Is Tourism Sustainable
The current issue and full text archive of this journal is available at www.emeraldinsight.com/1755-4217.htm
Is tourism with a low impact on climate possible?
Centre for Environmental Strategy, University of Surrey, Guildford, UK
Purpose – The purpose of this paper is to examine the impact of a range of different travel and tourism options, and quantiﬁes the carbon-dioxide emissions resulting from international vacations, breaking down emissions categories into those resulting from transport, accommodation and recreation. Design/methodology/approach – The paper uses summary data to review a range of possible vacation scenarios and examines their relative carbon-dioxide emissions in order to compare the relative climatic impact of different forms of tourism and vacation options. Findings – The paper concludes that intercontinental ﬂights and cruise ship travel are particularly carbon-intensive, which suggests that these two forms of tourism will be particularly vulnerable to any policy initiative to curb or price carbon emissions. Ends by considering whether climatically responsible international tourism is possible, and outlines some low-carbon options. Originality/value – The paper relates data on carbon emissions to the implications for tourism arising from climate change. Keywords Climatology, Tourism, Global warming Paper type General review
The nature of the challenge While much of the discussion about the climate change impacts of international travel focuses upon international aviation, all forms of transport result in the emissions of carbon dioxide. Increased carbon dioxide emissions also result from the accommodation facilities used by international tourists and the recreational activities they engage in. Scenario analysis of a range of possible international tourist trips shows that the greatest factor determining the carbon dioxide emissions of a trip is not the mode of transport selected but the distance travelled. The choice of accommodation can also have a signiﬁcant impact on the total carbon dioxide emissions resulting from the trip, with low-budget options the best choice. Non-air-based recreational activities engaged in by tourists were generally relatively small in their contribution to carbon dioxide emissions. Scenario analysis shows that carbon-neutral international travel is possible if consideration is given to minimising the carbon impact – travel distances need to be modest, people must use low carbon travel modes such as cycling, and they should stay in low-budget accommodation. Introduction International tourism clearly comes at a signiﬁcant cost to the environment, with the carbon dioxide emissions of air travel in particular receiving considerable media attention. However, air travel is not alone in having an environmental impact; other modes of transport also produce carbon dioxide emissions and other environmental effects, and the impacts from international tourism extend well beyond transport to both the accommodation and recreational activities used by international tourists.
Worldwide Hospitality and Tourism Themes Vol. 1 No. 3, 2009 pp. 274-287 q Emerald Group Publishing Limited 1755-4217 DOI 10.1108/17554210910980611
In 2006 there were approximately 846 million international tourist arrivals globally, with tourist arrivals being deﬁned as visitors who stay at least one night in the country visited (World Tourism Organisation, 2007). This was a 5.4 per cent increase on the previous year, with just over half (54.4 per cent) of these tourist arrivals being in Europe, 20 per cent in Asia and the Paciﬁc, and 16 per cent in the Americas. Africa and the Middle East each had less than 5 per cent of international tourist arrivals (World Tourism Organisation, 2007). Each component of international tourism, namely transport, accommodation and the recreational activities carried out by the tourists at their destination results in carbon dioxide emissions. While there are many carbon-calculators available on the internet, few of these provide details of the data they are based upon or how the calculations are made, making it difﬁcult to verify the relative impacts of each component of a vacation and thus how to minimise the overall impact. This paper will examine the literature on the environmental impacts of international tourism, and seek to quantify the carbon dioxide emissions resulting from international vacations, breaking down emissions categories into those resulting from transport, accommodation and recreation. Using these data, a range of possible vacation scenarios will then be examined for their relative carbon dioxide emissions in order to compare the relative climatic impact of different forms of tourism and vacation options. The paper ﬁnishes by asking whether climatically responsible international tourism is possible. The environmental impacts of international travel: transport Many internet based carbon-calculators that allow the assessment of overseas travel focus only upon the transport emissions while ignoring other possible sources