D
S
E
Working
Paper
A Spatial Agent-Based Model to
Explore Scenarios of
Adaptation to Climate Change
in an Alpine Tourism Destination
Stefano Balbi
Pascal Perez
Carlo Giupponi
Dipartimento
Scienze Economiche
Department
of Economics
Ca’ Foscari
University of
Venice
ISSN: 1827/336X No. 05/WP/2010
W o r k i n g P a p e r s D e p a r t m e n t o f E c o n o m i c s C a ’ F o s c a r i U n i v e r s i t y o f V e n i c e N o . 0 5 / W P / 2 0 1 0
ISSN 1827-3580
The Working Paper Series is availble only on line (www.dse.unive.it/pubblicazioni) For editorial correspondence, please contact: [email protected]
Department of Economics Ca’ Foscari University of Venice Cannaregio 873, Fondamenta San Giobbe 30121 Venice Italy
Fax: ++39 041 2349210
A Spatial Agent-Based Model to Explore Scenarios of Adaptation
to Climate Change in an Alpine Tourism Destination
Stefano Balbi
a, Pascal Perez
b,
Carlo Giupponi
aa
CEEM, Department of Economics, Ca’ Foscari University of Venice b
RMAP, Australian National University, Canberra First Draft: 1/3/2010
Abstract
A vast body of literature suggests that the European Alpine region may be one of the most sensitive to climate change impacts. Adaptation to climate change of Alpine socio-ecosystems is increasingly becoming an issue of interest for the scientific community while the people of the Alps are often unaware of or simply ignore the problem. ClimAlpTour is a European research project of the Alpine Space Programme, bringing together institutions and scholars from all countries of the Alpine arch, in view of dealing with the expected decrease in snow and ice cover, which may lead to a rethinking of tourism development beyond the traditional vision of winter sports. The research reported herein analyses the municipality of Auronzo di Cadore (22,000 ha) in the Dolomites under the famous peaks of the “Tre Cime di Lavaredo”. The local economy depends on tourism which is currently focused on the summer season, while the winter season is weak. As a whole, the destination receives approximately 65,000 guests per year with a resident population of 3,600 inhabitants. Since recently the Community Council is considering options on how to stimulate a further development of the winter tourism. This paper refers to a prototype agent-based model, called AuronzoWinSim, for the assessment of alternative scenarios of future local development strategies, taking into account complex spatial and social dynamics and interactions. Different typologies of winter tourists compose the set of human agents. Climate change scenarios are used to produce temperature and snow cover projections. The model is mainly informed by secondary sources, including demographic and economic time series, and biophysical data which feed-in its spatial dimension. Primary data from field surveys are used to calibrate the main parameters. AuronzoWinSim is planned for use in a participatory context with groups of local stakeholders.
Keywords
Alpine Winter Tourism, Spatial Agent-Based Model, Climate Change Adaptation
JEL Codes
Q
Address for correspondence:
Carlo Giupponi
Department of Economics Ca’ Foscari University of Venice Cannaregio 873, Fondamenta S.Giobbe 30121 Venezia - Italy Phone: (++39) 041 234126 Fax: (++39) 041 2349176
This Working Paper is published under the auspices of the Department of Economics of the Ca’ Foscari University of Venice. Opinions expressed herein are those of the authors and not those of the Department. The Working Paper series is designed to divulge preliminary or incomplete work, circulated to favour discussion and comments. Citation of this paper should consider its provisional character.
1. INTRODUCTION
Th e Alpine region in Europe is among the areas that are mo st rapidly affected by climate ch ange. In g eneral, the mean temperature o f this region has in creased up to +2°C fo r so me h igh altitude sites over the 1900-1990 period again st +0 .78°C in the last 100 y ears at a global level (IPCC 2007, ESFR ClimChAlp 2008). With a certain degree of local variability, glaciers h ave lost 50% o f their volume since 1850 and snow cover is decreasing espec ially at the lowest altitudes and in fall and spring. A c lear signal of climate change about precipitations is not detectable y et, but increasing risks o f extreme events have b een projected including floods, debris flows, avalanches, glacial hazards, and mo untain mass movements (Castellari 2008). The main expected imp acts on the Alps concern the hydrological conditions and water management, forests and biodiversity , agriculture, en ergy man ageme nt, and eventually tourism, which is the fo cus of this paper. While su mmer tourism is most probably going to be favored by climate ch ange (Bourdeau 2009), the World Tourism Organization started warning about the possible negative implications for winter tourism and sports since 2003 (UNWTO 2003). Nowaday s already 57 o f the main 666 ski resorts o f the
Eu ropean Alps are consid ered to not be snow-reliable1 (OECD
2007). However, climate change is also an opportunity for tho se resorts that are snow-reliable, as they will face less co mpetition in the future (Simpson et al. 2008, p . 62).
