• Auer, I., and Coauthors, 2007: HISTALP—historical instrumental climatological surface time series of the Greater Alpine Region. Int. J. Climatol., 27 (1) 1746.

    • Search Google Scholar
    • Export Citation
  • Brunetti, M., , M. Maugeri, , T. Nanni, , I. Auer, , R. Böhm, , and W. Schöner, 2005: Precipitation variability and changes in the Greater Alpine region over the 1800–2003 period. J. Geophys. Res., 111 , D11107. doi:10.1029/2005JD006674.

    • Search Google Scholar
    • Export Citation
  • Christensen, J. H., and Coauthors, 2007: Regional climate projections. Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 847–940.

    • Search Google Scholar
    • Export Citation
  • Costa, A. C., , and A. Soares, 2009: Homogenization of climate data: Review and new perspectives using geostatistics. Math. Geosci., 41 , 291305.

    • Search Google Scholar
    • Export Citation
  • Edwards, H. W., , M. F. Kostrzewa, , J. Jonkman, , D. Hopper, , S. Kinne, , J. Masters, , and E. Ruffel, 2001: Energy efficiency assessment report for Aspen Skiing Company snowmaking operations. Industrial Assessment Center, Department of Mechanical Engineering, Colorado State University Rep., 108 pp.

    • Search Google Scholar
    • Export Citation
  • Fauve, M., , H. Rhyner, , and M. Schneebeli, 2002: Preparation and Maintenance of Pistes. Handbook for Practitioners. Swiss Federal Institute for Snow and Avalanche Research Davos, 133 pp.

    • Search Google Scholar
    • Export Citation
  • Gruber, C., , and I. Auer, cited. 2008: Homogenisierung von Tagesdaten. Poster 1, 10. Österreichischer Klimatag, Vienna, Austria, Kulturabteilung der Stadt Wien, Wissenschafts- und Forschungsförderung and Österreichische Hagelversicherung. [Available online at http://www.austroclim.at/fileadmin/user_upload/ppt_10.Klimatag/Tagungsband_Klimatag2008.pdf].

    • Search Google Scholar
    • Export Citation
  • Hennessy, K., , P. Whetton, , I. Smith, , J. Bathols, , M. Hutchinson, , and J. Sharples, 2003: The impact of climate change on snow conditions in mainland Australia. CSIRO Atmospheric Research Rep., Aspendale, VIC, Australia, 47 pp. [Available online at http://www.cmar.csiro.au/e-print/open/hennessy_2003a.pdf].

    • Search Google Scholar
    • Export Citation
  • Jarvis, A., , H. I. Reuter, , A. Nelson, , and E. Guevara, cited. 2006: Hole-filled seamless SRTM data V3. International Centre for Tropical Agriculture (CIAT). [Available online at http://srtm.csi.cgiar.org].

    • Search Google Scholar
    • Export Citation
  • Jurković, A., , I. Auer, , and R. Böhm, cited. 2008: Gesamtschneehöhe – Vergleichende Zeitreihenanalyse. Poster 7, 10. Österreichischer Klimatag, Vienna, Austria, Kulturabteilung der Stadt Wien, Wissenschafts- und Forschungsförderung and Österreichische Hagelversicherung. [Available online at http://www.austroclim.at/fileadmin/user_upload/ppt_10.Klimatag/Tagungsband_Klimatag2008.pdf].

    • Search Google Scholar
    • Export Citation
  • Kendall, M. G., 1955: Rank Correlation Methods. 2nd ed. Griffin, 196 pp.

  • Kincaid, D. C., , and T. S. Longley, 1989: A water droplet evaporation and temperature model. Trans. ASAE, 32 , 457463.

  • Kromp-Kolb, H., , and H. Formayer, 2001: Klimaänderungen und mögliche Auswirkungen auf den Wintertourismus in Salzburg. Studie im Auftrag des Amtes der Salzburger Landesregierung, 33 pp. [Available online at http://www.salzburg.gv.at/pdf-kyoto-bericht-tourismus.pdf].

    • Search Google Scholar
    • Export Citation
  • Kuhn, M., 1987: Micro-meteorological conditions for snow melt. J. Glaciol., 33 (113) 2426.

  • Laternser, M., , and M. Schneebeli, 2003: Long-term snow climate trends of the Swiss Alps (1931–99). Int. J. Climatol., 23 , 733750.

  • Loibl, W., , A. Beck, , M. Dorninge, , H. Formayer, , A. Gobiet, , and W. Schöner, 2007: Reclip:more—Research for Climate Protection: Model Run Evaluation. Austrian Research Centers Systems Research Final Rep., 107 pp.

    • Search Google Scholar
    • Export Citation
  • Mann, H. B., 1945: Nonparametric tests against trends. Econometrica, 13 , 245259.

  • Marty, C., 2008: Regime shift of snow days in Switzerland. Geophys. Res. Lett., 35 , L12501. doi:10.1029/2008GL033998.

  • Olefs, M., , and F. Obleitner, 2007: Numerical simulations on artificial reduction of snow and ice ablation. Water Resour. Res., 43 , W06405. doi:10.1029/2006WR005065.

    • Search Google Scholar
    • Export Citation
  • Olefs, M., , and A. Fischer, 2008: Comparative study of technical measures to reduce snow and ice ablation in Alpine glacier ski resorts. Cold Reg. Sci. Technol., 52 , 371384. doi:10.1016/j.coldregions.2007.04.021.

    • Search Google Scholar
    • Export Citation
  • Petty, G. W., 2008: A First Course in Atmospheric Thermodynamics. Sundog Publishing, 337 pp.

  • Pröbstl, U., , and A. Prutsch, 2007: STRATEGE—Strategien für Wintersportorte im Zeichen des Klimawandels. Universität für Bodenkultur Wien, Vienna, Austria, 4 pp. [Available online at http://www.provision-research.at/cms/documents/files/Projekte/Stratege/071114_PT_Stratege.pdf].

    • Search Google Scholar
    • Export Citation
  • Scherrer, S. C., , and C. Appenzeller, 2006: Swiss Alpine snow pack variability: Major patterns and links to local climate and large-scale flow. Climate Res., 32 , 187199.

    • Search Google Scholar
    • Export Citation
  • Scherrer, S. C., , C. Appenzeller, , and M. Laternser, 2004: Trends in Swiss Alpine snow days: The role of local- and large-scale climate variability. Geophys. Res. Lett., 31 , L13215. doi:10.1029/2004GL020255.

    • Search Google Scholar
    • Export Citation
  • Scott, D., , G. McBoyle, , and B. Mills, 2003: Climate change and the skiing industry in southern Ontario (Canada): Exploring the importance of snowmaking as a technical adaptation. Climate Res., 23 , 171181.

    • Search Google Scholar
    • Export Citation
  • Scott, D., , G. McBoyle, , and A. Minogue, 2007: Climate change and Quebec’s ski industry. Global Environ. Change, 17 , 181190.

  • Sonntag, D., 1990: Important new values of the physical constants of 1986, vapour pressure formulations based on the ITS-90, and psychometer formulae. Z. Meteor., 40 , 340344.

    • Search Google Scholar
    • Export Citation
  • Steiger, R., 2008: Limits of snowmaking in a changing climate. Geophysical Research Abstracts, Vol. 10, Abstract EGU2008–A-02579. [Available online at http://www.cosis.net/abstracts/EGU2008/02579/EGU2008-A-02579.pdf?PHPSESSID=4997fdad47372d5e54162b98ef00254b].

