The future of the Olympic Winter Games in an era of climate change

Abstract

The Olympic Winter Games (OWG) stands as a symbol of international cross-cultural exchange through elite-level sport. As a mega-event with a significant reliance on a specific range of weather conditions for outdoor competitions, the OWG have developed several technologies and strategies to manage weather risk. Can these climatic adaptations cope with future climate change? Based on an analysis of two key climate indicators (probability of a minimum temperature of ≤0°C, and probability of a snow depth of ≥30 centimetres with advanced snowmaking capacity), this paper examines how projected changes to climate will impact the ability of the 19 previous host cities/regions to provide suitable conditions for outdoor competitions in the future. The results indicate that while the 19 former OWG hosts all have a suitable climate in the 1981–2010 period, only 11 or 10 (low–high-emission scenarios) remain climatically suitable in the 2050s, with as few as 6 in the high-emission scenario of the 2080s. The analysis reveals that climate change has important implications for the future geography of OWG host cities/regions as well as broader implications for participation in winter sport.

Keywords:

• Olympic Winter Games
• climate change
• sports tourism
• mega-events
• weather risk

Introduction

The Olympic Winter Games (OWG) will commemorate its first centennial in 2024. Over the past nine decades, this increasingly global celebration of winter sport has grown to become one of the world's mega-events. The economic, social and environmental impacts of the OWG vary significantly from games to games and have been widely debated (International Olympic Committee [IOC], 2012a; Mangan & Dyreson, 2010; Wallechinsky & Loucky, 2010). Nonetheless, the international prestige of hosting the OWG and positive Olympic legacy for host cities/regions that can result from massive infrastructure investment by higher levels of government, economic development and increased tourism can explain why cities/regions compete aggressively for the opportunity to host the Olympics.

The OWG is a mega-event with a significant dependency on weather conditions. Weather directly affects preparations for the games, outdoor opening and closing ceremonies, fairness of outdoor competitions, the ability to complete the full competition programme, spectator comfort, transportation, and visibility and timing of television broadcasts. The success of the games has often been partially attributed to favourable weather, while poor weather has been highlighted as one of the greatest challenges faced by organising committees (Rutty, Scott, Steiger, & Johnson, 2014). While much of the weather risk of historic OWG was overcome by moving competitions indoors, the specific climate and terrain conditions required by the diverse outdoor winter sports that comprise the OWG restrict the range of cities/regions that are capable of hosting the event.

Rutty et al. (2014) document the wide range of climatic adaptation strategies (e.g. transition of some competitions to indoor venues, snowmaking and advanced weather forecasting) that have been developed over the last 90 years to manage weather risk at the OWG. The need for weather risk management strategies by Olympic organisers has intensified, as the average February daytime temperature of OWG locations has steadily increased – from 0.4°C in the 1920–1950s, to 3.1°C in the 1960–1990s, to 7.8°C in games held in the twenty-first century (Rutty et al., 2014). Indeed, it would be difficult to imagine successfully completing the Olympic programme exclusively on natural ice and snow at the warmer host locations of the early-twenty-first century, compared to the colder and decidedly more alpine hosts of the early decades of the twentieth century. As weather-related impacts at recent OWG clearly reveal, there are limits to what current weather risk management strategies can cope with. These limits will be increasingly tested in a warmer world.

As the world comes together for the 22nd Winter Olympics in Sochi, Russia in 2014, the United Nations Intergovernmental Panel on Climate Change (IPCC) has begun to release the findings of its fifth Assessment on global climate change. The first report (IPCC, 2013) documents the observed changes in the global climate system, including a 0.85°C warming in global average surface temperatures between 1880 and 2012 and continued decline in Northern Hemisphere snow cover and glacier ice since the mid-twentieth century. With even stronger scientific confidence, the IPCC (2013) concluded that the ‘human influence on the climate system is clear. … (and) … has been the dominant cause of the observed warming since the mid-20th century’. The IPCC (2013) also emphasises that human-caused global climate change has just begun, and depending on future greenhouse gas (GHG) emissions, additional warming of global average surface temperatures of 0.3–4.8°C (relative to 1986–2005) is likely to occur by end of the twenty-first century. Critically, the IPCC (2013) anticipates that additional warming in the winter months will cause a further decrease in Northern Hemisphere snow cover and ice extent.

The implications of this projected climate change for winter sports and mega-events such as the OWG are unmistakable. Several studies have demonstrated the potential negative impact of future climate change on outdoor winter sports (see Scott, Hall, & Gössling, 2012 for a summary), but the implications of a warmer world for the OWG remains uncertain. This paper assesses whether projected climate change represents a long-term risk to the future viability of the OWG. Specifically, we examine which of the previous 19 host cities/regions would continue to have a climate suitable to once again host the full outdoor athletics programme of the OWG in the mid- to late-twenty-first century. Any differential climatic suitability among past host cities/regions would have implications for the IOC's consideration of bids to host future OWG.