of carbon dioxide. Air travel accounts for approximately 46 per cent of international tourist arrivals, with road-based transport accounting for 43 per cent, water-based transport accounting for 7 per cent and rail-based transport accounting for 4 per cent (World Tourism Organisation, 2007). Air travel Air travel for tourism has particularly signiﬁcant climate change implications even though some estimates suggest that international aviation currently accounts for just 1 per cent of total anthropogenic emissions of carbon dioxide (Olsthoorn, 2001). While the carbon dioxide emissions resulting from travel are relatively low in absolute terms compared to other anthropogenic emissions, and a signiﬁcant portion of this international aviation is for business purposes, the aviation industry has experienced very rapid growth in recent years. Total passenger numbers travelling by air are forecast to increase at an average annual rate of 4.6 per cent between 2005 and 2025 (Environmental Unit, 2007). According to the Tyndall Centre for Climate Change Research (2005), modelling results suggest that under a contract and converge policy aiming to achieve an atmospheric carbon dioxide concentration of 450 ppmv, if current growth rates continue aviation emissions by 2050 would exceed the entire permissible emissions for the UK as a whole, and looking at European Union as a whole, aviation emissions would make up 80 per cent of the permissible emissions. Even more seriously, carbon dioxide emissions make up only part of the contribution of aviation to climate change, as global warming
Tourism with a low impact on climate 275
from aviation also results from contrails, cirrus cloud formation and the emission of other greenhouse gases (Tyndall Centre for Climate Change Research, 2005). While there is considerable scientiﬁc uncertainly associated with such effects, such factors may mean that the full climate change effects of aviation may exceed that of the carbon dioxide emissions alone by a factor of 2-3.5 times. Owing to this great uncertainty, these other factors will not be considered in this paper, with the focus of climatic impacts to be on carbon dioxide emissions even though this is almost certainly a signiﬁcant underestimate of the full climatic impact of travel. The UK Department for Environment Food and Rural Affairs (DEFRA) estimates that a short-haul international air passenger typically is responsible for the emission of 130 g of carbon dioxide per kilometre travelled, based upon an average 0.65 load factor for the aircraft and an average journey of 500 km (DEFRA, 2007a). DEFRA estimates that a long-haul international air passenger is typically responsible for the emission of 105 g of carbon dioxide per kilometre travelled, based upon a load factor of 0.797 and an average journey of 3,500 km (DEFRA, 2007a). Road travel Road travel, most frequently by car, also has signiﬁcant climate change implications. Data from DEFRA suggests that the average car emits 205.9 g of carbon dioxide per kilometre travelled (DEFRA, 2007a). The average carbon dioxide emissions of new cars are declining over time as cars become more efﬁcient, with average emissions of new cars falling from 186 to 164 g of carbon dioxide per passenger kilometer between 1995 and 2003 (European Environment Agency, 2006). Even with this gradual reduction in emissions a single person driving to a vacation destination will be responsible for the emission of a signiﬁcantly larger amount of carbon dioxide than if they had ﬂown to the same destination, while for a family of four doing the same journey, the car journey will result in much lower emissions per person. Tourists, however, will tend to travel much further by air than by car, thus on trip by trip basis, carbon dioxide emissions from air travel for international tourism greatly exceed emissions for car travel. Bus travel is estimated by DEFRA (2007b) to result in emissions of 89 g/passenger km. Rail travel Carbon dioxide emissions from train journeys are affected by the power source of the trains and thus will be region speciﬁc. Trains powered by nuclear or hydro generated electricity will have much lower emissions than trains powered by coal generated electricity and both will differ from diesel powered trains. Transport Watch UK (2007) provides an estimate of 55 g of carbon dioxide per passenger kilometre for a high-speed intercity rail (travelling at 200 km/h) and 86 g of carbon dioxide per passenger kilometre for a TGV train (300 km/h), in both cases assuming a 30 per cent load factor. An alternative estimate of carbon dioxide emissions from train journeys is provided by the Association of Train Operating Companies (2007) which suggest that rail travel ranges from 54 g of carbon dioxide per passenger per kilometre for an electric train to 74 g of carbon dioxide per passenger per kilometre for a diesel train. DEFRA (2007b) also provides an estimate for inter-city trains of 60 g of carbon dioxide per passenger kilometre but does not specify whether this refers to electric or diesel trains.
