaw Peak, controlled by drift fence. Compare absence of over- be alternated as shown in figure 61, although hang with adjacent, uncontrolled cornices. exact layouts are not essential. The practical 70 .^s!^:

^

w^ ■~i t .it: - y-.^-

FIGURE 59.—Modern avalanche mound system, near Innsbruck, Austria.

height of mounds is generally 12 to 15 feet. How- 2. To determine the spacing of avalanche ever, 20 feet may be necessary in areas of excep- barriers, construct a right triangle using the slope tionally heavy snowfall. as the hypotenuse (fig. 63). The perpendicular Mounds have been successfully installed on leg is a contant 45 feet. Project the horizontal grades up to 2.5 degrees and experimentally up to leg until it intersects the line of the slope. The 35 degrees. The steeper the grade the longer the lengtli of the hypotenuse determines the spacing transition zone must be for an effective mound of the barriers, under average conditions, for the system. particular grade. Once the profile of the slide- 9.3 Barrier Design Factors. Due to variations path is known it is possible to plot the barrier sys- of terrain and climate, the following data must tem on a map or diagram. be taken as rules of tliumb rather than exact 3. The following rules of thumb apply to the formulas. Used in this manner, they are helpful construction of barrier fences (fig. 64). guides in designing avalanche barriers. 1. The pressure exerted by snowcreep against a a. Distance between posts—8 feet to 10 feet barrier, or any obstruction, can be estimated by b. Distance between bars—4 inches to 6 inches using the formula in figure 62. c. Fence angle—105 degrees on the uphill face Experiments with pressure gages have revealed d. Support—35 to 40 degrees attached two- that the force exerted at the ends of a barrier is thirds height of fence from two to two and a half times greater than at 4. Under average conditions, the wind-shadow the center. These effects are due to the plastic of a fence extends for a distance 11 times its nature of snow. height.

71 F-495221 FIGURE 60.—Earth mounds used as avalanche breakers near Girdwood, Alaska. Though a transition zone is absent on this slope, these mounds have proved an effective defense against wet slides. 9.4 Reforestation. From the long-range view- lanche defense project should include reforesta- point all autlioriries agree that a forest is tlie best tion if feasible. and most economical avalanche barrier. A cen- The practice in the Alps is to reforest imme- tury ago, according to Austrian engineers, there diately where mature timber has been destroyed was no avalanche Itazard at Innsbruck because of by avalanche action and install barriers to protect its screen of timber. The hazard has been created the seedlings. by forest fires and logging. The Austrians and Swiss are experimenting To be effective, the forest cover must be dense with methods of protecting seedlings. Posts set and extend to the avalanche release points. on the uphill side give good results. Other pro- Where forest cover lias been destroyed by fire, tective devices include tripods and wickets. In many cases, trees are planted in the shadow of logging, or climactic avalanche cycles, the problem complete avalanche defense systems. The Aus- is to reestablish tlie cover. Some great slidepaths trians favor the "island method'' of reforesting a are dormant for decades at a time and reforest denuded area or pushing timberline to higher themselves. Others are active every winter. elevations. They have learned from experience Snowereep often prevents the establishment of that the mortality rate is lower where seedling young trees. Reforestation on slidepaths, whether are planted near established growth, even if it is natural or by man, needs protection. Any ava- no more than a bush.

72 A V A LANCM6

> 4-5 FT.

PLAN

AVALANCME

TWNSITION SLOPE

PR,Of-iL-E;

FIGURE 61.—Avalanche breakers. > -fJ'A/'

FIGURE 63.—Avalanche fence layout. The separation de- pends on the slope angle, the fences being located at 4.j-foot contour intervals.

^j or HE/sur ' AttA Of ftWCt OR OBSTRUCTION ( HEIGHT TIMES WIDTH) : tfEtCTIVt LtNGTU OF SNOW MA55 A-L-W- SIN X ACTING ON ObSTaUCTION, (TUII££ TIMEÏ DEPTH OP i,NOw) ■ DEkJSITV OF SNOW MASS (r. Of WAT Eli IM THE SNOW TIMES 62 4-THe WT OE A CUfT Oi^ WAT£li ) ' ANGLE Of THE SLOPE

FIGURE 62.—The calculation of snowcreep pressure.

FIGURE 64.—Avalanche fence design, showing optimum angle of inclination and bracing arrangement.

