EGU Journal Logos (RGB) O O O Advances in pe Annales pe Nonlinear Processes pe n n n A A A Geosciencescc Geophysicaecc in Geophysicscc e e e s s s s s s Natural Hazards O Natural Hazards O p p e e and Earth System n A and Earth System n A Sciencesccess Sciencesccess Discussions Atmospheric O Atmospheric O p p e e Chemistryn A Chemistryn A and Physicsccess and Physicsccess Discussions Atmospheric O Atmospheric O p p e e Measurementn A Measurementn A c c c c Techniqueses Techniqueses s s Discussions O O p p Biogeosciencesen A Biogeosciencesen A cc Discussionscc e e s s s s Climate Ope Climate Ope n n A of the Past A of the Pastccess Discussionsccess Earth System Ope Earth System Ope n n A Dynamics A Dynamicsccess Discussionsccess Geoscientific Geoscientific O O p p Instrumentation e Instrumentation e n n A A Methods andcce Methods andcce Data Systemsss Data Systemsss Discussions GeoscientificOpe GeoscientificOpe n n A Model Development A Model Developmentccess Discussionsccess Hydrol.EarthSyst.Sci.Discuss.,10H,9y5d05ro–9lo53g1y, 2a0n1d3 O Hydrology and O D wd©owAi:w1u.0thh.y5od1rr(9os4)l-/e2ha0er1st3sh.d-sC-1yCs0t--A9s5tct0ri-i5bd-uis2tci0ou1ns3s3..n0eLtE/i1ca0e/rn9ts5hSe0 .c5S/i2ey0sn1tc3e/emspen Access EarthS cSieysntceemspen Access iscussio 10,95H0E5–S95S31D,2013 Discussions n P Op Op a TShciisendcisecsu(sHsEioSnSp)a.pPeleraiss/eharesOfebcreeteoantnhu enSdcceoirerrrneecsvepieoen AccewndfoinrgthfienajolupranpaelrHiyndHroEloOSgScyeiafanandv aSEiDlacaisribcetulhesns.Siconeyssen Accetem per Igmlipdainctgoofnsnsooiwl ss ss | redistribution O O D Impact of snowSoglid lEiadrthipen Acceng on soil SolidDi sEcuassriotnhspen Acce iscuss K.Meusburgeretal. ss ss io n redistribution for a sub-alpine area in P TitlePage O O a Switzerland The Cryospherepen Acc The CryoDsispchusesiornespen Acc per Abstract Introduction e e ss ss | Conclusions References K. Meusburger1, G. Leitinger2, L. Mabit1, M. H. Mueller1, and C. Alewell1 D is c Tables Figures 1EnvironmentalGeosciences,UniversityofBasel,Basel,Switzerland us s 2InstituteofEcology,UniversityofInnsbruck,Innsbruck,Austria io J I n P Received:15May2013–Accepted:17June2013–Published:18July2013 ap J I e Correspondenceto:K.Meusburger([email protected]) r Back Close | PublishedbyCopernicusPublicationsonbehalfoftheEuropeanGeosciencesUnion. D FullScreen/Esc is c u s s Printer-friendlyVersion io n P InteractiveDiscussion a p e r 9505 | Abstract D is c HESSD u Theaimofthisstudyistoassesstheimportanceofsnowglidingassoilerosionagent s s for four different land use/land cover types in a sub-alpine area in Switzerland. The 14 io 10,9505–9531,2013 n investigated sites are located close to the valley bottom at approximately 1500ma.s.l., P a 5 while theelevation of the surroundingmountain ranges is about2500ma.s.l. Weused pe Impact of snow two different approaches to estimate soil erosion rates: the fallout radionuclide 137Cs r gliding on soil andthe RevisedUniversalSoilLoss Equation(RUSLE).TheRUSLE modelissuitable | redistribution to estimate soil loss by water erosion, while the 137Cs method integrates soil loss due D is to all erosion agents involved. Thus, we hypothesise that the soil erosion rates deter- c K.Meusburgeretal. u 137 s 10 mined with the Cs method are h13ig7her and that the observed discrepancy between sio theerosionrateofRUSLEandthe Csmethodisrelatedtosnowgliding.Cumulative n snow glide distance was measured for the sites in the winter 2009/2010 and modelled Pa TitlePage p for the surrounding area with the Spatial Snow Glide Model (SSGM). Measured snow e r Abstract Introduction glide distance range from 0 to 189cm with lower values for the north exposed slopes. | We observed a reduction of snow glide distance with increasing surface roughness of Conclusions References 15 D thevegetation,whichisanimportantinformationwithrespecttoconservationplanning is and expected land use changes in