AND AVALANCHE ACTIVITY

Allnals of Glaciology 3 989 @ nternational Glaciological Society NFLUENCE OF STRONG NDS ON SNO DSTRBUTON AND AVALANCE ACTVTY by R. Meister (Eidgenossisches nstitut fur Schnee- und Lawinenforschung, eissfluhjoch, C- 726 Davos, Switzerland) ABSTRACT n a local range, crest winds were copared with winds at lower stations ake it possible initiate a drift-transport odel which would predict snow accuulation patterns on leeward slopes. Corrections the odel input were ade after consideration of detailed drift-flux easureents in the lowest 2 above snow surface. Good agreeent was found between the tal length of large avalanches in a path near the crest, the appropriate wind reading and the corrected snow-depth increents in the rupture zone. Control of ediu-sized avalanches likely cause injury skiers can be iproved with the proposed ethod. NTRODUCTON north, at the first pre-alpine ridge crest, the ean daily wind speed exceeds a 5 / s threshold value during 6% of all days of the year. This is docuented by readings of the station at santis (252 a.s..). At eissfluhjoch (269 a.s..), this threshold is reached on only % of all days. The Alpine wind field is influenced in a arked way by local pography and altitude. At Testa Grigia near Zeratt, for exaple, at an altitude of 348, the ean hourly wind speed exceeds 2 / s for h of the year. Local wind analysis for Davos Four sites were se lected for oniring of wind direction and represents a typical ountain situation drained the south-west and creeks slopes structuring the terrain and potential avalanche starting areas. the siultaneous speed. Figure with a ain valley with different side hence deterining Transport of snow under the influence of wind occurs either siultaneously with precptation or, during dry conditions, with pure blowing snow. Caused by the rough terrain, strong winds in Alpine regions show a arked eddy turbulence. Deposition patterns are predoinantly deterined by this coplex flow. Avalanche activity has been directly related strong winds in ore than 4% of all events recorded in the Davos area (de Quervain and Meister, 987). eather conditions glvg rise blowing snow are coon in ountainous regions, and skiers are aware of their influence when they are following tracks drawn shortly before in an undisturbed snow field. Drift-flux easureents have been ade not only in Antarctic regions (Budd and others, 966) but also in nuerous other areas of flat terrain associated with drift probles. Schidt (986) analysed transport rates of drifting snow and pointed out their relationship aerodynaic roughness length. Mountain observations (Fiihn, 98; Meister, 987) are coplicated by the bulged vertical wind-speed profile at crest lines and by sall obstacles just before the traps. Snow-depth distribution could easily be easured by aerial phograetric techniques, but these depend on there being low snow reflectivity (contrast) which is rather rare. Microwave reote-sensing techniques onir avalanche-critical snow paraeters have been presented by Matzler (987), but are still not in practical use. Fiihn and Meister (982) showed how indirect snow-cover data ay influence calculated avalanche run-out distances. The ost iportant steps in this procedure are () deterination of the snow-pack increent layer caused by snowfalls in the area, (2) evaluation of the axiu possible snow-slab height by the Coulob-Mohr criterion, and (3) estiation of the ost likely snow-slab area. n the present paper the first and last steps are exained in relation ountain winds. MOUNTAN ND OBSER V A nons The Davos area represents an inner Alpine ountainous region with rather reduced wind speeds. Further the Fig.. Local ap of the Davos area, with urban region cross-hatched, showing lake in the valley bot, and position of four sites (J, VF, MA and DA) used for local wind analysis. Qualitative wind roses show the ain wind directions. 95