Literature review

The evolution of the OWG

The OWG is the world's premier winter sporting event. This celebration of winter sport has grown from a modest gathering of 250 amateur athletes from 16 countries competing in 16 medal events at the 1924 games in Chamonix, France, to over 2500 athletes, representing 82 countries, competing in 86 medal events at the 2010 games in Vancouver, Canada (IOC, 2013). The OWG has always attracted large numbers of spectators and international tourists (Wallechinsky & Loucky, 2010), but has grown to become one of the world's sporting mega-events, with approximately 1.5 million tickets sold at each of the recent games in Salt Lake City, USA and Vancouver, Canada (IOC, 2012a).

The OWG has undergone a radical transformation over the last 40 years, since the decision to allow sporting professionals of all types to compete, promotion of the games to the world through partnerships with media and commercialisation through a wide range of private-sector sponsorships (Mangan & Dyreson, 2010; Preuss, 2004; Wallechinsky & Loucky, 2010). As a global mega-event, host regions and countries garner international prestige through the promotion of their natural and cultural heritage – both genuine and packaged as a tourism product (Garcia, 2008) – during television broadcasts to a worldwide audience of billions (e.g. the 2010 games reached 200 countries and a potential audience of 3.8 billion people worldwide [IOC, 2013]). The worldwide media broadcast revenues for the most recent games in Vancouver, Canada, exceeded US$1.2 billion (IOC, 2012a). The increased economic value of the games has influenced the timing of events, to better coincide with prime viewing times of the largest markets and broadcast sponsors) and increased the salience of weather risk management strategies to ensure both quality of conditions and timing of competitions (Rutty et al., 2014).

The OWG generates economic, social and environmental impacts on the host city and region. These negative and positive impacts have been widely debated elsewhere (Mangan & Dyreson, 2010; Wallechinsky & Loucky, 2010). Negative impacts can include crowding (Ritchie & Smith, 1991), price inflation (Ritchie & Aitken, 1984), increased crime (Mihalik & Cummings, 1995) and long-term national or state/provincial government debt (Preuss, 2004). Many examples exist of ‘white elephant’ sports infrastructure and large public debt associated with hosting the Olympics. The IOC has become increasingly mindful of these negative impacts and limits the qualifying host countries to those that have the resources and infrastructure to successfully host an Olympic Games without negatively impacting the region or nation. The IOC requires prospective cities/regions to include a legacy plan in the bid proposal, to demonstrate the long-term economic, social and environmental impacts the Olympics will have on the host region and has worked with host cities/regions/countries to document the long-term legacies of the games (IOC, 2012a).

Although the overall economic benefit of hosting the Olympics varies significantly from games to games, the Olympic legacy for host city/region is generally considered to be positive. This is in part because higher levels of government (federal and state/provincial levels) largely make the massive investment in infrastructure to host the games, leaving much improved transportation systems, additional housing, and sporting and tourism related infrastructure (Deccio & Baloglu, 2002; Gratton & Preuss, 2008; IOC, 2012b). The stature of hosting an OWG also brings long-term business development and increased tourism to host cities/regions (Bridges, 2008; IOC, 2012b; Jeong & Faulkner, 1996; Madden, 2002; Ritchie & Smith, 1991; Smith & Stevenson, 2009). It is the promise of these types of tangible economic and reputation benefits that drive cities and regions from around the world to engage in the highly competitive and lengthy process of bidding to host the OWG.

Climate change and vulnerability of winter sports

Concern about global climate change has increased worldwide and continues to feature prominently in high-profile international policy debates. Reviews of international climate change mitigation commitments conclude that the policy goal of restricting global warming to below 2°C is increasingly unlikely and the trajectory is towards a warming of +4°C or greater by the end of the twenty-first century (Peters et al., 2013). Although the consequences of climate change will vary geographically, it is inevitable that all nations and economic sectors will have to adapt to additional climatic change in the decades ahead. This has led to an explosion of interest in climate change impacts and adaptation research (Adger, Arnell, & Tompkins, 2005; Janssen, Schoon, Ke, & Borner, 2006; Stehr & von Storch, 2010).

With multiple sensitivities to climate and environmental changes, the outdoor sports industry, associated tourism sectors and mega-events such as the OWG are no exceptions. The rapidly growing literature on climate change and recreation/tourism (Becken & Hay, 2007, 2012; Hall & Higham, 2005; Scott, McBoyle, Minogue, & Mills, 2006; Scott et al., 2012) shows that the implications of climate change will vary by market segment and geographic region, and that all tourism destinations will need to adapt to climate change to minimise risks and capitalise on new opportunities, in an economically, socially and environmentally sustainable manner (Scott et al., 2012). For time-limited and weather-dependent mega-events, such as the OWG, the implications of climate change are even more significant, as events have substantially less flexibility to adopt activity or time substitution adaptations.