73 2332.81 SNOW AVALANCHES Chapter 10 10 AVALANCHE RESCUE. Prompt and or- cue caches must be properly stocked and easily ganized rescue operations are the only hope of accessible, and their location and contents known getting the victim out alive if a human being is to all who might have occasion to use them. buried in an avalanclie. There are records of per- Caches must be kept sealed until actually needed sons who lived as long as 72 hours while buried. for a rescue operation. The establishment of Ordinarily they are either killed instantly, or die properly stocked rescue caches must be one of the within a short period from suffocation, bodily in- first steps in developing a ski area where ava- jury, or shock. lanche danger may exist. Investigations of a number of avalanche acci- 1. The minimum equipment that should be dents, fatal and nonfatal, lead to the conclusion available at all times in a readily accessible loca- that 2 hours is the average survival limit. Snow tion is as follows: is porous and ordinarily contains enough air to 12 probes, sectional preferred; otherwise support life, thougli not consciousness. It appears metal rods, i/^-inch diameter, minimum 12 that in about 2 hours an ice mask, from condensa- feet long. Aluminum thick-wall conduit tion of the victim's breath, forms an airproof seal is recommended. around his face. Rescue operations are therefore 12 snow sliovels designed to get the victim out witliin the 2-hour 12 electric headlamps with carton of spare limit. Due to special circumstances which pro- batteries long the life of the victim—he may be in an air 100-foot climbing rope pocket—rescue oi)erations nuist not be abandoned Five yards of reel flagging cloth for at least 24 hours. 200 feet of i/4-incli nylon cord Successful avalanclie rescue opei'ations depend First aid kit (standard ski patrolman's kit upon special equipment, trained leadership, and satisfactory) manpower. Darkness and storm are normal ; Toboggan with blankets. Collapsible tobog- parties setting out to participate in a rescue ac- gan preferred, light enough to be back- tion must be prepared for the worst that weather packed and snow conditions have to offer. The rescue 2. The following additional equipment is leader must disi)atch only groups of men who are recommended : properly clothed, equipped, and in good physical Emergency rations condition to face the rigors of exhausting work Chemical hot pads for toboggan in deep snow at high altitudes. Ski climbers in assorted lengths Obviously the accident would not have hap- pened if avalanche hazard did not exist. This Additional probes same hazard condition is a source of danger to Gasoline stove the rescue party. Therefore, the highest degree Camp cook kit of caution is required both during the march to Mountain tent the accident scene and while rescue work is in Gasoline lanterns progress. In an avalanche rescue operation in Portable radio 19.58, two independent and uncoordinated parties 3. Highly trained German shepherd dogs have set out for the scene. One of them, following come into widespread use in the Alps. Time and high ground, dislodged a slide whicli buried a again they have located buried avalanche victims member of the other party below,.resulting in a rapidly and efficiently. Avalanche case histories second fatality. The strictest observance of all show that buried victims have sometimes been precautions for traveling in avalanche terrain located by untrained dogs who happened to be on must be enforced to prevent such accidents. Rescue opei-ations of all kinds can easily become the scene and instinctively joined in the search. disorganized uidess tliey are directed with a firm Therefore, whenever possible a dog should be taken hand. The most effective rescue is one which is on rescue operations, whether he has been trained carried out with military precision and rigid or not. discipline. Proposals to locate victims with mine detectors 10.1 Rescue Equipment, Rescue equipment is or other unusual means have not as yet proved needed imediately when an accident occurs. Res- practical. 74 10.2 Immediate Action. for extended rescue operations with everything 1. Snow ranger will take charge of rescue opera- except food. It may be dispatched in groups with tions and sound general alarm. He sliould not at least probes and shovels. hesitate to requisition any needed equipment or 2. The snow ranger will serve as leader after call for volunteers. All ski patrolmen and ex- notifying the county sheriff or designated disaster perienced skiers will be ordered to report to a authority. People not physically qualified for the central location. exliausting work of rescue should not be taken. 