Meister Strong winds and avalanche activity SCATTERGRAM ND DAVOS 987/88 36 -'-ii-----i--------"'-ii--"-i-----f-- t t- + + ----,------r-----t---------,_-----r-----t---' + Cl 27 ----------------------------------i----- "U U Cl Z + --+-----------i---+--t--+-----+----- *, '. +.fill '. + ++ + + t ",. + ++ +.+!+ + t -; + le --+-,------r-----t-; + -+-t---r-----r---+i.-'.;+.;. 4o + + +... t ----_r-----r-----t----------_r-----r-----t-----' +,+ 9 ---------r-----t---------_r-----r-----t-----.. ------------f----i-+-------.r.. +++++ -. t!+.+.,;.t. ++J;,... t -L L J Sl Sl Sl Sl l\ N NDDRECTON J (deg) Fig. 2. Coparison of free-field wind direction (J) and wind direction in valley bot (DA). 435 x hourly readings selected fro data collected between January and March 988. Sl D (T) Fig. 3. Test area around "Versuchsfeld" (VF) near Davos. P-P5 are positions of five cup aneoeters, S is snow-depth stick; Pluv. pluviograph. Distribution of snow depth on 3 March 988 is qualitatively represented, showing areas with less snow than at S (shaded parallel the east direction) and areas with ore snow than at S (shaded parallel the north direction). The instruentation used for the winter readings in 987-88 was as follows J "eissfluhjoch" crest positon at 2693 near the nstitute. A Pit gauge with rotatable flag, both ounted on a ast, easured peaks of 2 s gusts and in eans. VF "Versuchsfeld" (study site) plateau station at 2544 in a sall valley. A set of five aneoeters, P-P5, installed 2 above snow surface easuring the wind-speed conditions around a sall hill. P5, at the hillp, also easured wind direction. The readings were taken during the sae sequences as J. Daily conventional snow-cover observations were also at the user's disposal. MA "Mattenwald" sheltered forest site at 98. Fro a cup aneoeter in ean wind signals were transitted by radio over a distance of 6 k the nstitute. This wind station onired rupture-zone conditions of a gully avalanche path of 28 steepness. DA "Davos" valley station at 56. Operated fro an open field near Davos lake and undisturbed by buildings, the eteorological centre offers the sae dataset as at crest position. Crest readings obtained using the instruents described above yield the ain wind direction and the doinant velocity features for winter conditions. Monthly ean values of 4-7 / s are typical. As shown qualitatively in Figure, where secral wind directions and ean wind speeds were plotted, the local pography deterines both direction and speed of wind at lower stations. A clear correlation could be found for wind direction between the peak station (J) and the valley plane (DA), as shown in Figure 2. The 435 hourly ean wind directions chosen randoly fro 26 periods between 4 January and March 988 indicate a arked diversion of the north-west winds at the crest the north-east in the valley ground. The other doinant wind direction is south-east at crest position, and winds blowing fro that direction were frequently pressed close the ain axes of the valley, although about 25% of all observations did not confor this pattern. ith forestsite observations no specific dependence of wind direction relative the other stations could be found. ind speeds at the two lower stations differ considerably fro those observed at the crest. This is indicated by the low correlation coefficients shown in Table. Only a few situations have been found in which north-east wind gusts penetrate the lightly covered larch site (MA) just below the treeline; soe 3 the south-east three open channels are regularly filled with windtransported snow. This indicates strong wind influence near that site and, fro wind readings, gives warning of places unsuitable for walkers and skiers. A second proble, which had also been noticed in previous winters, arises fro the different tie spans of the snow srs, which had different onsets and different intensities at the range of sites studied. TABLE. ND-SPEED OBSERVATONS N TE DAVOS AREA (435 h MEANS, RANDOMLY SELECTED FROM TE PEROD 4 JANUARY 988 TO MARC 988). FOR LOCAL SCALE STUATON SEE FGURE Site ind speed ean Std. dev. Max. a b speed r2 (/ s) (/ s) (/ s) J YF MA DA 7.94 = x 3.24 4.69 = Yl 2.56.89 Y2.63 2.5 = Ys.85 8.5 2.2.75 -. 7..9.7 8.6.26.46.85.46.45 a,b = linear regression coefficients, r2 Yi = ax + b, i =, 2, 3. correlation coefficient, 96