Water based travel When calculating carbon dioxide emissions from international ferry-based travel, it is necessary to make decision on how to split allocations between car passengers, foot passengers and freight. Carbon Tracking (2008) suggest allocating emissions on the basis of their relative contribution to ferry companies revenue, with foot passengers providing 5 per cent of revenue, cars 40 per cent and freight 41 per cent. On this basis Carbon Tracking calculates that each car-kilometre of ferry travel generates 1,130 g of carbon dioxide, while each foot passenger-kilometre generates 141 g of carbon dioxide (Carbon Tracking, 2008). Cruise liner travel is also particularly carbon intensive. The Carnival Corporation, a global cruise ship operator with 85 ships, calculates in its 2007 Annual Environmental Management Report that 342 g of carbon dioxide per passenger kilometre were emitted from its ships during 2007 (Carnival Corporation, 2008). Non-fossil fuel based travel The lowest carbon dioxide emitting option for travelling signiﬁcant distances is cycling as this form of transport makes no direct use of fossil fuels. However, even cycling activity will cause additional carbon dioxide emissions compared to staying at home as the cyclist themselves will expend signiﬁcant extra energy cycling long distances each day, and thus unless the cyclist wishes to lose weight they will need to increase their food intake. A cyclist weighing 70 kg who cycles for around 5 h/day covering 100 km would expend approximately 3,000 additional kilocalories compared to being inactive at home (NHS Direct, 2008). If these additional calories were obtained by eating a high-starch food such as bread, this equates to eating 1.2 kg of bread per day in addition to the person’s normal diet. Production of such a quantity of bread ´ would result in the emission of approximately 912 g of carbon dioxide (Wallen et al., 2004). Thus, a cycle journey results, very approximately, in 9 g of carbon dioxide emitted per kilometre cycled. Table I summarises the estimations of carbon dioxide emissions from the different modes of transport. With the exception of cycling, a new car with four occupants clearly has the lowest emissions per kilometre travelled, while the train is the best option for one person or a couple. The environmental impacts of international travel: accommodation In most cases international tourism will entail a trip many times greater than a person’s regular daily travel patterns. Compared with the journey to a different country, regular daily travel (such as the journey to and from work) is relatively insigniﬁcant, and for many people will be comparable to the distance travelled to the airport or international rail terminal, or the daily travel while abroad doing recreational activities. Thus, the international journey is in addition to regular travel and causes additional carbon dioxide emissions than would otherwise have been made when considering the impact of the trip. However, with tourism accommodation things are less straightforward. A person will generally sleep in a bed, shower, eat, and use a range of domestic appliances (either directly or indirectly) whether they stay at home or in a hotel so it is really the marginal change in carbon dioxide emissions that are important when considering the net impact of the trip.
Tourism with a low impact on climate 277
Mode of transport Air travel Short-haul ﬂight Long-haul ﬂight Average car – single occupant New car – four occupants Bus Cycling Rail Intercity rail TGV Electric train Diesel train Water Single occupant car passenger on ferry Foot passenger on ferry Cruise liner passenger
Grams/CO2 per passenger kilometre 130 105 206 41 89 9 55 86 54 74 1,130 141 342
Source(s) DEFRA (2007a) DEFRA (2007a) DEFRA (2007a) European Environment Agency (2006) DEFRA (2007a) ´ NHS Direct (2008) and Wallen et al. (2004) Transport Watch UK (2007) Transport Watch UK (2007) Association of Train Operating Companies (2007) Association of Train Operating Companies (2007) Carbon Tracking (2008) Carbon Tracking (2008) Carnival Corporation (2008)
Table I. Summary of carbon dioxide emissions per passenger kilometre of different transport options
This net impact will result from the traveller’s home continuing to be maintained in their absence, perhaps with a signiﬁcant reduction in energy consumption depending upon its location, while at the same time the traveller may use facilities that they would not normally use. The largest environmental impact resulting from accommodation probably results from the need to build and maintain facilities for tourists that are in addition to everyday residential accommodation, which will heated (or cooled) and generally maintained whether or not they are fully occupied. As with transport, per capita carbon dioxide emissions will be minimised when occupancy rates are high. Other environmental impacts will result from the location in which accommodation facilities are built. Ski resorts, for example, are built in ecologically fragile mountain