The risks posed by climate change to snow-based sports/recreation, particularly the large international ski tourism industry, have received considerable attention in the scholarly literature, government assessment reports and the media. The climate change vulnerability of the ski industry has been examined to some extent by over 30 studies in 13 countries (Scott, Gössling, & Hall, 2012). This geographically and methodologically diverse literature has consistently projected decreased reliability of natural snow cover, shortened and more variable ski seasons, increased snowmaking requirements, contraction in the number of operating ski areas, altered ski tourism revenues and employment, and declining real-estate values of vacation properties. The extent and timing of these impacts depend on the rate of climate change, the types of adaptation considered, and the relative impact on competing ski tourism regions. International comparative analyses of major ski markets have demonstrated that climate change risk is not evenly distributed among or within regional ski tourism markets. Rather, they are specific destinations and ski areas that are at risk due to a variety of climatological, operational and physical characteristics (Scott & Steiger, 2013). This research points to a growing scarcity of specific environmental conditions, coupled with infrastructural, economic and political criteria that could support a mega-event, such as the OWG.

The current literature on mega-events, including the Olympics, points to limitations in how environmental impacts are assessed and how environmental remediation and restoration are implemented (Collins, Jones, & Munday, 2009; Laing & Frost, 2010). Despite a high level of ‘environmentally friendly’ rhetoric from the IOC and strong integration of environmental considerations into the bidding process (Cantelon & Letters, 2000), the actual assessment of the environmental impacts of an event such as the Olympics remains problematic (Collins et al., 2009). Dickson and Arcodia (2010) call for increased inquiry into how events adopt sustainability policies, specifically outlining key failures of the Olympic Games to consider environmental factors. As the global events sector remains a significant contributor to climate change, the movement from consideration to actual assessment and remediation of environmental impacts of the OWG is a needed initiative. The declaration of the 2010 Vancouver OWG as ‘carbon neutral’, with carbon offset programmes and policies, stands as an initially foray into reconciling the overall climate and environmental impacts of the OWG mega-event. With the implementation of climate adaptation technologies to ensure adequate conditions for Olympic competitions (Rutty et al., 2014), the sustainability and environmental impact of future OWG is an area of pressing research need.

Moreover, with the prominent role of snow and temperature-dependent sports at the OWG, there is a clear relevance of sports and weather research for further understanding the climate vulnerability of the games itself. Studies that have examined the impacts of weather on Olympic sports (Borghesi, 2007; Koch & Panorska, 2013; Martin, 1996; Peiser & Reilly, 2004; Verdaguer-Codina, Martin, Pujol-Amat, Ruiz, & Prat, 1995), cold weather athletic performance (Buhl, Fauve, & Rhyner, 2001; Gould, Greenleaf, Chung, & Guinan, 2002; Niinimaa, Shephard, & Dyon, 1979; Rammsayer, Bahner, & Netter, 1995) and winter tourism events (Scott et al., 2002; Scott, McBoyle, Minogue, & Mills, 2006) offer additional insights into how changing temperatures and snow conditions could affect the performance of Olympic athletes and the comfort of spectators. This study considers the insights from this body of work in the specific context of developing climate change vulnerability indicators and adaptation strategies.

Snowmaking as a winter sports climate adaptation

The technical production of snow as a way to improve or guarantee snow conditions for winter sports was first implemented in 1952 at a ski resort in the USA (Scott et al., 2012). Snowmaking technology has since become nearly universally employed by ski areas throughout Eastern Canada and the USA (Scott & McBoyle, 2007). Additionally, snowmaking is widely adopted in all major ski regions of the world, including Western North America, the European Alps, Japan, and Australia (Abegg, Agrawala, Crick, & De Montfalcon, 2007; Hennessy et al., 2008; Scott, 2006). The implementation of snowmaking has substantially reduced the vulnerability of these ski tourism markets to adverse weather and inter-annual climate variability (Dawson & Scott, 2013; Hennessy et al., 2008; Scott, Dawson, & Jones, 2008; Scott, McBoyle, & Mills, 2003; Scott, McBoyle, & Minogue, 2007; Steiger, 2010; Steiger & Abegg, 2013; Steiger & Stötter, 2013). Lake Placid in 1980 was the first OWG to introduce snowmaking, which allowed alpine events to run on schedule despite the worst snow drought in the eastern USA since 1887 (Lake Placid, 1980). Since the late 1980s, snowmaking capacity has been mandatory for OWG competition sites.