Meister Strong winds alld avalanche activity n winter, uch ore than in suertie when there is relatively well-undersod convectional behaviour, the valley winds are deterined by barrage pressure differences, frontal air streas, precipitation patterns, and air stratifications, which gether with the influence of the coplex Alpine terrain lead coplex wind-field patterns. Microscale wind field An attept was ade in winter 987-88 find flow patterns around a sall hill, with h = 5-, just beside YF and soe 5 below the nstitute. Figure 3 shows the positions of the five cup aneoeters (P-PS) relative the hillp. The advantage of the nearby experiental plot and the sheltered huts ust be ephasized. Because of the risk of disturbance fro ski pistes, P3 could not be oved further the east and P was influenced, at least at the beginning of the winter, by the presence of the elongated hut. All aneoeters were ounted approxiately 2 above the snow-cover surface and therefore their arrangeent required periodical adjustent. One wind flag only was ounted at PS, but for the days analyzed here this disadvantage was not liiting because the in eans at the index wind positon PS showed representative behaviour. For gust analyses including turbulence spectra it would be of interest be able identify all five wind directions. Our observations show that the incoing air flow is deflected both sides of the hill and that the effects of flow separation were ore pronounced under neutral lapse rates with south-east winds than with winds fro other directions, whilst shed vortices also occurred in stabledensity stratification. n low wind conditions of <2 / s the air strea tended flow around the side of the hill and in these conditions PS easured the lowest ean wind speed of all. The scattergra for wind direction presented in Figure 4 indicates two ain secrs which are ost affected by the valley pography. The readings presented were randoly selected fro data for the period fro January March 988. Splitting the data set with respect the two ain wind secrs, soe basic results can be derived for ean values and statistical indices in Table. Speed-up on the line P-P5-P4 is practically the sae for each of the two ain wind directions and P2-P3 is perpendicular these directions. Taken as a whole, the ean ratios of the in readings reain ore-or-ess constant in the range of -3 altitude around gentle hills. ND VF 987/88 SCATTERGRAM ---,------r-----t-----'-----,------r-----t-----' 25!. ---,------r-----t-----'-----,------r-----t-----' 2 ""E.. ---,------r-----t-----'-----,------r-----t-----' 5 rl + + + 4 +.',,+ + + -----------.,--' +. ----- --- a.. Ul '.+. ---,-----+ +." t -,--.----,-. + DRECTON PS Fig. 4. Scattergra readings at PS randoly selected and March 988. ----- + t.... 5 "'+..... -----, ++., (deg) of wind-direction and wind-speed (YF) with 4 x in readings fro data collected between January DRFT TRANSPORT The distribution of wind-blown snow in an air strea depends essentially on wind speed, snow-surface conditions, grain-size, air teperature, and huidity. Detailed theories on drifting separate tal drift rate in bed load and suspended load transport, the first containing the greater proportion of the transported ass (Schidt, 986). Applying the sae principles as are used for river suspended sedient transport (Kennedy, 988) or sand transport by winds (Bagnold, 956), the snow flux per unit width of strea ay be calculated using Zu Qs = f lzuz dz () zl where l z is the drift density at height z above surface and U z is the corresponding horizontal wind speed. The one-diensional diffusion equation yields (2) The index r is a reference height, u. is the the friction velocity, and k =.4 (von Karan constant). These equations, gether with a known profile of the vertical wind velocity coponent, would lead in a direct anner the tal transport. Probles arise with the lower integration height zl, which does not vanish near the ground for a logarithical wind profile, and also because vs' the ean fall velocity of suspended particles, is unknown and ust be estiated. Near the bed, ass transport occurs by creeping, rolling and saltating. The quasi-hydrostatic lift perpendicular the ean velocity direction causes a deforation of the strea lines just above the bed layer. Sall vortices are oved in a sinusoidal anner and