    • Search Google Scholar
    • Export Citation
  • Wallace, J. M., , and P. V. Hobbs, 2006: Atmospheric Science—An Introductory Survey. Academic Press, 483 pp.

  • World Meteorological Organization, 2008: Measurement of humidity. Guide to Meteorological Instruments and Methods of Observation, 7th ed., WMO, 91–120.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Location of the stations used in the analysis [digital elevation data from Jarvis et al. (2006)].

  • View in gallery

    Seasonal mean (October–April) wet-bulb temperature of the eight long-term stations (individual record lengths and data gaps are indicated in Table 1). Note the different ordinate scales.

  • View in gallery

    Wet-bulb temperature anomalies. Seasonal deviations (mean of October–April) from the 1961–90 long-term mean (bars) and corresponding 10-yr running mean value (solid line).

  • View in gallery

    Anomalies of the number of potential snowmaking days (daily mean wet-bulb temperature ≤−2°C). Seasonal deviations (sum of October–April) from the 1961–90 long-term mean (bars) and corresponding 10-yr running mean value (solid line).

  • View in gallery

    Trend analysis for seasonal means (October–April) of measured air temperature, relative humidity, calculated wet-bulb-temperature, and potential snowmaking days for Patscherkofel, 2247 m MSL. The spectrum of analyzed subperiods is delineated by a black triangle [e.g., the trend over the longest available time span (1948–2007) is located at the upper tip of the triangle with a window width of 59 yr and a central year equal to 1977.5]. White areas within the triangle symbolize trends that are not statistically significant at the 95% level (p value = 0.05). Colors indicate trend magnitude of the significant trends.

  • View in gallery

    Trend analysis for seasonal means (October–April) of potential snowmaking days for all long-term stations. The spectrum of analyzed subperiods is delineated by a black triangle [e.g., the trend over the longest available time span (depending on record length) is located at the upper tip of the triangle; see example in Fig. 5]. White areas within the triangle (solid black lines) symbolize trends that are not statistically significant at the 95% level (p value = 0.05). Colors indicate trend magnitude of the significant trends.

  • View in gallery

    Trend analysis for the month of (top) November and (bottom) December of potential snowmaking days for all long-term stations. The spectrum of analyzed subperiods is delineated by a black triangle [e.g., the trend over the longest available time span (depending on record length) is located at the upper tip of the triangle; see example in Fig. 5]. White areas within the triangle (solid black lines) symbolize trends that are not statistically significant at the 95% level (p value = 0.05). Colors indicate trend magnitude of the significant trends.

  • View in gallery

    Pdfs of wet-bulb temperature in (top) November and (bottom) December for all long-term stations for two periods: 1967/68–86/87 (solid line) and 1987/88–2006/07 (dashed line). The vertical gray bar indicates the threshold wet-bulb temperature for snowmaking (twlim = −2°C).

  • View in gallery

    Calculated snow production potential (fan gun) for November sums using calculated wet-bulb temperature and up-to-date technical characteristics of Austrian fan guns for all 11 stations with available hourly data.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 250 250 60
PDF Downloads 179 179 41

Boundary Conditions for Artificial Snow Production in the Austrian Alps

View More View Less
  • 1 Institute of Meteorology and Geophysics, University of Innsbruck, Innsbruck, Austria
  • 2 Central Institute for Meteorology and Geodynamics, Regional Office for Tyrol and Vorarlberg, Innsbruck, Austria
© Get Permissions
Full access

Abstract

To assess how meteorological conditions favorable for the production of artificial snow vary in time and space, wet-bulb temperatures are calculated using temperature and humidity data of 14 Austrian stations between October and April for 1948–2007 (station altitudes 585–3105 m MSL). Technical specifications of snow guns are used to define a wet-bulb temperature threshold value of −2°C for snowmaking and a relationship between wet-bulb temperature and snowmaking capacity. The Mann–Kendall nonparametric-trend test is used to detect monotonic long-term changes in air temperature, relative humidity, wet-bulb temperature, and number of snowmaking days. It is applied multiple times to overlapping time periods to capture significant trends on different time scales. Results show a marked, common air- and wet-bulb seasonal mean (October–April) temperature increase between +1.5° and +3.1°C from 1980 to 1990 for a majority of stations with no trends thereafter. The number of snowmaking days per season decreased by −20 to −34 for half of the stations in the period around 1979–2003. No altitudes were especially affected by changes in the analyzed variables. The estimated volume of produced artificial snow shows high interannual variability and exhibits no trends at an hourly resolution over the last two decades.

Corresponding author address: Marc Olefs, Institute of Meteorology and Geophysics, Innrain 52, 6020 Innsbruck, Austria. Email: marc.olefs@uibk.ac.at

Abstract

To assess how meteorological conditions favorable for the production of artificial snow vary in time and space, wet-bulb temperatures are calculated using temperature and humidity data of 14 Austrian stations between October and April for 1948–2007 (station altitudes 585–3105 m MSL). Technical specifications of snow guns are used to define a wet-bulb temperature threshold value of −2°C for snowmaking and a relationship between wet-bulb temperature and snowmaking capacity. The Mann–Kendall nonparametric-trend test is used to detect monotonic long-term changes in air temperature, relative humidity, wet-bulb temperature, and number of snowmaking days. It is applied multiple times to overlapping time periods to capture significant trends on different time scales. Results show a marked, common air- and wet-bulb seasonal mean (October–April) temperature increase between +1.5° and +3.1°C from 1980 to 1990 for a majority of stations with no trends thereafter. The number of snowmaking days per season decreased by −20 to −34 for half of the stations in the period around 1979–2003. No altitudes were especially affected by changes in the analyzed variables. The estimated volume of produced artificial snow shows high interannual variability and exhibits no trends at an hourly resolution over the last two decades.

Corresponding author address: Marc Olefs, Institute of Meteorology and Geophysics, Innrain 52, 6020 Innsbruck, Austria. Email: marc.olefs@uibk.ac.at

1. Introduction

Changing climatic conditions and the resultant changes in the cryosphere affect the winter tourism industry. Changes in seasonal snow conditions, glacier cover, or frozen water bodies can have direct consequences that necessitate investigating technological measures to adapt to these changes. An example of one such measure is to artificially increase the mass balance of snow and ice within glacier ski resorts (Olefs and Obleitner 2007; Olefs and Fischer 2008). Another example is artificial snowmaking, as ski tourism is a key economic factor. The operation of ski resorts is tied to snow reliability provided by natural or artificial snow. Both natural and artificial snow depend on the meteorological conditions, making the skiing industry climate sensitive.

Several studies investigated the effects of climate variability and climate change on different parameters related to the natural snow cover in the European Alps. In Switzerland, Laternser and Schneebeli (2003) found a significant decrease in average snow depth, snow cover duration, and the number of snow days from the early 1980s to 1999. Marty (2008) concluded that the number of days with a certain snow depth above a given altitude-dependent threshold value decreased in a steplike manner at the end of the 1980s with no trend since then until the present. Both studies suggest that these observations are due to increased winter temperatures rather than to changed precipitation conditions, which is corroborated by Scherrer et al. (2004; Scherrer and Appenzeller 2006). In Austria, Jurković et al. (2008) depicted region-dependent trends with negative trends of snow cover duration and snow days in the south, and no significant trend in the east and north.