Methods

Indicators for assessing the impact of climate change on the OWG

To assess which of the 19 locations that have formerly hosted the OWG would have a climate suitable to potentially host future games in the mid- to late-twenty-first century under projected climate change, several climatic indicators important to winter sports competitions were identified from the literature, and a content analysis conducted of the official post-games report submitted to the IOC by each from the Organising Committee from 1924 to 2010. Warm temperatures, rain, storms, fog, heavy snowfall and lack of snow were reported to have caused a range of impacts at the OWG outlined by Rutty et al. (2014). As indicated, much of the weather risk of former games has been overcome by moving many competitions indoors, leaving outdoor sports including alpine and cross-country skiing, ski jump and others vulnerable to natural conditions. Here, we focus on lack of snow and the presence of rain and warm temperatures. Other potentially relevant weather conditions, such as fog, heavy snowfall and storms are not adequately represented in global climate models, precluding local-scale projections. The indicators identified, with their relevance to winter sports competitions, specifically to the OWG which are typically held in the month of February, are summarised below:

1. probability of average maximum temperature ≥10°C in February – maximum temperatures above 10°C cause substantial deterioration of snow and ice quality, with a corresponding impact on durability (e.g. rutting) and fair/safe competition surfaces (e.g. soft and slow ski surfaces), however the occurrence is typically mid-day, so that scheduling of outdoor events can adapt, provided daily minimum temperatures are below freezing;

2. probability of average minimum temperature ≤0°C in February – when daily minimum temperatures are above freezing, snow and unrefrigerated ice surfaces continue to degrade and cannot refreeze overnight (snowmaking is not possible either), which can hamper the preparation of high-quality surfaces needed for elite competition and disrupt scheduling of competitions to ensure fair conditions for athletes;

3. number of days with liquid precipitation in February – rain at low temperatures is unpleasant for spectators and can also negatively impact snow and ice quality for competitions;

4. probability of sufficient snow base for skiing – sufficient snow depth is a precondition for snow-based competitions, especially alpine and cross-country skiing (a minimum snow depth of ≥30 centimetresin smooth terrain and ≥60 centimetres in rough terrain are identified in the literature – Abegg, 1996; Scott, 2003);

5. average natural snow depth on 1 February – natural snow fall has historically been a critical climatological factor for winter sports, but as snowmaking has become standard operating practice at elite international skiing events, the salience of this indicator for the operational feasibility of the OWG has declined and now is largely an aesthetic indicator, where snow cover provides an environment spectators would expect at the OWG;

6. average snow depth with snowmaking on 1 February – snowmaking has become an integral technology to supplement natural snow fall at the OWG and the capacity to make sufficient snow has supplanted natural snow fall as the key indicator of operational feasibility for ski areas;

7. number of snowmaking hours in January – efficient snowmaking is considered physically possible at temperatures below −5°C and with advanced snowmaking systems capable of producing a skiable snow depth of 10 centimetres of snow per day (Scott et al., 2008; Steiger & Abegg, 2013) a minimum of snowmaking 72 hours is needed to open a ski slope with 30 centimetre base or 144 hours for a 60 centimetre base.

This list of potential indicators was refined in three stages. First, indicators that would not prevent sporting competitions from taking place, but rather would require scheduling changes and reduce the winter aesthetic of the games, were eliminated from the analysis. These types of indicators included rainfall and natural snow fall/depth. Second, strong cross-correlations were found between the remaining temperature-based indicators, for example, where temperature thresholds also have important influences on precipitation indicators (i.e. when daily minimum temperature is above freezing, a considerable amount of precipitation falls as rain, further deteriorating snow and ice quality). These indicators were considered as secondary to the critical requirement of minimum daily temperatures of ≤0°C. Lastly, indicators of snowmaking capacity and hours are again not deemed as critical indicators, but rather as subsets of the critical indicator of probability of a snow depth of ≥30 centimetres with advanced snowmaking capacity. Without this level of snow depth (whether natural or artificial), outdoor events would not be possible.

These two indicators (probability of minimum daily temperatures of ≤0°C and the probability of snow depth ≥30 centimetres with advanced snowmaking capacity) were determined to provide the greatest insight into the climatic suitability of a city/region to host the OWG under current or future climate conditions. A previous host location was deemed climatically reliable if both indicators were achieved in 9 out of 10 winters (≥90% probability). If one or both indicators were achieved in less than 75% of winters, the location was considered unreliable for elite Olympic competitions. If one indicator was achieved ≥90% of winters and the other indicator was achieved only 75–89% of winters, or when both indicators were achieved 75–89% of winters, the location was classified as marginal/higher risk for the OWG. Each of the 19 former host locations was evaluated for its climatic suitability in the baseline or current climate normal period (1981–2010), as well as four future climate change scenarios (low- and high-emission scenarios for the 2050s and 2080s).

Data sources

Three sources of data (climate station data, climate change scenarios, and snowmaking operations model output) were used to operationalise the final set of two climate indicators at each of the 19 previous host cities/regions. Historic climate data were obtained from international (World Meteorological Organization) and national meteorological organisations (Deutscher Wetterdienst, Zentralanstalt für Meteorologie und Geodynamik, Meteo France, Environment Canada, Meteo Swiss, Hydrografisches Amt Bozen and Arpa Piemonte). Following standard practice in climatology, 30-year periods (‘climate normal') were defined for each host city, representing average (i.e. ‘normal') climate conditions over the time span. Daily data were obtained from 1981–2010, representing the baseline period for analysis. For each past host city/region, meteorological stations were chosen according to two criteria: (1) distance to host city or main competition sites, and (2) length and completeness of historical climate data record. Emphasis was placed on selecting meteorological stations within close proximity, due to the alpine nature of many host cities/regions and the need to accurately represent local climate. The climate station selected and the elevation at which the climatological analysis was conducted (i.e. station data were adjusted to the elevations of the majority of competitions using standard lapse rates) are identified in Table 1.