regions. They are costly to provide with services such as water or sewerage, due to their high altitude and extreme weather conditions. Perhaps, because of the difﬁculty in measuring the marginal impact of tourist accommodation relatively little has been written about this in the environmental literature. The CarbonNeutral Company (2008), a carbon off-setting company states that use of a UK hotel room for one night results in 34.32 kg of carbon dioxide emissions while an international hotel room results in 33.87 kg of carbon dioxide emissions. However, they do not explain how these ﬁgures were derived. ¨ Gossling et al. (2005) estimate the carbon dioxide emissions per night per bed resulting from several accommodation options, ranging from 20.6 kg of carbon dioxide per person per night for a hotel stay to 4.0 kg of carbon dioxide for a campsite. These estimates are based on the average energy consumption of each type of facility together with numbers of occupants. Thus, the actual carbon dioxide emissions will vary signiﬁcantly depending ¨ upon how the energy is produced, with Gossling et al. (2005) noting that they contain a moderate degree of uncertainty due to the limited data available.
Estimates of accommodation related energy use are provided by Becken et al. (2001) for New Zealand. Their estimates are given in mega-joules (MJ) per visitor night, but converted using a global average of 158.4 g of carbon dioxide per MJ (Schafer and Victor, 1999), their estimates equate to 24.6 kg of carbon dioxide per hotel night, 6.2 kg/night for a backpackers and 4.0 kg/night for camping (Becken et al., 2001). ¨ As with the estimate of the CarbonNeutral Company, the Gossling et al. (2005) and the Becken et al. (2001) estimates appear to refer to total emissions rather than the additional emissions resulting from the hotel stay compared to staying at home. Using data published by DEFRA (2007a), an average UK resident emits 2,687 kg of carbon dioxide per year from their home and their appliance use, thus emitting an average of 7.4 kg of carbon dioxide per day. Being absent from home will result in a reduction in such energy usage, although not usually to zero since many appliances are left in standby mode (which still consumes energy), and heating systems may be left on in order to prevent pipes freezing. Given this residual use, a temporarily unoccupied house might consume approximately 25 per cent of its normal energy consumption, with the amount being higher in winter and lower in summer. For the UK this residual energy usage equates to about 1.9 kg for carbon dioxide per night, meaning that camping would result in a net reduction in carbon dioxide emissions, and staying in a backpacker hostel would be nearly carbon dioxide neutral due to the savings being made at home. One further accommodation option available to international tourists concerned about their environmental impacts are “home stays”, where the travellers stay with an ordinary family in their home rather than in facilities speciﬁcally constructed (or adapted) and maintained for tourists. The easiest way solution is for travellers to stay with existing friends, however, there are also a number of websites which facilitate travellers staying with local people in the countries they visit by “couch-surﬁng” – sleeping on the couch or in the spare room of a local resident of the area they are visiting. Two of the most popular such site are www.couchsurﬁng. com and www.hospitalityclub.org, both of which function in a similar manner. These online forums are open for anyone to join by registering and creating for themselves a personal proﬁle. In advance of travelling the traveller will log on to the couch-surﬁng web site and search for people in the place they plan to visit before emailing a selection of the fellow couch-surfers in the place they plan to visit, explaining their travel plans and asking if they can visit or stay. Usually, the host will spend time with their guests, often having dinner together before their guests sleep on the sofa or spare bed. Typical accommodation can range from very basic student ﬂats through to rambling country houses with the quality of accommodation obviously dependent on the personal situation of the couch-surﬁng host. Under the rules of both the couch-surﬁng.com and hospitalityclub.org forums, no one is under any obligation to accept any guests they do not want, and no money is allowed to change hands between guest and host, with forum users generally expected to both host and be hosted. Safety is ensured by a feedback system and by examining a potential guest’s or host’s network of contact and friends within the forum. Many cities, like London or New York, have more than 1,000 couch-surfer hosts, other cities, like Gaza, Dili, Pyongyang, Kirkuk, Fallujah, Port Moresby or Honiara have less than ten, although these totals are constantly changing as the activity becomes more popular. Both networks are thoroughly international even if there are more couch-surfers in places with higher levels of internet access.