With the Historical Instrumental Climatological Surface Time Series of the Greater Alpine Region (HISTALP) dataset (Auer et al. 2007; online at http://www.zamg.ac.at/histalp/), these trend patterns can be explained. It contains homogenized records and a trend analysis of temperature, pressure, precipitation, sunshine, and cloudiness for the European Alps and its wider surroundings. These results show increased winter air temperatures for all regions and decreasing precipitation in the south throughout the last century. Similar negative trends for the temperate and subpolar Northern Hemisphere are found in Laternser and Schneebeli (2003).

The data analysis of Auer et al. (2007) explains the causes of the change in the European Alps. The results are characterized by a highly significant increase of air temperature in all seasons and regions. Precipitation trends have a lower statistical significance and are more complex to interpret, showing decreasing winter precipitation in all investigated subregions with greater drying in the south. A general problem of this precipitation trend analysis is the spatial and temporal heterogeneity of the statistical significance (lower p values than for air temperature) and of the magnitude of the observed trends. Brunetti et al. (2005) applied multiple trend analyses to overlapping time periods (i.e., running trend analysis) with the HISTALP dataset and demonstrate a high impact of the choice of the time window on trend analysis. Generally, both papers (Brunetti et al. 2005; Auer et al. 2007) show a more pronounced cold season positive temperature and negative precipitation trend in the period 1975–2000 compared with earlier times.

These findings together with projected temperature and precipitation changes for the European Alps (Christensen et al. 2007; Loibl et al. 2007) underscore the potential demand for increased artificial snow production as an adaptation strategy for ski resorts. As the potential production of artificial snow is a function of the meteorological conditions, the snowmaking industry itself may be concerned by rising air temperatures and modified moisture conditions. To our knowledge, two studies analyzed the past conditions and possible impacts of climate change on the production of artificial snow in Austria (Kromp-Kolb and Formayer 2001; Pröbstl and Prutsch 2007). They are both limited to a small regional scale and focus on a single factor to analyze snowmaking conditions. Different dry air temperature threshold values are used to establish the relationship between climate and artificial snow production. Snowmaking probabilities are calculated as well by dividing the number of days with a mean temperature below this threshold value by the total number of days in a defined period.

For two stations in the Austrian province of Salzburg, Kromp-Kolb and Formayer (2001) computed this probability for the period 1948–2000 using a −2°C threshold air temperature value. To simulate an increased temperature, they added 1° and 2°C to these observations, which revealed a probability decrease by 5%–10% per degree warming, depending on the altitude. Pröbstl and Prutsch (2007) investigate the change in snowmaking hours using a threshold of −3°C dry air temperature in the Austrian ski resorts Planai and Schladming. They detect a 40% decrease of possible snowmaking hours for the period 1988–2002 when compared with 1961–90. Other studies use climate station data and simple snow-modeling approaches, incorporating artificial snowmaking to simulate the season length of a ski resort in the past and in future scenarios in Canada and Austria (e.g., Scott et al. 2003, 2007, Steiger 2008).

The humidity conditions of the surrounding air are a key factor that should be considered in addition to air temperature in such studies, as the small liquid water droplets exiting the snow gun evaporate and consequently cool to the wet-bulb temperature of the air. The amount of cooling is directly proportional to the humidity of the air; therefore, less humidity means more evaporative cooling of the droplets and consequently better snowmaking conditions at same air temperatures. Wet-bulb temperature is thus a more appropriate variable to calculate snowmaking conditions than air temperature or relative humidity alone, as it is influenced by both. Since humidity is one of the World Meteorological Organization (WMO) standard variables recorded at climate stations it is readily available as well.

To our knowledge, there is one study (from Australia) that explicitly uses wet-bulb temperature as a composite of air temperature and humidity along with technical characteristics of snow guns to relate climate variables to artificial snowmaking capability (Hennessy et al. 2003). Another study (Edwards et al. 2001) recommends the use of ambient air temperatures instead of wet-bulb temperatures. The authors argue that melting conditions that may occur after the deposition of the frozen particles are better represented by ambient air temperatures.

As ski resorts and snowmakers are highly interested in the relationship between climate variables and artificial snowmaking, the present study intends to assess the evolution of these conditions over a wide range of regions and altitudes in Austria. To best determine the effects of these conditions on the physical process of snowmaking, we calculate wet-bulb temperatures, the industry standard, and use known technical characteristics of modern snow guns. (See supplemental material at the Journals Online Web site: http://dx.doi.org/10.1175/2010JAMC2251.s1.)

2. Materials and methods

a. Data

Air temperature and relative humidity records from 14 Austrian stations between 585 and 3105 m MSL were analyzed (see Table 1) and used to calculate wet-bulb temperatures. Time series begin at dates ranging from 1948 (earliest) to 1994 (latest). They are divided into manual (C) and automated (H) periods (see Table 1), consisting of three manual readings per day and automated hourly measured values, respectively. Hourly values are available for the last 8–25 yr starting in the early 1980s or mid-1990s (Table 1). Air temperature and relative humidity of the manual period were measured using different instruments: thermohydrographs from Fuess and Thies, liquid thermometers from Schneider and Janaer, as well as the Assman Psychrometer. For the automated period, linearized thermistors from Kroneis and Logotronic and a hair hygrometer from Lambrecht were used. General measurement accuracies are ±1% (relative humidity) and ±0.2°C (air temperature). All data were manually quality controlled by removing outliers and physically unrealistic values. Missing values were discarded from the trend analysis (see Table 1; data gaps). A homogenization procedure as used for the HISTALP dataset (Auer et al. 2007) would be preferable but was not applied as a method for daily time resolutions and is still under evaluation (Gruber and Auer 2008; Costa and Soares 2009; I. Auer 2009, personal communication). Stations were selected so as to represent key winter tourism destinations in Austria with as long as possible time series of relevant data. They are located in different altitudinal zones and are subject to different atmospheric flow conditions (Fig. 1).

For this study’s purposes, the local relative altitude of the stations—whether each station is atop a mountain or in a valley—is very important. The latter favors the formation of temporary or stationary inversion layers during winter months. As can be seen in Table 1, all stations were qualitatively defined as either valley or mountain stations with respect to small-scale topography.

Mean annual precipitation is determined largely by regional and altitudinal features (Table 2). Generally, it is highest for elevated stations located in the northern Alps (e.g., Galzig or Warth) and decreases toward the main ridge (Patscherkofel and Obergurgl), as moist air masses are blocked. Because of its high altitude, the Sonnblick (3105 m MSL) receives exceptionally high amounts of precipitation for its location. Kitzbühel (780 m MSL) has exceptionally high amounts of precipitation despite its low elevation, as the northern-limestone Alps are discontinuous north of the station. Mean annual air temperature is dominated by the altitude and varies between 7.5°C at Puchberg and −5.7°C at Sonnblick. For all stations with available climate data, annual and wintertime mean values (1961–90) of precipitation and temperature are shown in Table 2.

b. Wet-bulb temperature and artificial snowmaking

Artificial snow is produced by expelling small liquid water droplets (0.2 × 10−3–0.4 × 10−3 m diameter) from the snow gun through nozzles at high speed (>30 m s−1). These droplets must then freeze when exposed to the ambient atmosphere. Generally, there are two types of snow guns: air–water guns, in which large quantities of pressurized air are used to discharge the droplets, and fan guns that blow pressurized water as it leaves the nozzles to provide a wide dispersion. Pressures are approximately 25 000 hPa and mean velocities equal 30 m s−1 for fan guns and even higher for air–water guns. Droplet sizes are large enough to not be carried far away from the desired area and small enough to freeze during the travel time in the air. The very small size facilitates freezing via a high-heat transfer rate due to a large area–unit mass ratio. Time until refreezing depends on the energy balance of the droplets after leaving the snow gun providing that a sufficient number of freezing nuclei are present. The main components of the energy balance are (i) expansion cooling as a function of the pressure difference between the water (air) in the gun and the surrounding air, (ii) radiation balance as a function of incoming longwave and shortwave radiation, drop albedo, temperature, and emissivity, (iii) turbulent sensible-heat flux depending mainly on the temperature difference between the droplet and the air, and (iv) turbulent latent-heat flux depending mainly on the vapor-pressure gradients between the water droplet and the air.