Table 1. Climate stations and altitudes analysed at each OWG host city/region.

Host city/region	Year hosted OWG	Altitude (masl)	Station name	Station altitude (masl)	Altitude analysed^a (masl)	Lat	Long
Chamonix	1924	1042	Chamonix	1042	1042	46.62	7.47
St. Moritz	1928, 1948	1798	Segl Maria	1798	1798	46.16	9.77
Lake Placid	1932, 1980	591	Lake Placid	591	591	44.25	−73.99
Garmisch-Partenkirchen	1936	719	Garmisch	719	719	47.48	11.06
Oslo	1952	1	Tryvasshogda^b	514	300	59.98	10.67
Cortina d'Ampezzo	1956	1211	Toblach	1219	1700	46.73	12.22
Squaw Valley	1960	1899	Tahoe City	1899	1899	39.17	−120.14
Innsbruck	1964, 1976	578	Innsbruck	578	1600	47.26	11.38
Grenoble	1968	384	Grenoble St. Geoirs	384	1600	46.15	6.10
Sapporo	1972	26	Sapporo^c	26	600	43.07	141.33
Sarajevo	1984	630	Sarajevo^d	630	1300	43.87	18.43
Calgary	1988	1048	Kananaskis Pocaterra	1610	1500	50.71	115.12
Albertville	1992	328	Chamonix	1042	2100	46.62	7.47
Lillehammer	1994	440	Venabu	930	200	61.65	10.11
Nagano	1998	419	Nagano^c	419	1500	36.67	138.20
Salt Lake City	2002	1288	Heber	1704	2300	40.49	−111.43
Turin	2006	240	Prerichard^e	1353	1900	45.08	6.72
Vancouver	2010	10	Whistler Alta Lake	658	800	50.13	−122.95
Sochi	2014	30	Krasnaya Polyana^d	568	900	43.68	40.20

^aThe altitude analysed is that of the finish line of alpine skiing events held in each location (the most recent games, where a location has hosted the OWG more than once).

^bMissing data between 1976–1998. Temperature was extrapolated with data from station Oslo Blindern (94 metres above sea level, 59.94, 10.72) with monthly temperature lapse rates derived from periods with data at both stations. Oslo Blindern does not record precipitation, therefore missing precipitation data at Tryvasshogda were filled with precipitation data from station Bjornholt (360 metres above sea level, 60.05, 10.69).

^cDaily snow depth data were not of sufficient quality for Nagano and Sapporo, therefore SkiSim snow model performance was tested against the average number of days with snow on the ground (snow cover) and monthly snowfall, which were available from the Japan Meteorological Agency for the baseline period.

^dSnow depth data were not available, therefore standard SkiSim parameter values (based on previous model applications, e.g. Steiger, 2010; Steiger & Abegg, 2013) were used.

^eFor this station, only data from 1992–2011 were available. No suitable nearby station was available to fill this missed data period, so the analysis could only be conducted with a shorter ‘normals’ period.

Climate change scenarios for temperature and precipitation (monthly resolution) for each of the 19 host city/regions were obtained from the Coupled Model Intercomparison Project phase 5 (CMIP-5) (World Climate Research Program, 2013), which uses 24 global climate models to prepare simulations for the IPCC Fifth Assessment Report (Taylor, Stouffer, & Meehl, 2012). Scenarios for two future periods were used in the analysis: 2041–2070 (referred to as the central decade of the ‘2050s’), representing climatic conditions in the middle of the twenty-first century, and 2071–2100 (referred to as the central decade of the ‘2080s’), representing late-twenty-first-century conditions. To consider the possible range of future climates, the IPCC's Representative Concentration Pathways (RCP) emission scenarios were used, with RCP 2.6 representative of a low GHG emission future and RCP 8.5 representative of a high-emission future. The range of projected temperature change during the winter months (December–January–February) at each of the former OWG host locations is presented in Table 2. Given that the global climate models used in CMIP-5 provide projections of climate variables at a spatial resolution of approximately 250 km, the climate change scenarios were downscaled to the climate station representing each former host location (Table 1) using the LARS stochastic weather generator (Semenov, 2013; Semenov & Barrow, 1997; Semenov & Stratonovitch, 2010). This weather generator produces synthetic weather time series on a daily basis, keeping the characteristics of the individual weather stations.

Table 2. Projected winter (December, January and February) warming at each OWG host city/region.