Th e Alpine people are o ften unaware o f or simply ignore the problem, which is a commo n problem of climate change
1
In general, a ski resort is considered to be snow-reliable if, in 7 out of 10 winters, a sufficient snow covering of at least 30 to 50 cm is available for ski sport on at least 100 days between December 1 and April 15 (Burki at al. 2007).
ad aptation. Perhaps, at a very local level, climate patterns are not evident enough to question the develop ment model b ased on the “white dream” which has prevailed since the seventies. Indeed, a mo del of deve lopment based on snow , no matter if natural or artificial, is still somehow surviving notwithstanding the maturity o f the traditional ski product and the stagnation o f the market demand (Macchiavelli 2009). At this point very careful assessments should be carried on befo re any further snow-based development plan (WWF 2007).
ClimAlp Tour is a European project o f the Alpine Sp ace Programme, bringing together institutions and scholars from all countries of the Alpine arch, in view o f dealing with the expected decrease in snow and ice cover, which may lead to a rethinking o f tourism develop ment bey ond the traditional vision o f winter
sports2. The project analy ses 22 pilot are as with diverse
environ mental, social and economic conditions in order to provide a g lobal perspective on the Alpine tourism. Raising the aw aren ess o f the stakeholders including tourists, population and businesses on the impa ct o f climate change on tourist econo my o f the Alps and on possible ad aptation strategies is one o f the goals o f the whole project.
Th is paper explores the conceptualization phase o f an ag ent-b ased mo del (ABM) capable o f gathering the available h eterogeneous in formation and assess diffe rent scen arios of future local development, and eventually tourist supply, taking into account co mplex spatia l and social dy namics and intera ctions. Th is methodology has already proven to be successfu l for other tourism related issues (i.e. Perez et al. 2009). However, to our knowledge, this is one o f the first attempts to explore the
interactions between climate ch ange and winter mountain tourism
by means of an agent-oriented approach3. Our work is still in
progress and is meant to co mplement a decision support sy stem
(DSS)4, which will be imp lemented during a series o f
particip atory workshops.
2. CASE STUDY
Th e research reported herein analyses the mu nicipality of Auronzo di Cadore located in the province o f Belluno, in the Veneto region, in the north-east of Italy . It covers a v ast area (22.000 ha) which includes Misurina with its lak e and the mo st famou s mountain of the Dolo mites, namely the “Tre Cime di Lavaredo”, part o f the UN ESCO world heritage since 2009.
Th e village o f Auronzo (866m on the sea level) ho sts almost the entire population of the municipality of approximately 3,600 inhabitants. It is located in the middle of the v alley of the Ansiei riv er, which forms the artificial Santa Caterina Lake, due to the presence of an old d am. The lake basin is 3 k m long and is endowed with beach facilities that periodically host mo tor nautical and canoe compe titions. Misurina is a sma ll settlement at 25 km from the main village, placed at an altitude o f 1754 meters, just under the Tre Cime di Lavaredo peak, which is accessible both by means o f several mountain paths and by a toll regula ted carriageway .
Th e local econo my depends on tourism which is currently focused on the summer season, while the winter season is weak, with only 25% o f arriv als (Regione Veneto 2009). Indeed, hiking (200 k m
3
See also Sax et al. 2007.
4
o f signed mo untain paths and 10 alpine refuges) and relax are the main elemen ts o f attraction. The total hosting capacity is o f approximately 7.300 bed s of which around 1700 in the hotel sector and the remainder in the extra-hotel sector (B&Bs, lodgings, etc.). Th e 75% of the hotels’ beds concern structures with 1 or 2 stars. In 2008 63.700 arrivals and 305.400 tourist nights were registered, showing a slight decrease from the previous year. The last 10 years have witnessed the increase o f arrivals and the contraction o f av erage stay .