2. The informant or eyewitness should be ques- They can help by preparing and transporting tioned on exact location of accident. Even if in supplies. poor physical condition, eyewitnesses should re- 3. In the 30 to 60 minutes at his disposal, the turn to the accident location with the first party to snow ranger will organize and equip the followup point out where the victim was last seen. This party, notify autliorities outside the area of the is extremely important. accident and of possible need for further help. He 3. Dispatch the fii-st party. A snow ranger will obtain, if possible, the correct name or names does not necessarily lead this party. It must be of victims and appoint a leader to take charge of in command of a reliable ski mountaineer and any further reinforcements and equipment. first aid man, preferably trained in avalanche 4. Before beginning to jirobe, the followup rescue. It consists of not less than three persons. party must make a tliorough search of the slide. Five is a desirable number, large enough to do the They should foi-m a single line, shoulder to work, small enough to travel fast. Tliis party shoulder, and march back and forth across the should be dispatched within 15 minutes after slide debris kicking and scufling at the snow to un- receiving the alarm. cover any additional clues and possibly the victim ■i. The first party travels light witli whatever or his e([uipment, which might lie somewhere just equipment can be picked up at once. Speed is under the surface. Systematic probing of the slide the first consideration. begins at the tip and works up tlie fall line toward 5. Upon reaching the location of the accident, the last-seen point (fig. fî5). Rescuers are spaced the leader posts an avalanche guard, unless he shoulder to shoulder and probe eveiy square foot decides this is not necessary. Leaders are respon- (fig. 66 and 67). If the victim is not located on sible for the safety of their parties. Avalanche the fall line below the last-seen point, or the latter guards must be kept on duty if there is the slight- is not known, the entire debris cone nuist be sys- est danger of another slide in the same vicinity. tematically probed. It is very important to mark Rescuers must know where to go in case of an the probed areas as the search is completed, so that alarm. If the danger becomes critical, leaders no effort is wasted by duplication (fig. 68). Flags must not hesitate to call oft' operations. or other nuirkers foi- this purpose are an essential 6. Locate the spot where the victim was last part of the rescue equipment. observed and mark the "last seen"' point so that 5. The first party may be relieved and sent out it cannot be obliterated by wind or snowfall. of the area. lixhausted rescuers will become From this point downhill, the first party makes casualties themselves if allowed to remain. Thej'^ a hasty search of the slide surface for the victim can cai'ry messages giving an estimate of addi- or any of his equipment. tional equipment and assistance needed. 7. If any indication of the victim is found, begin 6. Shovel crews will accompany the probers, re- probing in the vicinity. If no indication is found, lieving them at intervals and digging in any likely begin to probe in likely locations. These places spots. They should be prevented from digging are obstructions in the slidepath such as trees, hai)hazardly and wasting energy. boulders, or transitions, also the tip and edges of 7. Small slides may be systematically probed in the slide. A human ]>ody is bulky and all other a few hours. Large slides may take much longer, things being equal, is likely to be thrown toward especially if the snow is deep and hard. If the the surface or the sides. probing action is unsuccessful, the slide must be 8. If the victim is found, give first aid. ITnless trenched. This work, where possible, should be danger of further avalanches is acute, the firet under the supervision of the county sheriff. party should not attempt to move tlie victim out 8. Trenches are dug parallel to the contour down of the area. Regardless of physical injuries, the to ground level or undisturbed snow at intervals victim must first be treated for suffocation of 6 feet. If sectional probes are available the in- and shock. Get the snow out of nose and mouth. terval can be increased to 10 feet. Digging begins Apply artificial respiration if breathing has at the tip of the slide and proceeds uphill. It is stopped. Get the victim waim, applying body best to space the shovel crews along one trench heat if nothing else. with frequent reliefs. In this way snow from one 10.3 Followup Action. trench can be thrown into the one just duo-. 1. In a well-organized ski area, the followup 9. The operation ceases to be of an emergency party can usually start in half an ho)ir. It must type if trenching is nece,ssary. A constant system start within an hour. It goes completely equipi)ed of relief crews must be organized.