Host city/region	DJF^a Average Tmax 1981–2010	2050s		2080s
		Low emissions (RCP 2.6)	High emissions (RCP 8.5)	Low emissions (RCP 2.6)	High emissions (RCP 8.5)
Chamonix	4.5	+2.5°C	+3.0°C	+2.4°C	+4.5°C
St. Moritz	−0.1	+4.1°C	+4.8°C	+3.9°C	+6.3°C
Lake Placid	−0.9	+2.2°C	+3.9°C	+2.2°C	+6.3°C
Garmisch-Partenkirchen	4.6	+4.1°C	+4.8°C	+3.9°C	+6.3°C
Oslo	0.7	+1.3°C	+2.4°C	+1.2°C	+4.2°C
Cortina d'Ampezzo	3.5	+4.1°C	+4.9°C	+4.0°C	+6.5°C
Squaw Valley	6.0	+2.1°C	+3.1°C	+2.2°C	+4.7°C
Innsbruck	6.0	+4.1°C	+4.9°C	+4.0°C	+6.5°C
Grenoble	7.1	+2.6°C	+3.2°C	+2.5°C	+4.6°C
Sapporo	0.0	+2.0°C	+3.0°C	+1.9°C	+5.0°C
Sarajevo	4.9	+2.0°C	+2.7°C	+2.0°C	+4.4°C
Calgary	2.6	+1.1°C	+2.4°C	+1.3°C	+4.3°C
Albertville	7.3	+2.5°C	+3.0°C	+2.4°C	+4.5°C
Lillehammer	−2.5	+1.3°C	+2.4°C	+1.2°C	+4.2°C
Nagano	4.3	+4.2°C	+5.3°C	+4.1°C	+6.9°C
Salt Lake City	6.1	−0.8°C	+0.9°C	−0.7°C	+2.8°C
Turin	9.7	+2.5°C	+3.0°C	+2.4°C	+4.5°C
Vancouver	5.9	+1.1°C	+2.2°C	+1.3°C	+4.1°C
Sochi	8.5	+3.6°C	+4.2°C	+3.6°C	+5.8°C

^aDecember, January, February.

Snowmaking and snow depth data were produced using SkiSim2 (Steiger, 2010), a ski operations simulation model, that incorporates both natural snowfall and advanced snowmaking capacities. SkiSim (1.0 and 2.0) has been used extensively to investigate the potential impact of climate change on the ski industry in North America (Dawson & Scott, 2013; Scott et al., 2003, 2006, 2008; Scott & Steiger, 2013) and Europe (Steiger, 2010; Steiger & Abegg, 2013; Steiger & Stötter, 2013). SkiSim 2 uses a degree-day-based snow model, with an integrated snowmaking module that allows for an analysis of the impact of climate change on ski operations, including snowmaking potential, impact of altitude on snow conditions (in 100 metre intervals) and snow–rain precipitation classification (for details, see Steiger, 2010). Key climate data inputs for SkiSim2 include daily temperature and precipitation. These variables were obtained from the historical climate record of the station nearest to each former host location and were extrapolated from the altitude of the particular climate station to the host city and main competition site altitudes, using a standard winter lapse rate of 0.4°C/100 metres (see Kunz, Scherrer, Liniger, & Appenzeller, 2007 for Switzerland, or Steiger & Abegg, 2013 for Austria) and 3%/100 metres for precipitation (Steiger, 2010). For some Olympic host cities, snow-based competitions were held at multiple altitudes (e.g. the start and a finish line for alpine skiing across several hundred metres of vertical terrain). In such cases, the highest finish line was used to represent the altitude of the competition. This choice is conservative with respect to the potential climate change risk, but reflects the potential for host cities to develop infrastructure and hold events at higher altitudes as an adaptation to poor snow conditions at a lower altitude.

Results

Figure 1 illustrates the probability that minimum daily temperatures in February were ≤0°C at each of the 19 former host cities/regions during the baseline period (1981–2010) and under the low- and high-emission scenarios for the 2050s and 2080s. In the baseline climate (1981–2010), all locations have ≥90% probability of average daily minimum temperatures that would remain below 0°C. In the 2050s, between 8 (low-emission) and 9 (high-emission) scenarios no longer fulfil this indicator of climate suitability. Locations such as Garmisch-Partenkirchen (Germany), Vancouver (Canada) and Sochi (Russia) are projected to achieve threshold in less than 75% of winters under the high-emission scenario and therefore are considered unreliable for hosting the outdoor sports programme of the OWG.

Figure 1. Probability of average daily minimum temperature in February ≤0°C.

In the 2080s, the number of former host locations able to achieve this indicator ≥90% of the time declines even further to between 9 (low-emission scenario) and 13 (high-emission scenario). Under the warmer high-emission scenario, three locations are projected to achieve this threshold less than 50% of winters (Garmisch-Partenkirchen, Germany; Vancouver, Canada and Sochi, Russia), while several others have a probability less than 75% (Innsbruck, Austria; Squaw Valley, USA; Chamonix, France; Grenoble, France; Sarajevo, Bosnia-Herzegovina and Oslo, Norway).