For both the latent and the sensible heat flux, the synoptic wind speed is far less important than the gradients of humidity and temperature as droplet speed mostly dominates the synoptic wind and is in the same order of magnitude from one snow gun to the other. Therefore the synoptic wind is not accounted for as a meteorological boundary condition in this study.

If we neglect air pressure changes, expansion cooling is not a function of the meteorological boundary conditions. Furthermore, the calculation of the radiation balance is not straightforward: drop albedo is difficult to determine, and there is a lack of in situ radiation measurements. For these reasons, the present study concentrates on turbulent heat fluxes to capture the influence of the meteorological boundary conditions on artificial snow production. Wet-bulb temperature is sufficient to capture both the latent and the sensible heat flux relevant for artificial snowmaking, as it is the temperature a volume of air reaches when water is evaporated into it to the point of saturation. It is assumed that the air supplies all the latent heat of vaporization (Petty 2008). Wet-bulb temperature is also the temperature a raindrop reaches when traveling though a layer of constant wet-bulb temperature (Wallace and Hobbs 2006). Thus, liquid water droplets exiting the snow gun reach the wet-bulb temperature of the ambient air within a few seconds given typical droplet sizes and speeds. As the traveling time through the air is 10–15 s, the droplets will have ample time to reach the wet-bulb temperature (Kincaid and Longley 1989; various snow gun producers 2009, personal communication).

Thus, the production potential of artificial snowmaking is proportional to the wet-bulb temperature. The drier the air, the higher the maximum allowable air temperature can be for artificial snowmaking because of increased evaporation. The threshold wet-bulb temperature twlim is the highest wet-bulb temperature that still allows the production of artificial snow. The maximum snow production is limited by the maximum possible water flow through the gun. Modern snow guns are equipped with temperature and humidity sensors that in situ calculate wet-bulb temperatures to immediately modify water flow as atmospheric conditions change. Up-to-date (year 2007) technical data of four top-selling Austrian snow-gun manufacturers (cited as “various snow gun producers”) were used to calculate linear regressions between the snow production potential in meters cubed of snow per hour and wet-bulb temperature tw (°C) for a fan (ppf) and air–water (ppaw) gun, respectively [see Eqs. (1) and (2)]. These are representative mean relations as we used data of the top-selling fan and air–water gun of each of the four manufacturers, respectively. These mean relations are valid for a water temperature of less than 2°C, a water pressure of 25 bar, and a density of 400 kg m−3 (±10%) for the produced snow. We explicitly use the term production potential because true production is limited by water availability, and because we do not consider sublimation or wind erosion losses in the air or after deposition. Such losses are estimated to be around 5%–15% for fan guns and 15%–40% for air–water guns (various snow gun producers 2009, personal communication). Possible melting of the artificial snow after deposition is a function of the detailed energy balance at the site and cannot be considered herein. Equations (1) and (2) have a maximum possible production potential, limited only by water flow, of 72 m3 h−1 for a fan gun and 51 m3 h−1 for an air–water gun, at a wet-bulb temperature of −14°C or lower. The actual wet-bulb threshold temperature value twlim for artificial snowmaking given by the snow gun manufacturers equals −1.5°C for both the air–water and the fan gun. For the entire study we use a rounded value of −2°C that defines the highest possible wet-bulb temperature to produce artificial snow. This value is equal to that used in the work of Hennessy et al. (2003), and is lower compared to that of Edwards et al. (2001; −1.1°C).

Given these technical limitations, Eqs. (1)(6) are valid in the range −14°C ≤ tw ≤ −2°C:
i1558-8432-49-6-1096-e1
i1558-8432-49-6-1096-e2
i1558-8432-49-6-1096-e3
i1558-8432-49-6-1096-e4
i1558-8432-49-6-1096-e5
i1558-8432-49-6-1096-e6
The production potential is then converted to a potentially produced volume of snow by multiplying by the number of snowmaking hours hs. This is the so-called potential snowmaking capacity ppfvol and ppawvol in meters cubed [Eqs. (3) and (4)]. To get a better idea of what these cubic meters mean, we relate them to the area that can be provided with a basic, skiable layer of artificial snow, the skiable snow area (ssa). The ssa is generally defined as a 0.3-m-thick snow layer with a density of 400 kg m−3 that is sufficient to provide skiable conditions (Fauve et al. 2002). The calculation of ssa consists of multiplying the produced volume of snow ppfvol or ppawvol by the mean density of artificial snow (400 kg m−3; various snow gun producers 2009, personal communication) and then dividing by the mass per unit area necessary for ssa (0.3 m × 400 kg m−3 = 120 kg m−2) using a mean density for groomed ski pistes of 400 kg m−3 (Fauve et al. 2002; Olefs and Fischer 2008). Therefore ppfvol and ppawvol are multiplied by 400/120 (3.33) to obtain the desired area in m2 ssaf and ssaw, respectively [Eqs. (5) and (6)].

Artificial snow production is still possible at air temperatures of +3°C, if relative humidity is as low as 30%. For typical values of relative humidity around 75%, the maximum air temperature for snow production is −1°C (see chart in section a of supplemental material).

c. Calculation of wet-bulb temperature

The psychrometric method is a WMO standard observation method to derive humidity variables from two temperature measurements (World Meteorological Organization 2008): dry- and wet-bulb temperature using a psychrometric chart based on the psychrometric equation. The difference between dry-air temperature and wet-bulb temperature increases with decreasing mean humidity. We calculate the thermodynamic wet-bulb temperature iteratively using the standard psychrometric equation (Sonntag 1990; the exact calculation method is given in section d of supplemental material). Tw is a function of relative humidity, dry air temperature, and air pressure (the latter is because of the evaporation process). A standard barometric formula is used to derive mean air pressure using the station altitude, as wet-bulb temperature has only a weak pressure dependency (see chart in section b of supplemental material). Finally, we refer to the thermodynamic wet-bulb temperature whenever we mention calculated wet-bulb temperatures in this study.

d. Trend analysis

In view of the two available types of data (manual and automated), we divide our analysis into long-term (1948–2007) and short-term (1982–2007) periods.

For the long-term analysis we calculate daily means of wet-bulb temperature for all eight stations (denoted C in Table 1; arithmetic mean of 0600, 1300, and 1800 UTC). Since manual readings were replaced by the automated hourly values for some stations, we need to extend the time series. To do so, we combine it with hourly data [denoted H in Table 1] using the same calculation method (arithmetic mean of 0600, 1300, and 1800 UTC). This enables us to construct a continuous time series for statistical analysis. We then use these daily means to decide whether snowmaking is possible or not on a given day. A potential snowmaking day is defined as a day with a mean wet-bulb temperature ≤−2°C.