Tourism with a low impact on climate 279
In carbon dioxide emission terms couch-surﬁng results in little if any additional carbon emissions; as no dedicated accommodation is being provided for the tourists. Any additional emissions resulting from hosting couch-surﬁng guests should be offset by a reduction in the guest’s own home being temporarily unoccupied. In addition to the environmental beneﬁts, the tourists themselves generally save money, of course, by avoiding accommodation costs. Perhaps, the greatest beneﬁt of couch-surﬁng, however, is that many participants report that it allows much more meaningful contact and understanding with the local community than is normally possible when visiting a foreign country as a tourist. Table II summarise the net carbon dioxide emissions for the different accommodation options. Net emissions refers to the change in carbon dioxide emissions of a stay in a hotel compared staying at home and assumes that the maintenance of the traveller’s house in the UK will result in the emission of 1.8 kg of carbon dioxide, a 75 per cent reduction from the average emissions of 7.4 kg/day – thus saving 5.6 kg of carbon dioxide per day to off-set against the emissions from the vacation accommodation. The environmental impacts of international travel: recreation Becken and Simmons (2002) calculated the average energy use per tourist for different tourism-related recreational activities and attractions in New Zealand. They divided activities into three categories; attractions, entertainment, and activities. Attractions included things such as museums and art galleries, amusement parks, breweries and natural attractions, such as hot pools. The attractions they considered ranged between 95 g of carbon dioxide per tourist for a visitor centre, and 27.6 kg of carbon dioxide for a wine trail (when converted from MJ values using global averages of grams of carbon dioxide per MJ of energy). Entertainment included cinema, live theatre, night clubs, casinos and shopping and ranged between 127 g of carbon dioxide for a tourist shop and 6.2 kg of carbon dioxide for an entertainment complex. Activities included things such as sky diving, heliskiing, diving, sailing, mountain biking, rock climbing, ﬁshing and guided walking, which ranged from 95 g of carbon dioxide for a horse riding operator and 459.8 kg of carbon dioxide for a ﬁshing operator offering helicopter trips (Becken and Simmons, 2002). The average attraction was responsible for 950 g of carbon dioxide per tourist, while for entertainment the average was 1,426 g, and for activities the average was 15.2 kg of carbon dioxide. With regard to activities, the precise nature of the activity had a very signiﬁcant impact on carbon dioxide emissions. Air-based activities resulted in 70 kg
Net carbon dioxide emissions (kilograms per person-stay per night) for a UK resident 28.3 15.1 19.1 20.8 0.55 21.6 0
Accommodation Hotel room Hotel room Hotel room Hotel room (average) Backpackers hostel Camping Couch-surﬁng
Source CarbonNeutral Company (2008) ¨ Gossling et al. (2005) Becken et al. (2001) Becken et al. (2001) ¨ Gossling et al. (2005) and Becken et al. (2001)
Table II. Net carbon dioxide emissions resulting from different forms of tourist accommodation
of carbon dioxide per tourist while adventure-based activities resulted in a much more modest 5.5 kg of carbon dioxide on average. The carbon dioxide estimates for activities derived by Becken and Simmons (2002) were based on a survey of tourism related operators in New Zealand and thus only included organised activities or maintained venues. A mountain biking operator, for example, will have to run some form of administrative ofﬁce and may transport bikes and riders to and from the start of an organised ride, with each of these resulting in carbon dioxide emissions. However, the same activity carried out by a group of individuals on their own in some situations would result in no direct carbon dioxide emissions if they were to go from and return to their hotel on their own. Similarly, a group of tourists visiting some wineries directly on route to their vacation destination would result in far fewer emissions than identiﬁed by Becken and Simmons (2002) for a winery tour run by a tourist operator. See Table III for a summary of estimates of carbon dioxide resulting from international tourist recreation. Emissions scenarios for international vacations Using the previously calculated data it is possible to calculate the net change in carbon dioxide emissions of a range of vacation options compared to staying at home in order to identify the relative signiﬁcance of the travel mode choice, accommodation choice and activity choice. To