The most critical factor for snow-based competitions, such as alpine, nordic and freestyle skiing, is a sufficiently deep snow base. The minimum snow base for alpine skiing is 30 centimetres, which is sufficient for smooth slopes on alpine meadows and track setting for classic nordic skiing on smooth trails. For alpine slopes with rocky surfaces, much greater snow depths of up to 1 metre are required for safe operations. This analysis utilised an optimistic requirement of only 30 centimetre snow depth to account for slope/trial grooming (smoothening) as a possible adaptation strategy. Figure 2 illustrates the probability that a snow depth of ≥30 centimetres, with advanced snowmaking capacity, could be achieved at each of the 19 former host cities/regions. Similar to the first indicator, in the baseline climate (1981–2010), all locations achieved this operational threshold in 90% of winters. Under projected climate change scenarios, some locations cannot guarantee a sufficient snow depth despite advanced snowmaking, as temperatures become too warm to produce enough snow. In the 2050s, three former host locations are projected to no longer be able to produce a sufficient snow base by the beginning of February in ≥90% of winters (Sochi, Russia; Squaw Valley, USA and Garmisch-Partenkirchen, Germany). The number of locations that can no longer fulfil this indicator of climate suitability for snow-based sports competitions remains the same under the low-emission scenario for the 2080s, but increased substantially to 11 under the high-emission scenario. Each of these 11 locations are projected to achieve this threshold of snow base in less than 75% of winters under the high-emission scenario, and are therefore considered unreliable for hosting the outdoor sports programme of the OWG.

Figure 2. Probability of snow depth with snowmaking ≥30 centimetres on 1 February.

As previously noted, this analysis was conducted at the altitude of the finish line where skiing events were previously held at each host location. A potential adaptation would be to relocate competitions to higher altitudes, where this is physically feasible (i.e. ski slopes at higher altitudes with a sufficient vertical drop for elite Olympic competitions). This adaptation was not considered here because the new elevation of the alpine slopes within an acceptable distance to other games venues cannot be anticipated.

The two indicators were compared to assess the overall climatic suitability of past host locations to reliably host future outdoor OWG competitions under changed climate conditions. Figure 3 presents the probability that each of the 19 former host locations is projected to achieve both indicators in the 2050s (Figure 3) and 2080s (Figure 4) under low- and high-emission scenarios. Locations that were able to achieve both indicators in ≥90% of winters and classified as climate reliable are found in the upper right shading in Figures 3 and 4. In the baseline period (1981–2100), all former host locations were classified as climatically reliable to host the outdoor competitions of the OWG. Despite some concerns about the climate conditions at the recent host locations of Vancouver (Canada) and Sochi (Russia), both were assessed to be climatically reliable using the indicators developed for this study, supporting the IOC's decision to award the games to these locations. A comparison of Figures 3 and 4 reveals that the number of past host cities/regions falling outside of the reliable climate zone increases in both the 2050s and 2080s, particularly under higher emission scenarios. More locations become climatically marginal/higher risk in the 2050s because of minimum daily temperatures exceeding 0°C. The number of locations with one variable being achieved in <75% of winters (white area in Figures 3 and 4) are classified as not climatically reliable to host outdoor sports programme of the OWG, which visibly increases in the higher emission scenario for the 2080s. This is because of the inability to produce sufficient snow, even with advanced snowmaking.

Figure 3. Probability of former host locations remaining climatically suitable for the OWG in the 2050s under climate change.

Figure 4. Probability of former host locations remaining climatically suitable for the OWG in the 2080s under climate change.

Table 3 provides a summary of the climate suitability rating for each of the 19 former host locations. In the baseline period (1981–2010), all locations were rated as climatically reliable. In the 2050s, the number of climate reliable locations decreased to 11 in the low-emission scenario and 10 in the high-emission scenario. The impact of projected climate change is far greater in the late-twenty-first century, with the differential impact of the two GHG emission pathways particularly notable. In the low-emission scenarios for the 2080s, 10 of the former host locations would still have reliable climate conditions. However, if the high-emission scenarios were realised, it is projected that less than one-third (six in total) of former host locations would remain climatically suitable for the games. Interestingly, of the remaining climate reliable locations under the highest emissions scenario, all global regions that have previously hosted Olympics are represented (Western North America – Calgary, Canada and Salt Lake City, USA; European Alps – St. Moritz, Switzerland, Cortina d'Ampezzo, Italy and Albertville, France; East Asia – Sapporo, Japan).

Table 3. Climate suitability rating of host cities/regions for future OWG.