We use the nonparametric, linear Mann–Kendall trend test (Mann 1945; Kendall 1955) to identify statistically significant monotonic long-term changes of the measured air temperature, relative humidity, and calculated wet-bulb temperature, and to quantify changes in the number of snowmaking days. We apply the trend test for overlapping periods of different lengths (10–59 yr) within the period 1948–2007. The periods were shifted by 1 yr (e.g., 10-yr periods were calculated for 1948–57, 1949–58, and so on; 50-yr periods for 1948–97 and 1958–2007). The total number of analyzed subperiods ntot depends on the record length of the station and can be calculated as , where n is the number of years or seasons available. For the five longest time series from 1948 to 2007 (Kitzbühel, Patscherkofel, Sonnblick, Villacher Alpe, and Zell am See) with 59 seasons, ntot amounts to 1275 subperiods that are trend tested. Each subperiod is defined by the central year as well as by its length, the window width. Trends were calculated for each month separately and for the cool season (October–April). We test significance at a value of 0.05. The magnitude of the significant trends is given in the respective units (humidity and temperature) per season. (Figure 5, described below, shows an example of such a running trend analysis for Patscherkofel.)

The short-term analysis only uses hourly data. We do not apply statistical trend tests as we do not have a long enough series of hourly data. For a qualitative evaluation, potential snowmaking hours and the snow production potential are calculated at 10 stations in the period 1994–2007. Despite the lack of statistical analysis, we consider this short-term analysis to be of great importance. The snow production potential is a function of wet-bulb temperature within the defined limits (−2° and −14°C) and is not just a function of the likelihood that the daily mean exceeds our threshold value. The short-term analysis therefore better approximates real snowmaking conditions.

3. Results

a. Long-term analysis

Calculated potential snowmaking days increase linearly with altitude (0–16 days per month for Puchberg, 22–31 for Sonnblick) and for mean seasonal sums (51 at Puchberg, 202 for Sonnblick) as can be seen in Table 3. In the case of small altitudinal differences between stations (e.g., Zell am See versus Kitzbühel or Patscherkofel versus Villacher Alpe), regional variations of temperature and humidity may dominate. Generally, the month of October is not suited for artificial snowmaking in altitudes below 2000 m MSL, as it has less than five snowmaking days in general. In November, all stations have at least five snowmaking days, which under present conditions is sufficient to provide a basic skiable layer (ssa). The number of potential snowmaking days is greatest in January with 16 and 31 days for Puchberg and Sonnblick, respectively. The highest natural variability occurs at the three stations at lower altitude in the core season of January–March with up to ±9 days.

Seasonal mean values (October–April) of wet-bulb temperature for eight long-term stations (Fig. 2) show that conditions favorable or disadvantageous for artificial snowmaking are observed simultaneously at most stations on the seasonal mean scale. Sonnblick has pronounced low values of wet-bulb temperature because of its high elevation. Wet-bulb interannual temperature ranges are smaller for mountain stations compared to valley stations. This is a previously observed phenomenon that is due to the greater interannual variability of the snow cover in valleys or plains, which has direct consequences on the heat budget. An exception is the valley station in Obergurgl, which has low temperature variability similar to Patscherkofel or Villacher Alpe.

To detect changes in mean wet-bulb temperature, seasonal deviations compared to the 1960–90 mean are shown in Fig. 3. A common feature of these deviations for all stations is a continuous positive anomaly from the late 1980s on and a maximum positive deviation of up to +3°C for the season 2006/07. Before the late 1980s, a few stations show periods of slightly negative anomalies around the 1960s or 1970s. The anomaly in the seasonal sums of snowmaking days (Fig. 4) compared to the 1960–90 mean sum shows similar patterns. There is a common decrease in the number of snowmaking days from the late 1980s on and record anomalies of −60 days in St. Anton and −40 days at Patscherkofel, Villacher Alpe, Kitzbühel, and Zell am See that reflect the exceptional season of 2006/07.

The patterns in Fig. 5 indicate that negative trends in wet-bulb temperature lead to positive trends in the number of snowmaking days during the central years 1960–75. The top panels in Fig. 5 demonstrate that this is mainly due to an increase in air temperature rather than to a decrease in relative humidity. These trends are also visible in Figs. 3 and 4. The opposite is true for the central year period 1980–1990 where trends are toward higher wet-bulb temperatures and fewer snowmaking days. Generally, from station to station, significant trends in the number of seasonal snowmaking days are found for different periods of time (Fig. 6). Only a minority of stations have significant trends in common over the same period of time.

If the trend analysis is calculated by individual months as in Fig. 7, trend patterns are similarly different from station to station. November and December are the most important snowmaking months as the season starts. While during November there are few significant trends, the month of December shows more and more synchronous trends between the various stations. On the other hand, the mean November and December sums of calculated potential snowmaking days for the period 1960–2007 compared to 1990–2007 are decreased for three stations (St. Anton, Villacher Alpe, and Patscherkofel) only (Table 4).

The maximum and minimum linear trends of all analyzed variables are depicted in Table 5 for seasonal means or sums (see Table e.1 in supplemental material for monthly resolution). On the seasonal scale the strongest positive trends in air temperature coincide for six out of eight stations regarding air temperature and five out of eight stations regarding wet-bulb temperature in the period from about 1980 to about 1990. Air temperature trend values for this decade range between +1.5°C at Kitzbühel and +3.1°C at Patscherkofel. For wet-bulb temperature these values equal +1.6°C at Obergurgl and +3.0°C at Sonnblick.

The strongest trends in the number of snowmaking days on a monthly scale (see Table e.1 in supplemental material) occur in October and November between the mid-1960s and 1970s, when the number of snowmaking days increased. There was no general decrease synchronous to all stations. The maximum trend values are found at Kitzbühel: −38 days for 1968–78 and +31 days for 1974–86, respectively. Moreover, the strongest trends in most of the cases do not coincide with the statistically most significant ones (lowest p values). The length of the significant trend periods is quite different in many cases (see Figs. 6, 7).

To create a practical tool for ski resorts, we calculated daily climatological snowmaking probabilities for each station based on the exceedance frequency of daily wet-bulb temperature values compared to twlim. This was done for the entire record and for two 20-yr periods [1967/68–86/87 and 1987/88–2006/07, October–April; see charts in sections e.5 (entire record) and e.6 (20-yr periods) in supplemental material]. Figure 8 shows the probability density functions (pdfs) of wet-bulb temperature in November and December for these two periods of 20 yr and their position relative to twlim. This allows us to quantify past changes in terms of extreme values and mean conditions and to understand modified climatological probabilities. Generally, extreme values can change because of changes of the mean or of the spread. During November, changes of mean values and spread are small and there are no patterns common to multiple stations. In December, differences between the two periods are much larger and there is an increasing mean value and decreasing spread for all the stations except Puchberg (decreasing mean value, increasing spread). For snowmaking days this translates into less negative trends in November compared with December (see Fig. 7) and smaller (greater) changes in probabilities, respectively. In December, during the period 1987/88–2006/07, the maximum of the pdfs is close to twlim in Kitzbühel and Zell am See. This is reflected both in increasing and decreasing snowmaking probabilities between the two periods (see section e.6 in supplemental material).