permit comparison, some basic assumptions need to be made. Each scenario will be for two people travelling together with the point of origin being suburban London. For all scenarios except Scenario 1, it is assumed that they will consume a similar amount of food to normal. Local transport usage at the destinations for self organised activities is assumed to approximately balance local transport usage had they stayed at home, such as commuting (Table IV). Discussion A typical UK resident staying at home, as noted in the accommodation section, will produce on average 7.4 kg of carbon dioxide each day from their house and its appliances. The average resident travels 13.39 km to work, thus doing a round trip of 26.78 km each weekday, with the overwhelming majority of such trips being by car (National Statistics, 2007). The home-based emissions were taken into consideration
Highest carbon dioxide emission (kg) per tourist participant/visitor from survey sample 27.6 6.2 459.8 Lowest carbon dioxide emission (kg) per tourist participant/visitor from survey sample 0.095 0.127 0.095
Tourism with a low impact on climate 281
Activity Attractions Entertainment Activities (all) Air based activities Adventure based activities
Average carbon dioxide emissions (kg) per tourist participant/visitor 0.950 1.426 15.2 70 5.5
Source: Becken and Simmons (2002)
Table III. Carbon dioxide emissions resulting from tourist recreation
Scenario Description 83 0 0 – 83 2.6
Table IV. Transport scenarios showing the carbon dioxide emissions from a range of tourist trip options compared to the carbon dioxide emissions resulting from staying at home
Carbon dioxide emissions in kilograms for two people Travel Accommodation Recreation Cruise Total Additional carbon dioxide emissions per traveler per day (kg) Comments Emissions from train, ferry and cycling each approximately equal in contribution to total emissions 184 7 15 – 206 17 333 250 15 – 598 50 Approximately one third of travel emissions due to use of car ferry (continued)
Return electric train from London to Dover, foot passage by ferry to Calais, and then cycling 1,600 km over 16 days through northern France, Belgium, The Netherlands, western Germany, Luxembourg and back through northern France to Dover before returning to London; couch-surﬁng for accommodation each night Return TGV train from London to Amsterdam, staying in a backpackers hostel for six nights and visiting an average of eight attractions around the city over the course of the week Return car journey (average car) from London to Amsterdam via ferry crossing, staying in a hotel for six nights and visiting an average of eight attractions in around the city over the course of the week
Carbon dioxide emissions in kilograms for two people Additional`al
Travel Accommodation Recreation Cruise Total
when calculating the accommodation emissions and it was assumed that the tourist would do local travel at their destination approximately comparable to what they would have done at home, hence the calculated emissions for each scenario are in addition to existing personal emissions. The scale of these existing personal emissions, which are slightly in excess of 10 kg/person/day, are useful for providing a benchmark to compare against the resulting emissions for each travel scenario. Scenario 1 clearly had the lowest level of carbon dioxide emission – 83 kg in total and 2.6 kg/person per day, which itself was probably a slight overestimate since local travel the destinations covered on the trip would almost certainly be carried out by bicycle and hence would be lower than an average person’s daily regular travel back in the UK, possibly even cancelling out the 2.6 kg of emissions resulting from the train, ferry and cycling. However, even without allowing for this probable overestimation, 2.6 kg of carbon dioxide per person per day is a modest increase on the benchmark for comparison of 10 kg/person/day. Scenarios 2 and 3 both describe a very similar trip, Scenario 2 being the more luxurious ﬂying to Amsterdam and staying in hotel and Scenario 3 being the budget version – taking the train to Amsterdam and staying in a backpacker hostel. The majority of the difference in the two scenarios, however, is due not to driving rather than taking the train, but due to staying in a hotel rather than a backpacker hostel, showing how with relatively short trips, the choice of accommodation has a much greater impact on carbon dioxide emissions than the choice of transport. Scenario 2 had a per person carbon dioxide emission increase of 17 kg/day, thus resulting in a nearly tripling of emissions, while Scenario 3 resulted in additional emissions of 50 kg/person/day, a six-fold