see create conditions which then allow displaceent of single particles. The greatest iportance ust be attached the utual contact areas between neighbouring grains, since the snow-surface conditions for drift are deterined by those grains which can be oved freely without previously needing displace their neighbours. A fresh attept was ade during winter 987-88 clarify the situation which arises during Alpine snowsrs, in particular with respect the significance of precipitation. During dry conditions, which at the chosen test field near eissfluhjoch occur ostly with south-east winds, drift ay be reduced bed-load transport, suspended particles higher than.5 above snow surface being rather rare. Coparison with the situation at an Alpine ridge crest (Schidt and others, 984) showed a bulged vertical ass-flux profile, with a axiu O.S-.S above the snow surface. Soe doubts were raised about the nature of the drift density in the.2 region of the drift closest the ground. n order especially exaine these lower levels, the whole installation with asts and equipent was oved YF eliinate the influence of flow separation at the sharp crest. P2 (Fig. 3) offers optial conditions for the oniring of all wind directions and speeds. n addition the five freely turnable Melior rocket tubes (inlet orifice F = 2 ) three glass funnels (F = 26 2 ) were ounted on a sall ast in order ove the down the level. A filter-fabric sock (F = 24 2 ) copleted the equipent. The results of 2 runs are presented in Table ll. Precipitation rate was estiated with the help of a heated precpitation gauge (Joss-Tognini type SM), a pluviograph and of an electronic snow-depth gauge at the nearby study site. Surface-snow density readings were perfored before or during the runs on the windward side, but always after the runs on the lee side. Drift transport, Q, in the lowest 5 above the terrain was estiated by the use of the Equations () and (2) for each run with stepwise profile integration, the lower liit zl chosen as Zo = roughness length, and the upper liit Zu chosen as S.O. The ost interesting observations ade during the exposure ties were No. 4 and 5 these sr events occurred with clear sky, 97

Meister Strong winds and avalanche activity TABLE. ND OBSER V A TONS AROUND A SMALL LL AT VERSUCSFELD (VF), DA VOS, 254 a.s.l. (4 RANDOMLY SELECTED Q in MEANS RECORDED FROM JANUARY TO MARC 988). FOR TOPOGRAPCAL STUATON SEE FGURE 3 File "E" (winds fro secr 45-24, nuber of observations 768) Site ind speed Std. dev. Max. a b r2 ean (/s) (/s) (/s) P 3.6.77.3.69.37.97 P2 4.69 2.48 3.8.97.6.98 P3 4.27 2.28.4.89.3.97 P4 4.58 2.35..88.45.9 P5 4.68 2.53 3.3 DD P5 ( 3 ) (26 ) File "" (winds fro secrs -44 and 25-36, nuber of observations 642) Site ind speed Std. dev. Max. a b r2 ean (/s) (/s) (/ s) P 3.54 2.8..68.57.92 P2 4.24 2.72 2..9.29.96 P3 4.2 2.72 2.2.9.22.98 P4 3.88 2.66.4.88..98 P5 4.39 2.96 2.9 DD P5 (286 ) (29 ) a,b linear regression coefficients, r2 = correlation coefficient, y = ax + b with x = speed P5 and y = speed Pi with i =... 4. TABLE ll. SNODRFT MEASUREMENTS AT VERSUCSFELD (VF) STE P2, NTER 987-88 Run no. Date Start Dura- ind Air Precip- Surface snow Lowest Dr'ft Rearks tie tion DD FF tep. itation density gauge transport (P5) (P2) rate windward lee position Q ( in) (deg) (/s) (C) (/h) (kg/ s ) (kg/ s ) () (g/ s) 2 3 4 5 6 7 8 9 2 3 4 5 6 7 8 9 2 5 Jan 88.3 35 28 6. -2. 9.3 8. ski tracks (depth =. ) filled within 3 in 5 Jan 88 5.3 9 9 5. -. 9. 5. 5 gusts with u> 2 / s 5 Jan 88 6.3 937 38 5.4-4. 9.3 22.4.2 snow eroded during run 6 Jan 88 8.5 6 3 2.5-5. 6.3 25.4 ski tracks (depth =.5 ) filled within 2 in (visibility reduced) 6 Jan 88.4 3 34 6.2-3. 39. 4.9 7 Jan 88 6.5 96 32 5.4-4.2 73.3.9 constant wind direction 26 Jan 88. 356 2 4.9-5. 84 37..4 Graupel 3 Jan 88.5 5 288 6.8 -. 8 55. 5.7 Sastrugi with ean height =.7 3 Jan 88 3. 64 28 5.8 -.9 3 24.2 32. surface patterns filled within in Feb 88 4.3 2 39.4-5. 5 84. 58.4 ski tracks (depth =.4 ) filled within 5 in Feb 88 5.3 42 7.6-6. 78 7.7 25.9 3 Feb 88 8.3 3 6 6.8-8. 4 33. 6.7 no constant wind direction 5 Feb 88 8.3 32 38 5.7-9. 24 238..5 ski tracks (depth =.7 ) partly filled after run 6 Feb 88 6. 33 22 6.2-6. 29. 62.3-3 gust per in 7 Feb 88 9.4 69 3 8.3-9 2.5 88 7. 67.3 no sastrugi, little turbulence 9 Feb 88 5.3 29 259 7. -2 2. 44 82. 38.4 fluctuating wind direction 29 Feb 88 8.5 5 28 3.6-7.4 88 33.5 5.8 soothed surface 29 Feb 88.2 4 289 6.5-8.6 24 39.5 4.4 strong erosion, waved deposition patterns 8 Mar 88 8.5 49 274 6.6-8. 92 2. 5.5 Mar 88 9.3 6 275 6. -2.5 7. 23.8 fluctuating wind direction 98