Host city/region	2050s		2080s
	RCP 2.6 (low emission)	RCP 8.5 (high emission)	RCP 2.6 (low emission)	RCP 8.5 (high emission)
Albertville (F)	Reliable	Reliable	Reliable	Reliable
Calgary (CDN)	Reliable	Reliable	Reliable	Reliable
Chamonix (F)	Marginal–high risk	Marginal–high risk	Marginal–high risk	Not reliable
Cortina d'Ampezzo (I)	Reliable	Reliable	Reliable	Reliable
Garmisch-Partenkirchen (GER)	Not reliable	Not reliable	Not reliable	Not reliable
Grenoble (F)	Marginal–high risk	Marginal–high risk	Marginal–high risk	Not reliable
Innsbruck (A)	Reliable	Marginal–high risk	Marginal–high risk	Not reliable
Lake Placid (USA)	Reliable	Reliable	Reliable	Marginal–high risk
Lillehammer (N)	Reliable	Reliable	Reliable	Marginal–high risk
Nagano (J)	Reliable	Reliable	Reliable	Not reliable
Oslo (N)	Marginal–high risk	Marginal–high risk	Marginal–high risk	Not reliable
Salt Lake City (USA)	Reliable	Reliable	Reliable	Reliable
Sapporo (J)	Reliable	Reliable	Reliable	Reliable
Sarajevo (BIH)	Marginal–high risk	Marginal–high risk	Marginal–high risk	Not reliable
Sochi (RUS)	Not reliable	Not reliable	Not reliable	Not reliable
Squaw Valley (USA)	Marginal–high risk	Not reliable	Not reliable	Not reliable
St. Moritz (CH)	Reliable	Reliable	Reliable	Reliable
Turin (I)	Reliable	Reliable	Reliable	Not reliable
Vancouver (CDN)	Marginal–high risk	Not reliable	Not reliable	Not reliable

Conclusion

The confluence of the 22nd OWG and the release of the final two reports of the 5th IPCC Assessment in early 2014 provide an important opportunity to consider the long-term implications of global climate change for the world of sport and the collective cultural global heritage symbolised by the Olympic Movement. It is evident from the results of this analysis that the many climatic adaptations employed by recent OWG Organising Committees to manage the risks of weather-related disruption of outdoor competitions begin to reach the limits of effectiveness at some locations under projected climate change. The capacity of snowmaking to ensure adequate snow conditions (even a highly optimistic 30 centimetre snow base) for ski competitions is a particular source of climate change vulnerability for future OWG. As a result, the findings indicate that projected climate change would adversely impact the capacity of approximately half of the former OWG host cities/regions to host the games by mid-century.

This finding suggests that the IOC choice of Vancouver and Sochi to host recent games may prove very fortuitous, as the climatic capacity of these locations to host the games by mid-century is degraded, respectively becoming high risk and not reliable under even the low-emission scenario. The differential vulnerability of former host cities/regions also has implications for potential bids to host future OWG. For example, Munich, together with sports venues at Garmisch-Partenkirchen, has considered a potential bid for the 2022 games, though this analysis indicates that Garmisch-Partenkirchen is not considered to be climatically reliable by the 2050s. With the climatic capacity of this region to host the OWG degraded by mid-century, climate change places greater impetus on a bid to host the games over the next two decades.

Recognising that the format and technologies supporting winter sports in the later decades of this century will clearly be different from today, they will nonetheless continue to be founded on snow and ice as they have for the past 100 years. It is clear that the cultural legacy of the world's celebration of winter sport is at increased risk if the warmer climate scenarios of the late-twenty-first century occur. With additional warming continuing in the early-twenty-second century under a high-emission pathway, the number of former OWG locations capable of hosting the games would continue to decline. In other words, in a substantially warmer world, celebrating the second centennial of the OWG in 2124 would become increasingly challenging.

This analysis provides some initial insight into the implications of climate change for the long-term future of the OWG, but raises several other research questions. With fewer traditional winter sports regions climatically able to host OWG in a warmer world, will the unifying cultural benefits be degraded? What are the broader implications for participation in winter sport and eventually participation by countries in the OWG? What other regions of the world that have never hosted the OWG could replace the past host locations that are anticipated to be no longer climatically suitable? Will new winter sports powers and winter tourism regions develop in emerging markets? What are the long-term implications for the scale of the OWG as a mega-tourism and media event? Will the IOC need to move away from recent decision to award the OWG to large/resort cities with nearby mountainous areas capable of providing alpine skiing venues to more traditional, but smaller, alpine cities? Are these smaller cities capable of once again hosting the OWG given the tremendous growth in the number of athletes and spectators? Are there other adaptation technologies that could overcome the climate change vulnerabilities identified in this study? For example, could the development of truly artificial snow that is not temperature dependent (as opposed to machine-made snow that physically responds to temperature and solar radiation like natural snow) resolve the central remaining vulnerability of adequate snow for ski competitions?

The IOC has officially recognised the environment as the third integral dimension of Olympism, alongside sport and culture. Since then, the OWG has shown leadership in championing new technologies of the low-carbon economy, with no other major sporting event being able to demonstrate carbon-neutrality for over a decade. This leadership should be commended as it may foster other climate compatible mega-events and tourism. Nonetheless, much more will be required of all nations if the goal of the international community is to limit the warming of global average temperatures to less than +2°C over pre-industrial times through a rapid transition to a low-carbon economy.