the Alps. Our hypothesis was confirmed, the differ- cu Tables Figures s ence of RUSLE and 137Cs erosion rates was correlated to the measured snow glide s distance (R2=0.73; p<0.005). A high difference (lower proportion of water erosion ion J I P 20 compared to total net erosion) was observed for high snow glide rates and vice versa. ap J I The SSGM reproduced the relative difference of the measured snow glide values be- er tween different land use/land cover types. The resulting map highlights the relevance Back Close | ofsnowglidingforlargepartsoftheinvestigatedarea.Basedontheseresults,wecon- D FullScreen/Esc clude that snow gliding is a key process impacting soil erosion pattern and magnitude is c in sub-alpine areas with similar topographic and climatic conditions. u 25 s s Printer-friendlyVersion io n P InteractiveDiscussion a p e r 9506 | 1 Introduction D is c HESSD u While rainfall is a well-known agent of soil erosion, the erosive forces of snow move- s s mentsarehardlyknown.Particularlywetavalanchescanyieldenormouserosiveforces io 10,9505–9531,2013 n thatareresponsibleformajorsoilloss(Gardner,1983;Ackroyd,1987;Belletal.,1990; P a 5 Jomelli and Bertran, 2001; Heckmann et al., 2005; Fuchs and Keiler, 2008; Freppaz pe Impact of snow r et al., 2010) also in the avalanche release area (Ceaglio et al., 2012). gliding on soil Besides avalanches another important process of snow movement affecting the soil | redistribution surfaceissnowgliding(InderGandandZupancic,1966).Snowglidingistheslow(mm D is to cm per day) downhill motion of a snowpack over the ground surface caused by the c K.Meusburgeretal. u s 10 stress of its own weight (McGraw-Hill and Parker, 2002). Snow gliding predominantly◦ sio occurs on south-east to south-west facing slopes with slope angles between 30–40 n (In der Gand and Zupancic, 1966; Leitinger et al., 2008). Two main factors that control Pa TitlePage p snow glide rates are (i) the wetness of the boundary layer between the snow and soil e r Abstract Introduction cover and (ii) the ground surface roughness determined by the vegetation cover and | rocks (McClung and Clarke, 1987; Newesely et al., 2000). So far, only few studies Conclusions References 15 investigated the effect of snow gliding on soil erosion (Newesely et al., 2000; Leitinger Dis et al., 2008). A major reason for this shortcoming is the difficulty to obtain soil erosion cu Tables Figures s rates caused by snow processes. In steep sub-alpine areas soil erosion records (e.g. s io J I with sediment traps) are restricted to the vegetation period because avalanches and n P 20 snow gliding can irreversibly damage the experimental design (Konz et al., 2012). ap J I e Recentlyfirstphysicallybasedattemptstomodeltheerosiveforceofwetavalanches r weredone(Confortolaetal.,2012).Nosimilarmodelexistsforsnowgliding.However, Back Close | the potential maximum snow glide distance during a targeted period can be modelled D FullScreen/Esc with the empirical spatial snow glide model (SSGM) (Leitinger et al., 2008). The mod- is c elling of this process is important to evaluate the impact of the snow glide process on u 25 s s Printer-friendlyVersion soil erosion at larger scale. io n Soil erosion rates can be obtained by direct quantification of sediment transport in P InteractiveDiscussion a the field, by fallout radionuclides (FRN) based methods (e.g. Mabit et al., 1999; Ben- p e r 9507 | mansour et al., 2013; Meusburger et al., 2013) and by soil erosion models (Nearing D is et al., 1989; Merritt et al., 2003). Since the end of the 1970’s empirical soil erosion c HESSD u s modelssuchastheUniversalSoilLossEquation(USLE;WischmeierandSmith,1965; s io 10,9505–9531,2013 WischmeierandSmith,1978),anditsrefinedversionstheRevisedUSLE(RUSLE;Re- n P 5 nardetal.,1997)andtheModifiedUSLE(MUSLE;Smithetal.,1984),havebeenused a p worldwide to evaluate soil erosion magnitude under various conditions (Kinnell, 2010). e Impact of snow r These well-known models allow the assessment of sheet erosion and rill/inter-rill ero- gliding on soil | sion under moderate topography. However, they do not integrate erosion processes redistribution D associated with wind, mass movement, tillage, channel or gully erosion (Risse et al., is c K.Meusburgeretal. 