E N l! E N l! Feb 988 RUN NR. ( tn) - -2 l a PRECPTATON RATE= - -F_ - 7 Feb 988 RUN NR. 5 ( 69 tn) U,.- 7.6 /s AZM 42 deg TEMP- -6 QC. /h M MELLOR'S F FUNNEL S SOCK a 25.9 9 -- s... F, M DRFT DENSTY n z (g/3) PRECPTATON \ " u,.- "- M r 'M " 8.3 /s AZM 3 deg TEMP- -9 QC RATE= 2.5 /h ELLOR'S F FUNNEL S SOCK - \r a- 67.3 9-2 '" i'f _c;,."'" '" -- s Meister Strollg willds alld avalanche activity No. 3 hardened snow surface, snow transport only in 25 gusts with speeds of ore than 2 / s, grain-type distribution ratio 8 at the level of.. No. 5 Fluffy, weak, new snow was easily erodable. nitiall y erosion was.2 at the ast but later sall sastrugi were filled and terrain becae soother again. Grain-type distribution at.5 = 8 2 ; at.38 = 5 5 ; at.2 = 5 4 (see Fig. 5b). No. 8 Erosion with hard sastrugi on windward side and greater than 5 on the lee side; deposition up. Figure 5 presents outstanding differences in density profiles in relation free-field precipitation rate. Pure drifting occurs at wind speeds of about 8 / s only in the first.5 above snow surface, but with precipitation there is an asyptic approach precipitation rate (Fahn, 98). Assuing a ean fall velocity of / s, the precipitation rate of 2.5 / h or 25 g/ 2 h is equivalent an undistu rbed drift flux of.7 g/ 2 s at ground level. Subtracting the drift transport originating fro freefield precipitation (Q ) fro the easured drift-transport rate (Qs) gives an pridication of the additional snowtransport rate, which ay also be verified by the snow redistribution around the hill. A representative equation for drift-transport rates has been used, and in Figure 6 all 2 drift-transport rates, Qre.t' obtained fro the readings of VF tests during the winter 987-88 have been plotted, gether with data fro the 75 runs collected in previous winters (978-79 984-85) at the crest position, and then copared with ean wind speed at a reference height chosen as 2. in order facilitate coparison with the two data sets. For crest readings, this coparison iplied a bulged profile type adaptation. The large scattering cannot be explained by inaccuracies in precipitation easureents alone, but is ore likely have resulted fro snow surface conditions such as snow density or sintering behaviour. SNO DEPOSTON AND REDSTRBUTON e ake th e assuption that all lifted aterial is deposited within a liited distance behind the crest or behind obstacles such as buildings or blocks. The ost iportant facr is the acceleration rate near the ground, represented by (3) -3-2 b - DRFT DENSTY n z (g/3) Fig. 5. Snow-drift densities, nz, as a function of height, z, above snow surface (a) south-east winds without precipitation, (b) north-west winds with ean precipitation rate of 2.5 /h. and erosion was as pronounced as at the test site (loss of.5 of tal snow depth between 8. and 2. h of 6 January 988). No. 7 preceded by a short shower with graupel 2-3 in size; this is rare in wintertie. No. 8 ost snow transport ok place during 2 gusts with wind speed greater than 5 / s, flux was only oderate between gusts. No. 9 estiated ass ratios of crystal types (new large crystals felt-like grains rounded grains) was 3 at the.2 level and 2 at the.3 level. No. at the.5 level sall rounded fragents approxiately.2 in diaeter were doinant. Sooth surface and erosion fro weak, unsettled, newly fallen snow. No. lower wind speed and hard windward snow produced less snow transport than previous run, but bed load in lowest.8 was higher (see Fig. 5a). No. 2 turbulence in wind direction was very pronounced; unstable conditions diinished trap efficiency. 3 F -" r- f- le f- > 2 E +' " ) L- a w a rr rr a. '3 Ul z a rr - a(x) du(x)/ dt 2 3! ;f 2 2. t ili. a 3 El 5 5 MEAN NDSPEED U,. (/s) (4) 2 25 Fig. 6. Drift transport rate, Qre't' in lowest 5 above snow surface as a function of ean reference wind speed, U r. Nubers indicate integers of corresponding precipitation rates (/ h). 99