75 F-495222 FIGURE 66.—Close-interval probing requires insertion of FIGURE 65.—Arrangement of a probing line on an ava- the avalanche probe by the right foot, between the lanche path. It i.s much easier for the probers to feet, and then by the left foot. The prober then ad- keei) in a straight line if two men hold a string to vances one step and repeats the sequence. Such close mark l(K*ation for the probes, and move this marker probing is esx)ecially important in deep avalanche forward each time the line advances. debris (see fig. 67).

10.4 Avalanche Accident Report. Time followup party arrived at scene of 1. Date, time, and location of accident. accident. 2. Names of victim and other members of party, Procedure at slide. with any information on their mountain Time and location of victim when found, in- experience. juries sustained. 3. Summary of events leading up to accident; Cause of death or injury. departure point, route, and objective of Time operation was concluded. party. 6. Weather and avalanche hazard background; 4. Eyewitness account of accident, if available. wind, temperature, and snow data; restric- Otherwise, observer's deductions based on tions in force, if any; type of slide and extent. tracks and any other evidence. Important 7. Terrain data, including maps and diagrams. points are: location of skier in relation to 8. Recommendations and conclusions. release point of slide; how was slide released ? 10,5 Case Histories. 5. Summary of rescue operations. Times and 1. Case History No. 1 (fig. 69). names are imjwrtant. a. Memher.'< of party: A and B, guests at Time of accident. , excellent skiers; en- thusiastic but inexperienced ski moun- Time report of accident received and from taineers. C and D, ski instructor and whom. ski patrolman, respectively, excellent Time finst party dispatched. leader. Num- skiers and experienced mountaineers. ber in party. Events leading up to accident: C and Time first party arrived at scene of accident. D were going to test the Run, Time followup party dispatched. Leader. known to be dangerous at that time. A Number in party. and B went along at their own request.

76 If / -;>jí£íA

dtipiu buried victim

deflection of proof; witfi deptfi

FIGURE 67.—Why a close probing interval is necessary.

FIGURE 68.—String marker method used to guide probing operations. The string shown tied to the ski pole is moved one foot ahead after the complete line has been probed. The other string stays in place as a perma- nent marker for the area probed.

77 \ 4/KIER/ENTERED AREA HERE /y AT HEAD OF DRAU/. -/; /'i r- / I W

Poji+ion of Aierr //.• TA. ^K

i-^ ///-.. ''% ;,/ ^\ ^V

L^ '///,..

'/, ^. ''--/^ ^ .s- / >/ /• >iï5Uteîji'4"" ^^^"^ Ledge , 10' hiqh Poritiort^of 'B' when sV^t'Q. /tjr+«J- \^;;V^>=. 'A'found h€re,af+€r+renchinq. . • •• Trenche/. ?^jfc;- Area probed. /// Tip of ifTx\\àe buried v^ bij 2nd j-lidc-"

V, % ^--/ir. % ^'^(^''"Ony

FIGURE 69.—Avalanche accident. Case 1.

78 b. The accident: C skied onto the slope, fol- and it was blocked by snowslides. The lowed immediately by A and B. Slab, weather was cloudy at the elevation plus dry snow, avalanche fractured where the pickup truck and the snow- while all three were on the slope. C plows were operating. The plow was made an experienced mountaineer's typi- working in a slide; the pickup was cal test run of a doubtful slope: a short parked a short distance behind. An fast dip in and out of the danger zone. avalanche observer had earlier warned Only tlie edge of the slide touched him. the highway crew of the hazard and B and A went farther out onto the slope. recommended no traffic of any kind. B, nearest to the release point, managed d. The accident: An avalanche on the same to ski out to one side. A attempted to track as the one that had already blocked outrun it straight down the gully. He the road buried both the plow and the went over a rock ledge, apparently fell, i:)ickup. The rotary plow, being higher and was picked up there by the slide. and heavier than the pickup and close B, looking for A, went out onto another to the iiphill bank, withstood the force avalanche slope and started a second of the slide. C and D were trajjped in slide which buried the tip of the first one. the cab but not injured. The pickup, c. Rescxie operation.^: Rescue operations were parked near the doAvnhill bank, took the prompt, considering the distance from full force of the slide. All glass was help, and were well organized. Accident broken. A and B were packed in the happened at 2 p.m. First party reached cab as if in concrete. The avalanche the scene at 3:80 p.m. Followup party was probably a temperature-released reached the scene at 9:45 p.m. Victim slab. was found at 5 a.m. the following morn- e. Rescue operations: E, the driver of the ing. Elapsed time: 15 hours. Cause of supply truck, arrived on the scene shortly death : suft'ocation. Immediate action after the accident. He saw the exhaust was unsuccessful. Systematic probing stack of tlie rotary, which projected out was unsuccessful. Body was finally lo- of the snow, and freed (• and D. The cated by trenching. tliree then departed to get help without d. Weather and hazard data: Successive attempting to locate the pickup. On the storms followed by prolonged cold. way out of the canyon they encountered Snow had not settled and was unstable. the three skiers traveling on foot, and e. ConrJusion><: The following errors were the told them what had liappened. The direct cause of accident : skiers proceeded to the avalanche, founcl (1) A and B went onto the slope at once the pickup by means of its radio mast, instead of waiting for C to complete and dug out A and B with their skis his test nm. and hands. A recovered quickly, but B (2) A and B went too far onto the slope. was dead. (3) A tried to outrun the avalanche. f. Weather and hazard: See above. (4) The second slide started by B compli- g. Terrain data: The accident happened cated rescue operations. Under the where a known major avalanche path conditions set, the only possible means crosses a highway. of locating A quickly would have been h. Conclusions: The accident was caused by a trained dog. failure to heed a definite hazard warning 2. Case History No. £. and failure to recognize the possibility a. Location: An alpine highway. of a repeater avalanche. It was faulty b. Members of party : A, highway crew fore- technique to have both vehicles and all man and B, in a pickup truck; C and D, members of the party exposed to the plow operators in a rotary snowplow; same slide. If C and D had remained E, supply truckdriver; 1, 2, 3, skiers behind to look for their companions traveling the highway on foot. while E went for help, there would prob- c. Events leading up to the accident : At high ably have been no fatality. elevations a prolonged and severe storm i. Recommendation : All men working in haz- was in progress. Because of the hazard ardous areas should receive avalanche the highway was closed to public travel. rescue training.