10 1993;Mabitetal.,2002;Kinnell,2005)andalsosnowimpactduetomovementormelt- u s ing is not considered (Konz et al., 2009). Several models have been tested for steep s io alpinesiteswiththeresultthatRUSLEisreproducedthemagnitudeofsoilerosion,the n relativepatternandtheeffectofthevegetationcovermostplausible(Konzetal.,2010; Pa TitlePage p e Meusburgeretal.,2010b).TheerosionratederivedfromRUSLEcorrespondstowater r Abstract Introduction erosion induced by rainfall and surface runoff and hence in our site to the soil erosion 15 | processes during the summer season without significant influence of snow processes. Conclusions References D In contrast, the translocation of FRN reflects all erosion processes by water, wind is c Tables Figures and snow during summer and winter season and thus is an integrated estimate of the u s s total net soil redistribution rate since the 1950s (the start of the global fallout deposit). io J I 137 134 n Anthropogenicfalloutradionuclides(e.g. Cs, Cs)havebeenusedworldwidesince 20 P a decades to assess the magnitude of soil erosion and sedimentation processes (Mabit p J I e and Bernard, 2007; Mabit et al., 2008; Matisoff and Whiting, 2011). The most well- r Back Close known conservative and validated anthropogenic radioisotope used to investigate soil | 137 redistribution and degradation is Cs (Mabit et al., 2013). D FullScreen/Esc For (sub-)alpine areas the different soil erosion processes captured by RUSLE and is 25 c the 137Cs method result in different erosion rates (Konz et al., 2009; Juretzko, 2010; us s Printer-friendlyVersion Alewell et al., 2013). However, this difference might also be due to several other rea- io n sons such as the error of both approaches, the non-suitability of the RUSLE model for P InteractiveDiscussion a this specific environment and/or the erroneous estimation of the initial fallout of 137Cs. p e r 9508 | In this study, we aim to investigate, whether the observed discrepancy between ero- D sionratesestimatedwithRUSLEandtheonesprovidedbythe137Csmethodcanbeat isc HESSD u least partly attributed to snow gliding processes. Since vegetation cover affects snow ss gliding, four different sub-alpine land use/land cover types were investigated. The sec- ion 10,9505–9531,2013 P 5 ond objective of our research is to assess the relevance of snow gliding processes at a p catchment scale using the Spatial Snow Glide Model (SSGM). e Impact of snow r gliding on soil | redistribution 2 Materials and methods D is c K.Meusburgeretal. u 2.1 Site description s s io n The study sites are located in Central Switzerland (Canton Uri) in the Ursern Valley P TitlePage a 10 (Fig.1).TheelevationoftheW–Eextendedvalleyrangesfrom1400upto2500ma.s.l. p e Themeanannualrainfall,averagedbetween1986and2008,is1516mm.Themeanair r Abstract Introduction ◦ temperature measured at an altitude of about 1480m a.s.l is 3.1 C (MeteoSchweiz). | Conclusions References The valley is snow covered from November to April with a mean annual snowfall of D 443mm in the period 1986 to 2008. Drainage of the basin is usually controlled by isc Tables Figures u 15 snowmelt from May to June. Important contribution to the flow regime takes place dur- ss ingearlyautumnfloods.Thelanduseischaracterisedbyhayfieldsnearthevalleybot- io J I n tom(from1400toapproximately1600ma.s.l.)andpasturingfurtherupslope.Siliceous P a slopedebrisandmorainematerialisdominantatoursites,andformsCambicPodzols p J I e r (Anthric) and Podzols (Anthric) classified after IUSS Working