Meister Strollg willds and avalallche activity with u(x) adapted fro a potential flow odel (Fohn and Meister, 983). n a drift flow, erosion or resuspension occurs when a(x) < and deposition when a(x) > O. Threshold wind speed for transport of snow is deterined priarily by the degree of cohesive bonding (Schidt, 98). All aterial pass ing the crest line as additional load, Qrest' will be deposited at a distance which will be deterined by terrain roughness, aspect, and wind velocity. elzenbach (93) has di stingui shed two cornice types in relation speed and kineatic viscosity of the air. e has entioned a press ure-influ e nced type, in which grains adhere at the windward site, and a rolling type in which particl es deposited in a static sec tion are afterwards roll ed gether as an enseble. t is th e second type of cornice or snow accuulation that is of interest us. Around sall two-diensional obstacles, Tabler (975) was able obtain good results with a stepwise calculation of the developent of an equilibriu snow profile. This odel has bee n tes ted in Alpine terrain, but has the proble that one of th e os t iportant unknown paraeters is the se ttleent of the relatively loose wind-blown snow. Over sharp-edg ed borders along roadsides one can observe the typical depo si tion features which adjust the strealine direction during the drift process. The upper borde r of the ixing zone le ngthen s the forward-lying te rrain; sall crevasses can be fill ed in this anner, but with th e low viscosity of th e deposi ted snow the snow ass se ttles and strealine equilibrati on u st begin again in the next sr. At the beginning of winter, sall terrain irregularities are filled during the first srs and for these al so a settleent effect has be taken in account. ind-hardened slabs usuall y have high viscosities (Meister, 985), so that loose la ye rs below will se ttle ore easily due radiati o n effects. This is a situation which can lead additional stre sses. Surveying snow dep th around ridge slopes (Fohn and Meister, 983) has helped ak e it poss ible easure the tal snow transport in windy periods. The relevant equation is (5) was in the range of -2. Snow-depth distributions at VF for 3 March 988 have been plotted on Figure 3. Unshaded areas indicate regions with a snow depth of 2-3 ; the reference level (S VF) was 2.94 for that day. Areas with less snow (-2 and - ) are shaded parallel the abscissa and areas with ore snow shaded (2-3, 3-4, 4-5 ) are at right-angles it. Although the ean wind vecr aounts only.2 / s, accuulation areas are situated in the south-east of the hill and erosion zones lie the west. t is deduced that the greatest catch of drifting snow ust occur in periods with precipitation. The data relating deposition zones point flow separation by shed vortices. The question arises whether there are soe doinant terrain zones on which snow accuulation happens in a siilar anner each winter. The tal input ass of snow plays an iportant role; it can range fro 4 5 of water equivalent at an altitude of 25. ind regies al so vary fro winter winter. Sall troughs are usually filled during the first sr as are lee slope and cornice accuulations above about 2. This can be observed in springtie with ablation features. A V ALANCE ACTVTY The influence of snow transport on avalanche activity has be considered fro two different aspects (a) teperature-gradient snow etaorphis depends ainly on tal snow depth, and thus it differs between places with heaped snow and those with swept hups; (b) surplus stresses in rupture zones are due ainly wind-transported snow asses. The ost critical situations for windde pendent avalanche occurrences are during and in the 24 h following srs. The build-up phase of accuulation in cobination with precipitation and drift is of ajor iportance. A dangerous track in the Davos area is Salezer Tobel, situated just west of Davos lake (Fig. ). Using data fro the ten ost iportant events in the past years at that path, which are presented in Table V and Figure 7, the soothing effects were estiated by