Concerning trends in the seasonal sums of snowmaking days in the other months, most significant negative trends occur in March, followed by December and January (not shown). These trends do not fall into common periods of time and there are more negative trends than positive ones.

b. Short-term analysis

In Table 6, the step from a daily to an hourly time reveals micrometeorological phenomena due to local small-scale topography. Seasonal mean sums of potential snowmaking hours increase with altitude. Mean monthly sums plotted against elevation (not shown) reflect the seasonal variation of the vertical temperature gradient: weaker increases in December and January, stronger ones in spring and autumn. During October and April, conditions are unfavorable for artificial snowmaking. During October, below 1700 m MSL there are less than 3 days or 72 snowmaking hours, for April this altitudinal limit is ∼1200 m MSL. For Puchberg (585 m MSL) there is a well-defined increase from October to November with a minimum of 124 h or around 5 days in November. The best snowmaking conditions are again found in the month of January. The variability of daily and hourly values depends on the season and on the altitude. In November there is a typical standard deviation of 150 h (around 6 days) in altitudes of around 1500 m MSL with mean sums between 277 and 318 h. If one disregards Puchberg, natural variability during November varies in a small range (115–157 h).

The altitudinal dependence of calculated potential snowmaking capacity for a fan and an air–water gun (mean values for 1994–2007) changes in a similar way through the season as was observed for snowmaking hours (see Tables e.2 and e.3 in supplemental material). The only difference between snowmaking capacity and snowmaking hours occurs during December and January, when there is a marked increase in snowmaking capacity with altitude, as lower wet-bulb temperatures allow higher snow production until the upper limit for water flow is reached (−14°C). During November, ppfvol (see Table e.2 in supplemental material) varies with altitude from 3074 m3 (1.02 × 104 m2 of basic layer) at Puchberg to 15 199 m3 (5.06 × 104 m2 of basic layer) at Patscherkofel for one fan gun. The respective natural, interannual variability ranges from 2323 m3 (0.77 × 104 m2 of basic layer) to 7217 m3 (2.4 × 104 m2 of basic layer). Below altitudes of 2000 m MSL, standard deviations are high; in November, the standard deviations are about half of the means. The highest volumes were calculated at Villacher Alpe and Patscherkofel in January with more than 28 000 m3 (9.32 × 104 m2 of basic layer), which is around 50% of the maximum possible volume that could be produced in a month with continuous wet-bulb temperatures of −14°C.

Figure 9 presents the time series of potential snowmaking capacity for the month of November at 11 stations. In contrast to the long-term analysis, there is great similarity between the various stations. No general trends are visible over this period of time. The dominating feature is the interannual variability.

4. Discussion

The use of ambient air temperatures instead of wet-bulb temperatures as proposed by Edwards et al. (2001) to assess snowmaking conditions over a wider time frame (including possible melting after the production) was not followed in this study. We agree that the temperature of the snow, immediately after production, is often close to 0°C because of the release of heat during the freezing process (Fauve et al. 2002). Thus, a positive energy balance would lead to melting of some of the produced snow. Bearing in mind that snow can stay frozen up to +10°C (Kuhn 1987) and that we cannot quantify the exact site-specific energy balances at the time of particle deposition in the scope of this study, we prefer to use wet-bulb temperatures.

The sensitivities of calculated wet-bulb temperature to measured air temperature and relative humidity depend both on the temperature and relative humidity range. The sensitivity of wet-bulb temperature to changes in air temperature increases with increasing relative humidity. Likewise, a change of relative humidity has greater impacts for higher air temperatures than for lower ones. At −10°C measured air temperature, wet-bulb temperature changes by 0.12°C for a relative humidity change of ±4% (a typical year-to-year standard deviation). At freezing temperature (0°C) this sensitivity doubles. Between 70% and 80% relative humidity, the sensitivity of wet-bulb temperature for a change in measured air temperature of ±1°C (typical standard deviation) raises from 0.87° to 0.90°C °C−1. Thus, the impact of air temperature on wet-bulb temperature is higher than the impact of relative humidity. Thus, changes in air temperature are the predominant cause of changes in wet-bulb temperature, which was shown in Fig. 5 for Patscherkofel. Nevertheless, especially for air temperatures close to twlim, relative humidity changes can have an impact on the number of snowmaking days or hours and thus on the potential snowmaking capacity. Additionally, the weak air pressure dependency of wet-bulb temperature (see psychrometric equation in section d of supplemental material) may have some implications, as the analysis includes stations of different altitudes (585–3105 m MSL). Under the same air temperature and humidity conditions, the production capacity of artificial snow is slightly greater in higher altitudes. When comparing snowmaking conditions at stations in different altitudes, this dependency must be taken into account. Therefore we consider altitude as one of the key factors in our study—this is in contrast to “global climatologies” that are less suitable for ski resorts and snowmakers.

We modified the threshold temperature twlim by ±1°C, which does not affect our main conclusions. On the other hand, there is a strong sensitivity of the number of significant trends to a change in the p value (maximum of 25 trends per percent change in the p value). Finally, missing data in the period 1973–78 in Kitzbühel may have influenced our trend analysis and possibly lead to the extraordinary strong trends observed in snowmaking days compared to the other stations (see Table 6).

Small-scale local topography is poorly represented, meaning that our conclusions are only valid for the exact location of the respective station. Extrapolation of our results to altitudinal bands or expositions has to be done with caution as conditions for artificial snowmaking can change completely because of thermal and dynamical boundary conditions in response to local topography (i.e., inversion layers and warm foehn winds).

This study shows that altitudinal differences have the biggest impact on artificial snowmaking. An influence of inversion pools and the resulting more favorable conditions for artificial snowmaking at very low altitudes as reported by Pröbstl and Prutsch (2007) could not be documented with the available data. We suggest that this might be due to the large spatial difference between the analyzed stations and to the role of relative humidity. A vertical profile of stations, subject to same local meteorology (e.g., thermally induced along-valley winds), would be required to investigate this in more detail as done in Pröbstl and Prutsch (2007).

Our result of increased winter air temperatures in all Austrian regions agrees with Auer et al. (2007). The steplike decrease of snow depth in the late 1980s in Switzerland (Marty 2008) corresponds well to the strong winter temperature rise between 1980 and 1990 found in this study. For Austria, decreasing snow trends are only found for southern stations (Jurković et al. 2008), which does not correlate with air temperature or wet-bulb temperature trends calculated here. Our results are not especially comparable to other Austrian studies of artificial snowmaking, as they tend to focus on specific regions or use different variables derived from ambient air temperature.

5. Conclusions

Changes in the meteorological conditions for artificial snowmaking at 14 Austrian stations between 585 and 3105 m MSL were calculated. The results reveal a synchronous and marked increase in seasonal (October–April) mean air temperatures and wet-bulb temperatures from 1980 to 1990. Air temperatures rose at six out of eight stations, with changes from +1.5° to +3.1°C, and wet-bulb temperatures rose at five out of eight stations, with changes from +1.6° to +3.0°C. This is the strongest rise from 1948 to 2007 and there are no significant trends after 1990. This signal is also visible in the number of snowmaking days, but synchronous trends between the stations are less frequent. On a monthly scale of snowmaking days, synchronous trends between the stations are more frequently positive than negative. These positive trends are observed between the mid-1960s and mid-1970s. The decrease in snowmaking days is greatest in March, followed by December and January. Only a few stations show a decrease of snowmaking days between the 1960–2007 and 1990–2007 mean periods. Frequency distributions of wet-bulb temperatures show increasing mean values and a decreasing spread in December for seven out of eight stations. Generally, there are no typical changes in pdfs for specific regions or altitudes.