Group (2006). Back Close Of the 14 experimental sites, 9 are located at the south-facing slope and 5 at the | 20 north-facing slope at altitudes between 1450 and 1600ma.s.l. Four different land D FullScreen/Esc is use/cover types with 3–5 replicates each were investigated: hayfields (h), pastures c u (p), pastures with dwarf shrubs (pw), and abandoned grassland covered with Alnus ss Printer-friendlyVersion viridis(A).VegetationofhayfieldsisdominatedbyTrifoliumpratense,Festuca,Thymus ion serpyllum and Agrostis capillaries. For the pastured grassland Glubelaria cordifolia, P InteractiveDiscussion 25 a p Festuca sp. and T. serpyllum dominate. Pastures with dwarf shrubs are dominated by e r 9509 | Callunavullgaris,Vacciniummyrtillus,Festucaviolacea,AgrostiscapillariesandT.ser- D is pyllum. At pasture sites of the south exposed slope, which are stocked from June to c HESSD u s September, cattle trails traverse to the main slope direction. s io 10,9505–9531,2013 n 2.2 Snow glide measurement P a p e Impact of snow We measured cumulative snow glide distances with snow glide shoes for the winter r 5 gliding on soil 2009/2010. The snow glide shoe equipment was similar to the set-up used by In der | redistribution GandandZupancic(1966),Neweselyetal.(2000)andLeitingeretal.(2008).However, D no data logger was used to capture the snow glide rates for specific time intervals isc K.Meusburgeretal. u during the winter. The set-up consisted of a glide shoe and a buried weather-proof s s box with a wire drum. Displacement of the glide shoe causes the drum to unroll the io 10 n wire. The total unrolled distance was measured in spring after snowmelt. To prevent P TitlePage a p entanglement with the vegetation, the steel wire was protected by a flexible plastic e r Abstract Introduction tube. | Conclusions References 2.3 Assessment of soil redistribution pattern based on RUSLE and the 137Cs D is 15 method cu Tables Figures s s For 7 sites, RUSLE and 137Cs based erosion rates were available from Konz io J I n etal.(2009)andforthe6additionalsitesweappliedthesamemethodsforsoilerosion P a assessment with 137Cs and RUSLE than in Konz et al. (2009). The 137Cs measure- p J I e r ments were decay corrected for comparison purpose. Back Close | 2.3.1 137Cs to assess total net soil redistribution 20 D FullScreen/Esc is c A2inch×2inchNaI-scintillationdetector(Sarad,Dresden,Germany)wasusedtomea- u s 137 s Printer-friendlyVersion sure the in-situ Cs activity. The detector was mounted perpendicular to the ground io n ataheightof25cmtoreducetheradiusoftheinvestigatedareato1m.Measurement P InteractiveDiscussion a time was set at 3600s and each site was measured three times. p e r 9510 | Thedetectorwascalibratedagainstgammaspectroscopylaboratorymeasurements D with a 20% relative efficiency Li-drifted Ge detector (GeLi; Princeton Gamma-Tech, isc HESSD u Princeton,NJ,USA)attheDepartmentforPhysicsandAstronomy,UniversityofBasel. ss 137 io 10,9505–9531,2013 The resulting measurement uncertainty on Cs peak area (at 662 keV) was lower n P 5 than8%(errorofthemeasurementat1-sigma).Gammaspectrometrycalibrationand a p quality control were performed following the protocol proposed by Shakhashiro and e Impact of snow r Mabit (2009). A detailed description of the calibration procedure of the in-situ detector gliding on soil 137 | is provided by Schaub et al. (2010). For the conversion of the Cs inventories to soil redistribution D erosion rates we used the model as described by Konz et al. (2009). is c K.Meusburgeretal. u s 10 2.3.2 Assessment of soil redistribution by water erosion using the RUSLE sio n The RUSLE (Wischmeier and Smith, 1978) is an empirical model originally developed P TitlePage a in the United States. Several adapted versions for other regions as well as for differ- pe r Abstract Introduction ent temporal resolutions have been developed and applied with