increase. ` Scenarios 4 and 5 also described a similar trip, namely a week’s skiing at Val d’Isere. In both scenarios hotel accommodation was assumed but in Scenario 4 the primary form of transport was TGV train while in Scenario 5 air transport was used. Both of these scenarios resulted in very similar carbon emissions, suggesting that it is not so much a case that ﬂying is signiﬁcantly worse on a per kilometer basis than driving or taking the train, but rather, that ﬂying is worse for the environment because it facilitates much longer distance travel. The per capita daily additional emissions for Scenario 4 were 59 kg while for Scenario 5 they were 64 kg. Scenario 6, a week-long Mediterranean cruise, resulted in three times the additional emissions of the Alpine skiing trip and per traveller per day emissions were 18 times the benchmark ﬁgure for staying at home. While the greater ﬂight distance had some impact, the major reason for the signiﬁcant increase in carbon dioxide emissions compared to skiing was due to the cruise itself, which is not a particularly carbon-efﬁcient form of transport or accommodation. Scenarios 7 and 8 both were trips involving long intercontinental ﬂights but Scenario 7 was a week long visit to a beach resort in the Seychelles while Scenario 8 was a three month backpacking journey around South America. On a trip by trip basis, the backpacking journey around South America was by far the worst, resulting in more than ﬁve tonnes of carbon dioxide. On a per day basis, however, this scenario resulted in approximately a four-fold increase compared to staying at home, due to the ﬁve tonnes being averaged over a long period of time. Thus, if the Seychelles-visiting tourists of Scenario 7 had returned to London and for the next three months been responsible for the average carbon dioxide emissions of a UK resident, they would still have produced less carbon dioxide than the
South-American visiting backpackers of Scenario 8. Scenarios 7 and 8 show that intercontinental ﬂights inevitably produce very carbon-intensive tourist trips regardless of any other factor. Conclusions The analysis of the range of travel options presented here shows that international travel and tourism does not inevitably result in large increases in carbon dioxide emissions. With trains, planes and cars, the load factor is critical in determining carbon dioxide emissions per traveller, although with planes and trains, this is outside of the control of the traveller. Car travel would have come out signiﬁcantly better than air or train travel in the scenario analysis if three passengers rather than one had been assumed, just as plane or train travel would have come out better than the other two if full occupancy had been assumed when the basic carbon dioxide emission rates per passenger-kilometre were calculated. With short-distance travel, choice of accommodation and recreational activities becomes very important. Camping, couch-surﬁng and backpacker hostel accommodation are good choices from a carbon (and ﬁnancial) perspective and can thus contribute to low-carbon tourism. The environmental beneﬁts of choosing one of these low-carbon alternatives over hotel accommodation can more than outweigh the additional carbon dioxide emissions from using a less carbon-efﬁcient form of transport for short-distance trips. Recreational activities were shown by the scenario analysis to generally be relatively small in their carbon dioxide emissions compared to transport and accommodation-related emissions. However, where tourists engage in air-based recreational activities, such activities could be expected to be relatively signiﬁcant as a total proportion of the emissions resulting from a trip, except trips involving intercontinental travel. It is clear, therefore, that environmentally concerned travellers do have options open to them that allow carbon-neutral or nearly carbon-neutral travel and as an added beneﬁt, such travel options tend to be much lower in cost and provide signiﬁcantly greater interaction with the local people in the countries being visited than other forms of travel generally permit. Cycling is clearly the most carbon-friendly practical means of travelling internationally (and has signiﬁcant additional health beneﬁts) but travelling by train, plane or car a similar (relatively modest) distance is not substantially more carbon intensive when compared to the carbon emissions resulting from staying at home. The most important factor determining carbon dioxide emissions is the total distance travelled, which does make intercontinental travel a serious problem with regard to climate change, but other forms of low-carbon tourism offer attractive alternatives.