Equation (6), in relation avalanche length and snow-depth increents (L'l ) in the adjacent precipitation period, after taking account of the ass (L'M) of the wind-transported snow. Avalanche width and gully effect were also paraeterized. The readings at the valley station (L'l DA) ake no sense. The 979-8 avalanche event occurred- in spring, when the whole snow-pack was weakened because of its oisture. The unusual avalanche at 28 March 988 was triggered by explosives after evacuating people in exposed buildings. According the snow-depth increents, the rupture height aounted ore than.5 but the avalanche, which did not start in the upperost zone, had a relatively short tal length because of wet snow in the run-out zone. The question arises as whether that sae avalanche would naturally have started soe hours later. which showed that the snow ass transported the leeside in a given tie equals half the difference between the increent of water equivalent (in ) on the lee slope (L'l l) and the windward slope (L'l w). The equation holds only if the ridge syste as a whole catches the sae tal ass as a reference plot in the sae vicinity. Earlier surveys have helped investigars explore the additional deposition the leeside of alpine ridges. Additional deposition is equivalent drift transport, Qrest' and deposition length depends on slope angle, wind speed and drift density. Dyunin and Kotlyakov (98) have presented a snow-sr loading and tal solid discharge of drifting snow which, according their observations, grows with the third power of wind speed. ith a refineent related precipitation rate, Pr ( / h), ean wind speed at reference height U r (/s), and threshold wind speed vtr ( /s ), we get a ass of wind-transported snow L'M lee slopes of CONCLUSONS Threshold wind speed, vtr ' is found be 5 / so The reliability of this equation was tested with winter readings fro 978-79 98-82; the best agreeent was found after splitting wind direction in two ain secrs (north-west and south-east), and obtaining ean wind speeds for tj = 8 h. This was found by power-spectru analysis give a first-order peak in energy exchange (Meister, 987). For winter 987-88, Equation (6) underestiated additional snow load at the leeward point, P4, shown in Figure 3 by a facr of 2.8. At that site, situated soe 4 the lee of the sall hill, a tal water-equivalent of 52 accuulated in the period up 3 March 988. Starting with the onset of snow on Noveber 987, the experiental plot (VF) showed 824 ; calculation using Equation (6) gives a tal for wind-blown additional surplus snow of 6 for eastward facing slopes; ean depletion distance of wind-transported snow at ridge crests These studies have been exclusively concerned with the Davos area, yet for practical purposes conclusions ay be drawn which apply equally for different ountain regions. ith the observed large differences of ridge crest and valley winds on a local scale, no application of diagnostic odelling as reported elswhere (Tesche, in press) can at the oent be recoended for Alpine pography. More eteorological analyses are necessary. Drift transport sets in at a threshold wind speed of 4-5 /s, but depends on turbulent fluctuation in the near-surface air and on the density of the upperost snow layer. Density profiles of wind-transported snow in the lowest 5 above a level snow surface show that ost load is ebedded in the lowest.5 above terrain. ith precipitation, the density adjusts the free-field rate above 5, but vanishes above -2 without precptation. Deposition patterns of snow are predoinantly conditioned by terrain irregularities. The settleent of the tal snow depth prevents an adjustent equilibriu conditions. To iprove rupture-height evalution in potential avalanche zones, a siple odel 2