more or less success | (Kinnell, 2010). Despite its well-known limitation (highlighted in our introduction), we Conclusions References selected RUSLE because of the lack of simple soil erosion models specific for moun- D 15 is tain areas and moreover because of its better performance when compared to the c Tables Figures u s other existing models (Konz et al., 2010; Meusburger et al., 2010b). The RUSLE can s be calculated using the following equation: ion J I P a A=R·K ·LS·C·P (1) pe J I r Back Close where A is the predicted average annual soil loss (tha−1yr−1). R is the rainfall- runoff- | 20 erosivity factor (Nh−1) that quantifies the effect of raindrop impact and reflects the rate D FullScreen/Esc of runoff likely to be associated with the rain (Wischmeier and Smith, 1978). The soil isc u erodibilityfactorK (Nhkgm−2)reflectstheeaseofsoildetachmentbysplashorsurface ss Printer-friendlyVersion flow. The parameter LS (dimensionless) accounts for the effect of slope length (L) and ion slope gradient (S) on soil loss. The C-factor is the cover factor, which represents the P InteractiveDiscussion 25 a effects of all interrelated cover and management variables (Renard et al., 1997). pe r 9511 | For comparability between the RUSLE estimates of Konz et al. (2009) and the ones D assess in this study we used the same R factor of 97Nh−1. The K factor was cal- isc HESSD u culated with the K nomograph after Wischmeier and Smith (1978) using grain-size ss io 10,9505–9531,2013 analyses and carbon contents of the upper 15cm of the soil profiles. Total C content n P 5 of soils was measured with a Leco CHN analyzer 1000, and grain size-analyses were a p performed with sieves for grain sizes between 32 and 1000µm and with a Sedigraph e Impact of snow r 5100 (Micromeritics) for grain sizes between 1 and 32µm. L and S were calculated gliding on soil | after Renard et al. (1997). The support and practice factor P (dimensionless) was set redistribution D to 0.9 for some of the pasture sites because alpine pastures with cattle trails resemble is 10 small terrace structures, which are suggested to be considered in P (Foster and High- cu K.Meusburgeretal. s fill, 1983). For all other sites, P value was set to 1. The cover-and-management factor s io C was assessed for sites with and without dwarf shrubs separately using measured n P TitlePage fractional vegetation cover (FVC) in the field. a p e Forinvestigatedsiteswithoutdwarfshrubs(USDepartmentofAgriculture,1977)the r Abstract Introduction C factor can be estimated with: 15 | Conclusions References C=0.45·e−0.0456·FVC (2) D is c Tables Figures u s for sites with dwarf shrubs the following equation was used: s io J I n C=0.45·e−0.0324·FVC (3) P a p J I e The FVC was determined in April and September using a grid of 1m2 with a mesh r 2 Back Close width of 0.1m . The visual estimate of each mesh was averaged for the entire square | 20 meter. This procedure was repeated four times for each plot. The maximum standard D FullScreen/Esc deviation was approx. 5%. is c u s s Printer-friendlyVersion 2.4 Spatial modelling of snow glide distances io n P InteractiveDiscussion Weusedthespatialsnowglidemodel(SSGM,Leitingeretal.,2008)topredictpotential ap 2 e snow glide distances for an area of approximately 30km surrounding our study sites. r 25 9512 | The SSGM is an experimental model, which includes 6 main parameters: the forest D stand (0; 1), the slope angle (◦), the winter precipitation (mm), the slope exposition E isc HESSD u (0; 1), the static friction coefficient µ (–) and the slope exposition W (0; 1). Slope and ss s io 10,9505–9531,2013 aspect were derived from the digital elevation models DHM25 and below 2000ma.s.l. n P 5 the DOM (Swisstopo). The DOM is a high precision digital surface model with 2m a p resolution and an accuracy of ±0.5m at 1σ in open terrain and ±1.5m at 1σ in terrain e Impact of snow r withvegetation.TheDHM25hasaresolutionof25mwithanaverageerrorof1.5mfor gliding on soil | the Central Plateau and the Jura, 2m for the Pre-Alps and the Ticino and 3 to 8m for redistribution D the Alps (Swisstopo). Winter precipitation was derived from the MeteoSchweiz station is c K.Meusburgeretal. 