The season 2006/07 had the worst conditions for artificial snowmaking of any year over the period of observation (1948–2007). Generally, snowmaking conditions are best in January. In November, the standard deviation of snowmaking hours is generally not a function of altitude, and we calculate a minimum of more than 120 snowmaking hours or 5 days at all elevations. The interannual variability of snowmaking capacity is very large and there are no trends visible in the last 25 yr of hourly data.

Thus, meteorological effects of climate change on artificial snowmaking are visible on a daily scale of the last 60 yr but are not detectable on an hourly scale in the last 25 yr. The calculated trends have no region or elevation dependence. This is one of the first studies to analyze past meteorological conditions, as they affect snowmaking over a large range of regions and altitudes in complex mountainous terrain. Our findings are therefore unique and important for the skiing and snowmaking industry. Finally, the study provides a basis to better understand climate change effects in complex topographies.

Acknowledgments

We greatly acknowledge the various snow gun producers who provided us with the technical characteristics of state-of-the-art snow guns. This study was funded by the Professional Association of the Austrian Cable Cars and the Institute of Meteorology and Geophysics, University of Innsbruck. The authors are much obliged to Ekkehard Dreiseitl and Michael Kuhn for helpful comments that improved the clarity of the paper and to Saul Kinter and Lorna Raso for editing contributions. We also thank the anonymous reviewers and the editor for their valuable comments and suggestions.

REFERENCES

  • Auer, I., and Coauthors, 2007: HISTALP—historical instrumental climatological surface time series of the Greater Alpine Region. Int. J. Climatol., 27 (1) 1746.

    • Search Google Scholar
    • Export Citation
  • Brunetti, M., , M. Maugeri, , T. Nanni, , I. Auer, , R. Böhm, , and W. Schöner, 2005: Precipitation variability and changes in the Greater Alpine region over the 1800–2003 period. J. Geophys. Res., 111 , D11107. doi:10.1029/2005JD006674.

    • Search Google Scholar
    • Export Citation
  • Christensen, J. H., and Coauthors, 2007: Regional climate projections. Climate Change 2007: The Physical Science Basis, S. Solomon et al., Eds., Cambridge University Press, 847–940.

    • Search Google Scholar
    • Export Citation
  • Costa, A. C., , and A. Soares, 2009: Homogenization of climate data: Review and new perspectives using geostatistics. Math. Geosci., 41 , 291305.

    • Search Google Scholar
    • Export Citation
  • Edwards, H. W., , M. F. Kostrzewa, , J. Jonkman, , D. Hopper, , S. Kinne, , J. Masters, , and E. Ruffel, 2001: Energy efficiency assessment report for Aspen Skiing Company snowmaking operations. Industrial Assessment Center, Department of Mechanical Engineering, Colorado State University Rep., 108 pp.

    • Search Google Scholar
    • Export Citation
  • Fauve, M., , H. Rhyner, , and M. Schneebeli, 2002: Preparation and Maintenance of Pistes. Handbook for Practitioners. Swiss Federal Institute for Snow and Avalanche Research Davos, 133 pp.

    • Search Google Scholar
    • Export Citation
  • Gruber, C., , and I. Auer, cited. 2008: Homogenisierung von Tagesdaten. Poster 1, 10. Österreichischer Klimatag, Vienna, Austria, Kulturabteilung der Stadt Wien, Wissenschafts- und Forschungsförderung and Österreichische Hagelversicherung. [Available online at http://www.austroclim.at/fileadmin/user_upload/ppt_10.Klimatag/Tagungsband_Klimatag2008.pdf].

    • Search Google Scholar
    • Export Citation
  • Hennessy, K., , P. Whetton, , I. Smith, , J. Bathols, , M. Hutchinson, , and J. Sharples, 2003: The impact of climate change on snow conditions in mainland Australia. CSIRO Atmospheric Research Rep., Aspendale, VIC, Australia, 47 pp. [Available online at http://www.cmar.csiro.au/e-print/open/hennessy_2003a.pdf].

    • Search Google Scholar
    • Export Citation
  • Jarvis, A., , H. I. Reuter, , A. Nelson, , and E. Guevara, cited. 2006: Hole-filled seamless SRTM data V3. International Centre for Tropical Agriculture (CIAT). [Available online at http://srtm.csi.cgiar.org].

    • Search Google Scholar
    • Export Citation
  • Jurković, A., , I. Auer, , and R. Böhm, cited. 2008: Gesamtschneehöhe – Vergleichende Zeitreihenanalyse. Poster 7, 10. Österreichischer Klimatag, Vienna, Austria, Kulturabteilung der Stadt Wien, Wissenschafts- und Forschungsförderung and Österreichische Hagelversicherung. [Available online at http://www.austroclim.at/fileadmin/user_upload/ppt_10.Klimatag/Tagungsband_Klimatag2008.pdf].

    • Search Google Scholar
    • Export Citation
  • Kendall, M. G., 1955: Rank Correlation Methods. 2nd ed. Griffin, 196 pp.

  • Kincaid, D. C., , and T. S. Longley, 1989: A water droplet evaporation and temperature model. Trans. ASAE, 32 , 457463.

  • Kromp-Kolb, H., , and H. Formayer, 2001: Klimaänderungen und mögliche Auswirkungen auf den Wintertourismus in Salzburg. Studie im Auftrag des Amtes der Salzburger Landesregierung, 33 pp. [Available online at http://www.salzburg.gv.at/pdf-kyoto-bericht-tourismus.pdf].

    • Search Google Scholar
    • Export Citation
  • Kuhn, M., 1987: Micro-meteorological conditions for snow melt. J. Glaciol., 33 (113) 2426.

  • Laternser, M., , and M. Schneebeli, 2003: Long-term snow climate trends of the Swiss Alps (1931–99). Int. J. Climatol., 23 , 733750.

  • Loibl, W., , A. Beck, , M. Dorninge, , H. Formayer, , A. Gobiet, , and W. Schöner, 2007: Reclip:more—Research for Climate Protection: Model Run Evaluation. Austrian Research Centers Systems Research Final Rep., 107 pp.

    • Search Google Scholar
    • Export Citation
  • Mann, H. B., 1945: Nonparametric tests against trends. Econometrica, 13 , 245259.

  • Marty, C., 2008: Regime shift of snow days in Switzerland. Geophys. Res. Lett., 35 , L12501. doi:10.1029/2008GL033998.

  • Olefs, M., , and F. Obleitner, 2007: Numerical simulations on artificial reduction of snow and ice ablation. Water Resour. Res., 43 , W06405. doi:10.1029/2006WR005065.

    • Search Google Scholar
    • Export Citation
  • Olefs, M., , and A. Fischer, 2008: Comparative study of technical measures to reduce snow and ice ablation in Alpine glacier ski resorts. Cold Reg. Sci. Technol., 52 , 371384. doi:10.1016/j.coldregions.2007.04.021.

    • Search Google Scholar
    • Export Citation
  • Petty, G. W., 2008: A First Course in Atmospheric Thermodynamics. Sundog Publishing, 337 pp.

  • Pröbstl, U., , and A. Prutsch, 2007: STRATEGE—Strategien für Wintersportorte im Zeichen des Klimawandels. Universität für Bodenkultur Wien, Vienna, Austria, 4 pp. [Available online at http://www.provision-research.at/cms/documents/files/Projekte/Stratege/071114_PT_Stratege.pdf].

    • Search Google Scholar
    • Export Citation
  • Scherrer, S. C., , and C. Appenzeller, 2006: Swiss Alpine snow pack variability: Major patterns and links to local climate and large-scale flow. Climate Res., 32 , 187199.