Meiste,. SLrong winds and avalanche activity TABLE V. SALEZERTOBEL, DAVOS MOST MPORTANT AVALANCE EVENTS N TE NTERS 978-79 987-88 Date Adjacent Length idth Spping s' Trigger ls VF l VF l DA lm - period L B level ethod (d) () () ( a.s.l.) () () () () () 3 Mar 79 3 8 2 68 natural.75 86 6 7 9 Apr 8 2 2 5 65 natural. 23 7 2 2 lan 8 6 7? 58 75 natural.69 4 35 Dec 8 4 6? 57 5 natural.9 8 48 25 6 Jan 83 3 2 2 55 25 natural.45 76 59 76 9 Feb 84 3 22 6 55 475 natural.88 8 52 45 3 Feb 85 3 95 5 66 natural.3 8 4 6 8 Mar 86 2 2 4 68 natural.3 5 4 3 Mar 87 5 2 5 64 natural.59 98 69 9 28 Mar 88 5 4 25 58 75 explosive.82 49 7 5 Adjacent period = length of previous precipitation period. s' = avalanche run-out distance; data fro F()hn and Meister (982). e 3 < U - Z s w 2ee 5 lee * - VF + o - VF - - DA t U t* t v, (') ('),... N "-*lt "; er u z '' vo "- - - r - l- Q. 5 - w - C5-3 Z (/) CD - 't,t "- u, o o 5 5 2 25 OBSERVED AVALANCE LENGT L () Fig. 7. Salezer Tobel avalanche path. Total avalanche length, L, of axiu events for winters 978-79 987-88 shown in relation snow-depth increents ( ) of the adjacent precipitation periods. l J)A is accuulated water equivalent of new snow at the site Davos (56 ); l_yf is accuulated water equivalent of new snow at VF (254 ); lm is accuulated additional drift of new snow as calculated using Equation (6). brings additional snow loads leeward areas. ind speed, divided in three 8 h periods per day, ain wind secrs, and precipitation rates refine the odel input. For the purposes of iproving forecasting ethods valley readings of wind speed and snow depths are unsuitable data, but the best results were gained with crest wind easureents and snow-depth readings in sheltered places near the avalanche rupture zones. ACKNOLEDGEMENTS This work was part of project No. at FSAR, sub-titled "Snow transport". The author wishes thank all ebers of the nstitute for helpful support either in field capaigns or in evaluation work. Special thanks go Direcr laccard for revising the text, P. eilenann who provided coputer software, and E. ug for secretarial work. REFERENCES Bagnold, R.A. 956. The flow of cohesion less grain in fluids. Philos. Trans. R. Soc. London, Ser. A, 249, 235-297. Budd,.F.,.R.. Dingle and U. Radok. 966. The Byrd Snow Drift Project outline and basic results. n Rubin, M.., ed. Studies in Antarctic eteorology. ashingn, DC, Aerican Geophysical Union, 7-34. (Antarct. Res. Ser. 9.) Dyunin, A.K. and V.M. Kotlyakov. 98. Redistribution of snow in the ountains under the effect of heavy snow-srs. Cold Reg. Sci. Technol., 3(4), 287-294. F()hn, P.M.B. 98. Snow transport over ountain crests. J. Glacial., 26(94), 469-48. F()hn, P.M.B. and R. Meister. 982. Deterination of avalanche agnitude and frequency by direct observation and/or with the aid of indirect snowcover data. Beitr. ildbachersions- und Lawinen-forsch. Mill. Forstl. BUlldes-Ve,.suchsansl. ien, 44, 27-228. ' f()hn, P.M.B. and R. Meister. 983. Distribution of snow drifts on ridge slopes easureents and theoretical approxiations. Ann. Glacial., 4, 52-57. Kennedy,.F. 988. Coputation of river suspendedsedient discharge revisited. Schweizer ingenieur und Architekt, 6(6), 5-53. Meister, R. 985. Density of new snow and its dependence on air teperature and wind. Ziircher Geog,.. Schr., 23. 73-79. Meister, R. 987. ind systes and snow transport in Alpine pography. nternational Association of ydrological Sciences Publicatioll 62 (Syposiu at Davos 986 Avalanche Foratioll. Moveent alld Effects), 265-279. Quervain, M. de and R. Meister. 987. 5 years of snow profiles on the eissfluhjoch and relations the surrounding avalanche actvty (9336/ 37-985/ 86). nternational Association of ydrological Sciences Publication 62 (Syposiu at Davos 986 - Avalanche Foralion. Moveent and Ef/ects), 6-8. Schidt, R.A. 98. Threshold wind-speeds and elastic ipact in snow transport.. Glacial., 26(94), 453-467. Schidt, R.A. 986. Transport rate of drifting snow and the ean wind speed profile. Boundary-Laye,. Met eo,.ol., 34(3), 23-24. Schidt, R.A., R. Meister and. Gubler. 984. Coparison of snow drifting easureents at an Alpine ridge crest. Cold Reg. Sci. Technol., 9(2), 3-4. Tabler, R.D. 975. Predicting profiles of snowdrifts in pographic catchents. Proc. est. Snow Call/., 43, 87-97. Tesche, T.. n press. Nuerical siulation of snow transport, deposition and redistribution. p,.oc. est. SnolV Coni. elzenbach,. 93. Untersuchungen iiber die Stratigraphie der Schneeablagerungen und die Mechanik der Schneebewegungen nebst Schlussfolgerungen auf die Methode der Verbauung. iss. Vera//. Dtsch. Osterr. Alpenvereins, 9. 2