10 locatedinAndermatt.WeusedtheresultfromaQuickBirdlandcoverclassificationwith u s aresolutionof2.4m(subsequentlyresampledto5m)aslandcoverinput(Meusburger s io et al., 2010a). Combining this land cover map with a land use map (Meusburger and n P TitlePage Alewell, 2009), it was possible to derive the parameter forest stand. To each of the 4 a p investigated land cover types a uniform static friction coefficient (µs) was assigned. er Abstract Introduction The static friction coefficient can be derived by: 15 | Conclusions References F D µs= Fr (4) iscu Tables Figures n s s where F is the normal force that can be calculated with ion J I n P a Fn=m·g·cosα (5) pe J I r Back Close wheregthegravitationalconstant(9.81ms−1),αistheslopeangle(◦)andmtheweight | of the snow glide shoe (in our study 202g). D FullScreen/Esc 20 is The initial force (F ), which is needed to get the glide shoe moving on the vege- c r u ® s tation surface, was measured with a spring balance (Pesola Medio 1000g). To ob- s Printer-friendlyVersion io tain representative values of F the measurement was replicated 10 times per sam- n r P InteractiveDiscussion ple site and subsequently averaged. The parameter estimates the surface roughness, a p which integrates the effect of different vegetation types and land uses on snow gliding. e 25 r 9513 | A detailed description of the model and its parameters has been provided by Leitinger D is et al. (2008). c HESSD u s Themodelwascalibratedwiththemeasuredsnowglidedistancesand285mmwin- s io 10,9505–9531,2013 terprecipitation(sumoftheprecipitationfromDecember2009toApril2010).Afterthe n P 5 calibration, potential snow glide distances with long-term average winter precipitation a p of 443mm (years 1959 to 2010) were modelled. e Impact of snow r gliding on soil | redistribution 3 Results and Discussion D is c K.Meusburgeretal. u 3.1 Snow glide measurements 2009/2010 s s io For each site the static friction coefficient as measure for surface roughness was de- n P TitlePage a 10 termined in autumn prior to the installation of the snow glide shoes. Lowest surface p e roughness was observed for the hayfields, followed by sites covered with Alnus viridis r Abstract Introduction on the north exposed slope, where only little undergrowth was observed (Table 1). | For the pastures with dwarf-shrubs, the two mean monitored values differed (0.04 and Conclusions References D 0.07) but were similar to that of pastures without dwarf-shrubs (0.06 to 0.07). Slightly isc Tables Figures u 15 higher values were observed for the dense undergrowth of Alnus viridis sites on the ss south exposed slope (0.07–0.08). io J I n Themeasuredsnowglidedistancesofthedifferentsitesvariedfrom1to189cm(see P a Table 1). A main proportion of this variability can be explained by the slope exposition p J I e r andthesurfaceroughness(seeFig.2).Withincreasingsurfaceroughness(expressed Back Close as the static friction coefficient; µ ) the snow glide distance declines. This decrease | 20 s is more pronounced for the south exposed slope (y =−1547.2µ +172.93; R2=0.50; D FullScreen/Esc p=0.036). For the north exposed slope the snow glide distancses and the variability isc u s are lower. Approximately 80% of the observed variability on the north exposed slope s Printer-friendlyVersion can be explained by the surface roughness (y =−622.17µ +43.09; R2=0.82; p= ion s P InteractiveDiscussion 25 0.033). For the south exposed slope, the snow glide distances are higher and only a p 50% of the variability can be explained by the difference in surface roughness. The e r 9514 |