    • Search Google Scholar
    • Export Citation
  • Scherrer, S. C., , C. Appenzeller, , and M. Laternser, 2004: Trends in Swiss Alpine snow days: The role of local- and large-scale climate variability. Geophys. Res. Lett., 31 , L13215. doi:10.1029/2004GL020255.

    • Search Google Scholar
    • Export Citation
  • Scott, D., , G. McBoyle, , and B. Mills, 2003: Climate change and the skiing industry in southern Ontario (Canada): Exploring the importance of snowmaking as a technical adaptation. Climate Res., 23 , 171181.

    • Search Google Scholar
    • Export Citation
  • Scott, D., , G. McBoyle, , and A. Minogue, 2007: Climate change and Quebec’s ski industry. Global Environ. Change, 17 , 181190.

  • Sonntag, D., 1990: Important new values of the physical constants of 1986, vapour pressure formulations based on the ITS-90, and psychometer formulae. Z. Meteor., 40 , 340344.

    • Search Google Scholar
    • Export Citation
  • Steiger, R., 2008: Limits of snowmaking in a changing climate. Geophysical Research Abstracts, Vol. 10, Abstract EGU2008–A-02579. [Available online at http://www.cosis.net/abstracts/EGU2008/02579/EGU2008-A-02579.pdf?PHPSESSID=4997fdad47372d5e54162b98ef00254b].

    • Search Google Scholar
    • Export Citation
  • Wallace, J. M., , and P. V. Hobbs, 2006: Atmospheric Science—An Introductory Survey. Academic Press, 483 pp.

  • World Meteorological Organization, 2008: Measurement of humidity. Guide to Meteorological Instruments and Methods of Observation, 7th ed., WMO, 91–120.

    • Search Google Scholar
    • Export Citation
Fig. 1.
Fig. 1.

Location of the stations used in the analysis [digital elevation data from Jarvis et al. (2006)].

Citation: Journal of Applied Meteorology and Climatology 49, 6; 10.1175/2010JAMC2251.1

Fig. 2.
Fig. 2.

Seasonal mean (October–April) wet-bulb temperature of the eight long-term stations (individual record lengths and data gaps are indicated in Table 1). Note the different ordinate scales.

Citation: Journal of Applied Meteorology and Climatology 49, 6; 10.1175/2010JAMC2251.1

Fig. 3.
Fig. 3.

Wet-bulb temperature anomalies. Seasonal deviations (mean of October–April) from the 1961–90 long-term mean (bars) and corresponding 10-yr running mean value (solid line).

Citation: Journal of Applied Meteorology and Climatology 49, 6; 10.1175/2010JAMC2251.1

Fig. 4.
Fig. 4.

Anomalies of the number of potential snowmaking days (daily mean wet-bulb temperature ≤−2°C). Seasonal deviations (sum of October–April) from the 1961–90 long-term mean (bars) and corresponding 10-yr running mean value (solid line).

Citation: Journal of Applied Meteorology and Climatology 49, 6; 10.1175/2010JAMC2251.1

Fig. 5.
Fig. 5.

Trend analysis for seasonal means (October–April) of measured air temperature, relative humidity, calculated wet-bulb-temperature, and potential snowmaking days for Patscherkofel, 2247 m MSL. The spectrum of analyzed subperiods is delineated by a black triangle [e.g., the trend over the longest available time span (1948–2007) is located at the upper tip of the triangle with a window width of 59 yr and a central year equal to 1977.5]. White areas within the triangle symbolize trends that are not statistically significant at the 95% level (p value = 0.05). Colors indicate trend magnitude of the significant trends.

Citation: Journal of Applied Meteorology and Climatology 49, 6; 10.1175/2010JAMC2251.1

Fig. 6.
Fig. 6.

Trend analysis for seasonal means (October–April) of potential snowmaking days for all long-term stations. The spectrum of analyzed subperiods is delineated by a black triangle [e.g., the trend over the longest available time span (depending on record length) is located at the upper tip of the triangle; see example in Fig. 5]. White areas within the triangle (solid black lines) symbolize trends that are not statistically significant at the 95% level (p value = 0.05). Colors indicate trend magnitude of the significant trends.

Citation: Journal of Applied Meteorology and Climatology 49, 6; 10.1175/2010JAMC2251.1

Fig. 7.
Fig. 7.

Trend analysis for the month of (top) November and (bottom) December of potential snowmaking days for all long-term stations. The spectrum of analyzed subperiods is delineated by a black triangle [e.g., the trend over the longest available time span (depending on record length) is located at the upper tip of the triangle; see example in Fig. 5]. White areas within the triangle (solid black lines) symbolize trends that are not statistically significant at the 95% level (p value = 0.05). Colors indicate trend magnitude of the significant trends.

Citation: Journal of Applied Meteorology and Climatology 49, 6; 10.1175/2010JAMC2251.1

Fig. 8.
Fig. 8.

Pdfs of wet-bulb temperature in (top) November and (bottom) December for all long-term stations for two periods: 1967/68–86/87 (solid line) and 1987/88–2006/07 (dashed line). The vertical gray bar indicates the threshold wet-bulb temperature for snowmaking (twlim = −2°C).

Citation: Journal of Applied Meteorology and Climatology 49, 6; 10.1175/2010JAMC2251.1

Fig. 9.
Fig. 9.

Calculated snow production potential (fan gun) for November sums using calculated wet-bulb temperature and up-to-date technical characteristics of Austrian fan guns for all 11 stations with available hourly data.

Citation: Journal of Applied Meteorology and Climatology 49, 6; 10.1175/2010JAMC2251.1

Table 1.

Metadata of the 14 stations used for the climatological analysis. Time series resolution is divided in a period with three manual readings per day (C) and automated hourly measured values (H). Years with too many missing data are discarded from the analysis (data gaps). The station type (M = mountain station, V = valley station) is important for micrometeorological reasons; x = no data gaps.

Table 1.
Table 2.

Climatological key variables of the 14 stations used for the long-term analysis (MAT = mean annual air temperature; PSUM = mean annual precipitation sum; PSUMW = mean winter precipitation sum; MWT = mean winter air temperature; MWRH = mean winter relative humidity; MWTW = mean winter wet-bulb temperature). Annual mean values (MAT, PSUM) are calculated for the period 1961–90; winter mean values are valid for October–April within the respective available time span; x = no data.

Table 2.
Table 3.

Monthly mean and standard deviation (in parentheses) of the potential number of snowmaking days for 1960–2007.

Table 3.
Table 4.

Sum of calculated potential snowmaking days (November + December) for three mean periods of all eight long-term stations.

Table 4.
Table 5.

Maximum (first row of variable) and minimum (second row) values of all linear trends of snowmaking days calculated for overlapping periods of different lengths (windows) at the respective station using the Mann–Kendall trend test. Trend values (boldface) are given in units per season (October–April) of the four analyzed meteorological variables relevant for artificial snow production at the significance level of 95%. The two dates per trend value indicate the start and end year for the period for which the trend is calculated (x = trend below significance level).

Table 5.
Table 6.

Monthly mean sums and standard deviations (in parentheses) of calculated potential snowmaking hours, common period 1994–2007, for 10 short-term stations.

Table 6.

* Supplemental information related to this paper is available at the Journals Online Web site: http://dx.doi.org/10.1175/2010JAMC2251.s1.

Supplementary Materials

Save