79 2332.81 SNOW AVALANCHES Chapter 11

11 AREA SAFETY PLANNING. Use of an Vertical photomaps have a special value in locat- area exposed to avalanche hazard requires a sys- ing cross-country routes. Generally it is possible tematic snow safety phin. This phm is tailormade to i)ick out the major slidepaths at once and make to fit tlie area. A safety plan for an isolated a tentative route selection. structure like a mine building could scarcely be The purpose of the terrain analysis is to identify applied to a highway extending for many miles. the slidepaths and the zones of safety, to esti- Neither plan would have much value for a com- mate the effect of natural barriers, and to deter- plex development such as a winter sports area. mine the hazard pattern of the area. This pattern All safety plans, regardless of the type of use will govern the location of improvements. Build- or character of the area, have the same objective— ings are placed to take advantage of the natural maximum use with adequate safety. Knowing barriers. If this is impossible for any reason, the the type of use contemplated or already developed, fact must be noted and recommendations made for the observer proceeds according to the following: protection by other means. It is seldom possible, 1. Terrain analysis for instance, to survey an alpine highway on an 2. Climate and weather analysis entirely protected route (fig. 71). However, a 3. Hazard classification hazard reconnaissance prior to construction will 4. Operations plan. suggest ways of reducing the number of avalanche 11.1 Terrain Analysis. Terrain analysis should paths which must be crossed. be made on the ground. It is desirable for the For a winter sports development, the layout of observer to see the area in both winter and sum- the ski runs is as important as the location of mer (fig. 70). Photomaps of excellent quality buildings (figs. 72 and 73). Minimum hazard are almost a necessity for this work. Topogra- areas must be identified. Lift sites are chosen so phical maps are of limited value for analysis, since that they not only open up desirable ski slopes but also make it possible to protect the people they lack sufficient detail. They can be used to using them. The lift line itself requires careful record the results after an analysis has been made. study so that terminals and intermediate towers The low oblique photomap is the most desirable, will not exposed to avalanche danger as in figure followed by ground level panoramas and verticals.

LkLk F-495224 ¡•iGVKE7().~Terrai>t analysis: (1) Major slide paths. (2) Lesser slide paths. (3) Narrow gully. (4) Cornice for- niation. (o) Convex .slope. (6) Concave slope. (7) Safe area, no avalanche danger. (8) Cliffs and steep r^Lf'*'r'^ (fractiire zone). (!)) Gentle, open slopes—heginner's .ski area. (10) Pos-sible lift location. (11) Arrow Ro • »..^"""r;!"? ^^^ tenninal just beyond pass. Almost all sloi)es shown are accessible from this lift. Albion üa.sin. Alta, L tah.

80 Ordinarily, an original climate and weather analysis must be based on superficial data: some temperature, maximum snow depth, and total pre- cipitation records ; some word-of-mouth history of avalanche occurrence. To these, the inspector adds his own observations in the field : prevailing wind directions, snow types, and storm charac- teristics. This information, even though frag- mentary, should permit him to form some opinion on the predominating types of hazard. 11.3 Hazard Classification. After analysis of the terrain, climate, and weather, the next step is a hazard classification of the area (fig. 75). Iden- tifications must be accurate and should be plotted on a map. All factors must be considered : grade, contour, and dimensions of the various slopes; natural barriers; exposure to wind and sun; type of use ; the avalanche history of the location, and the protective measures that will be available. Four degrees of liazard are recognized : 1. Minimum hazard. This classification indi- cates practical absence of hazard. Exam- ples : A building fully protected by natural or artificial barriers; a slope not likely to avalanche. 2. Low intermittent hazard. This indicates occasional exposure to avalanches of dangerous size. Examples: A slope not steep enough to avalanche in dangerous volume except under extreme conditions ; a FIGURE 71.—An example of avalanche hazard. A major avalanche path, with vertical relief of nearly 4,000 structure close enough to a slidepath to be feet, overhangs the access highway leading to a major damaged under climax avalanche condi- ski resort. Protective measures used are road clo- tions. sures during high hazard, and artificial release of avalanches by artillery fire. 3. High intermittent hazard. This indicates areas frequently subject to avalanches of dangerous size. Example: A high-angle The trend in lift design is changing as a result slope of sufficient dimensions so that hazard of damage to lift lines caused by avalanches. It is likely to exist with every major storm used to be thought necessary to install a lift on or under delayed-action avalanche condi- many towers so that the cable would closely follow tions. (The distinction between low and the land profile. Numerous towers multiply the high intermittent hazard is frequency.) exposure to avalanche hazard—and snowcreep, which can be just as destructive. Modem design 4. High intermittent hazard, not controlled. utilizes a minimum number of towers and long Indicates highly hazardous areas which spans extending over the avalanche paths. are not feasible to control. Protection of the lift line and structures from The classification of an area may be changed if avalanche hazard and snowcreep should always protective measures are applied. Thus an area be fully considered. would take a lower classification if it were put 11.2 Climate Analysis. It is more difficult to under stabilization by use, explosives, or barriers. make a climate and weatlier analysis than a terrain The reverse situation is also possible. analysis. In a new area, the data are usually in- 11.4 Operations Plan. The final step in snow complete. The kind of information an avalanche safety planning is to formulate a plan of opera- observer wants generally is not recorded. Some tions. Where a number of different activities are information is always obtainable by studying the involved, such as a ski area and a highway, with available records, by questioning inhabitants or several cooperating agencies, the operations plan people familiar with the area, and by studying the can be complex. Ordinarily it is a brief and conditions on the ground. Persistence and ingenu- simple outline of procedure. A snow ranger, well ity are required. In one case, a snow ranger trained and thoroughly familiar with his 'area recommended minimum lift clearances on the basis seldom needs to look at it. He knows it by heart.' of snow course records at a station 20 miles away The same is not true for his successor or a tem- but in a similar location. The recommendation porary replacement. Either could easily get into was generally thought to be excessive, but it was trouble without a systematic plan of action. The accepted and proved to be correct the next winter. following is a plan outline : 81 Snow Safety Operations Plan V. Table of Facilities I. Objective 1. Explosives magazine and field caches II. Ivesponsibility 2. Headquarters 1. Leadership and control ;i Gun towers 2. Finance 4. Shelters ;3. Personnel 5. Barriers 4. Equipment and facilities. III. Table of Organization VI. Procedures 1. Snow Safety Leader 1. Area hazard classification 2. Assistants 2. Protective measures ;'). Keinforcenients. 3. Opening and closing procedures IV. Table of liquipment 4. Restrictions procedures: signs Rifles 5. Sequence of avalanche control opera- 2. Explosives tions ;5. Snow-Weather-Avalanche observation (>. Search and rescue equipment Where national-forest land is involved, a snow 4. Communications 5. Rescue equipment safety plan approved by the forest supervisor is (i. Transportation, wheeled and oversnow customarily a permit requirement. It covers ski 7. Siffns. patrol functions and plans for handling other

.iiJijL

F-+nr,226 KiGUKE 72.—A new .ski slope in tlie process of development. (1) Lift line partially cleared. (2) Only observed avalanches on the mountain originate in this area of scattered timber. (.3) A natural clearing. (4) An ava- lanche Kully flankins the new area. (5) Area of heavy timber, very steep but completely stable. (6) Construc- tion road. Switchbacks .screened behind timber. Arrow at right, prevailing wind direction. The sloi>e faces north.

82 hazards related to ice, breakable crust, insufficient woukl insure a sound basic safety plan taking- snow cover, high wind, low visibility, and lift advantage of natural barriers for the protection operations. of major installations, and reducing the need for Where possible a snow safety reconnaissance expensive artificial protection measures to a should be made in advance of development. This minimum.

F-49D227 FIGURE 73.—The slope shown in figure 72 after develop- ment. ( 1 ) The liftline, held to minimum width for wind protection. All defective trees removed. (2) A ski trail cut around the mountain. (3) Ski trail partially cleared around the other side of the moun- tain. Timber between trails left for avalanche pro- tection. Tree clumi>s below are al.so left so that no avalanche path the full length of the slope is created. (4, 5, 6) The.se isolated trees and part of the clump were later removed. (7) Jumping hill. Tree screen retained for wind and avalanche protection and to act as a natural barrier between skiers and the avalanche F-495228 gully. FIGURE 74.- -Avalanche destruction of a lift tower.

tfA

■^4»\^

'■• ^- "^Ài--- Vd^^f t'^^i^ A^ 'A^w^■ .V .^

ill Liiii ' *>)'l

,, F-49.5220 i'lGURE 75.—Hazard classification of a major ski area. 1, Minimum hazard. 2, Low intermittent hazard 3 Hieh intermittent hazard. 4, High intermittent hazard not controlled (not illustrated here).' '

83 Bibliography

The followinjí list encompasses some technical Fuchs, Alfred references on snow and avalanches available in the July 1957. Effects of Explosives on Snow. SIPRE English language. There exists a large body of Special Report No. 23. Nakaya, U. literature in otlier languages—principally Ger- 1954. Snow Crystals. Cambridge: Harvard Uni- man, French, Italian, Russian, and Japanese—in versity Press. tlie form of texts and articles or paper-s in period- Rikhter, G. icals. A more complete bibliography of this liter- August 1954. Snow Cover, its Formation and Prop- erties. SIPRE Translation No. 6. ature is available on request from the Forest Roch, Andre Supervisor, Wasatch National Forest, Salt Lake April 1956. Mechanism of Avalanche Release. SIPRE City, Utah. Translation No. 52. The classic work in Englisli has long been Seligman, G. "Snow Structure and Ski Fields" by Gerald 1936. Snow Structure and Ski Fields; With an Ap- pendix on Alpine Weather by C. K. M. Douglas. Seligman. This well-illustrated book has been out London : MacMillan and Co. Ltd., 555 p. of print for a long time, and accessible copies Yosida, Z. are rare. 1955 and 19.56. Physical Studies on Deposited Snow, I. Some of tlie more important Swiss texts have Thermal Properties II. Mechanical Properties. Con- be«n made available in English by the U.S. Army tribution to the Institute of Low Temperature Corps of Engineei-s' Snow, Ice and Permafrost Science. Sappro, Japan: Hokkaido University No. 7 Research Establishment (referred to below as and No. 9. SIPRE). Most important of these is ''Snow and Various authors 1951. Review of the Properties of Snow and lee. its Metamorphism," a standard reference on snow SIPRE Report No. 4. mechanics published in Switzerland in 1939. For additional papers and texts on the general Bader, H., R. Haefeli, E. Bûcher, et al. January 1954. Snow and its Metamorphism, SIPRE subject of snow and ice, see List of Publications, Translation No. 14. U.S. Army; Snow, Ice and Permafrost Research Bueher, E. Establishment ; Corps of Engineers, July 1959. February 19.")6. Contributions to the Theoretical Foundations of Avalanche Defense Construction, The technical journal which carries papers on SIPRE Translation No. 18. various aspects of snow physics and occasionally Fuchs, Alfred on avalanches is the Journal of Glaciology, pub- July 1955. Avalanche Conditions and Avalanche Re- search in the United States with Recommendations lished twice a year by the British Glaciological for Future Work, SIPRE Technical Report No. 29. Society.

84 U.S. GOVERNMENT PRINTING OFFICE: 1961 O—577868