Notwithstanding the p resence o f two small downhill ski area s and two cross-country ski centers so me hotels don’t even open for the winter season. Th e four ski-lifts o f Mount Agudo, which reach a max imu m elevation o f 1600 m, supply seven ski-trails covering 15 k m. In the locality o f Palus San Marco, ju st at halfway between Auronzo and Misurina, there is the Somadida Forest, one o f the largest o f the p rovince, which beco mes a cross-country ski centre (with nine loops of 52.5 k m in total) during the winter season. Th e Marmarole sled-dog centre and an ice-kart circuit are also placed Palus. In addition, Misurina, which has an hosting capa city o f approximately 500 beds is endowed with the two ski-lifts o f Col de Varda (fro m 1756 m to 2220 m) that supply five ski-tracks, and 17k m of cross-country ski loops.
Since recently the Co mmun ity Council is considering options on how to stimulate a furth er dev elop ment of the winter tourism. Th ere exist several proje cts o f ski-areas develop ment. The mo st amb itious is located in Marzon valley , a few km fro m the main village, which would connect the valley to the ski-a rea o f Misurina (with an average altitude over 2000 m). The main problem at stake is how to develop winter tourism in the next 40 y ears, in a context of climate ch ange and market demand that is
not favorable. In particular, the spatial heterogeneity given by the bipolarity o f the case study , suggested the elaboration o f a sp atially explicit ABM.
F i g u r e 1 . M a p o f t h e M u n i c i p a l i t y o f A u r o n z o d i C a d o r e
3. MODEL DEVELOPMENT
Th e dev elop ment o f a conceptual model h as b een carried out to support the design o f the ABM by integrating three methods gravitating around different research groups o f the heterogeneous ABM scientific co mmun ity . The ARDI (Actor, Resources, Dynamics, and Interactions) method belongs to the co mpanion mo deling tradition, mainly applied to natural research man ageme nt. It can be extremely e fficient for ju mpstarting the process of visual formalization o f the do main model (Etienne 2006). The ODD (Overview, Design concepts, D etails) protocol belongs to the individual-based mod eling branch o f ecology , but
is g aining further diffu sion in so cial science. Differently from the other methods, it consists o f a narrative description o f the various elements of an ABM, contributing to a mo re rigorous fo rmulation phase (Grimm et al. 2006). Finally , the UML (Unified Modeling Language) belongs to the co mputer science tradition and is probably the most effective methods, preceding the coding phase, which can guarantee th e full replicability o f the model (Boch et al. 1999).
Th ese methods have b een firstly applied in the order in which are presented, subsequently leading to a more iterative approach . By means of ARD I we id entified the tourist fa cilities as the main resources o f the sy stem, th e winter tourists as the acting agents and the meta-economic agents as the economic units in charge of accounting the economic flows for each tourism sub-sector. The application of ODD led to the d efinition of the tourists’ heterogenic b ehavior and o f the future sc enario s. Finally , through UML, we were able to formalize in diagrams all the details. The clear advantage is that the modeler is endowed with a set o f tools mu tually checking the model internal consistency with regards to both its static and the dynamic aspects.
3.1 Representing Heterogeneo us Market Behavior
Ev en though winter sports, and especially downhill skiing, are still th e essence of winter mo untain tourism the market h as reached its matu rity and is challenged by (a) loss o f shares in the tourist market in the Alpine countries throughout Europe, (b) co mpetition from other tourist destin ations, (c) the growing econo mic and territorial divide between large and small resorts, (d) the need for huge new investmen ts against the background o f a
reduction o f public funding, (e) new recreational practices (freesty le and freeride), (f) the ageing o f th e tourist population, (g) de mand for environ mental quality , (h) the changed notion o f
resort5, (i) the inclination toward shortened and repe ated holiday s,
(j) b ehavioral unpredictability , due to wheatear forecasts, and fin ally (k) the search for new markets (Minghetti, 2002, D aidola, 2006, Bourdeau 2009, Macchiav elli 2009). However, according to
Camanni [2002 ] and Daidola6 [2007], a new light ski industry,
with less investments and more flexibility with reg ard to climate conditions, is possible and the small resorts may thus be advantaged.
ABMs can be of great value in dealing with su ch dynamics because they are particularly well suited to incorporate heterogeneity of behavior. Our approach focu ses on the simulation o f tourists’ response (demand-side) to alternative strategies o f development o f tourism facilitie s o f the de stination (supply-side) in order to provide the local stakeholders (residents, entrepreneurs, local tourism organizations, co mmu nity council), with quantitative indicators of possible futures which depend on their collective decisions.
Drawing on the above cited literature and esp ecially on secondary data from marketing surveys o f the tourism statistical
observatories o f Trentino and Alto Adige7 (Provincia Autonoma di
Trento 2007, Provincia Autono ma d i Bolzano 2009), we crea ted a
5
“The idea of the resort as a unity of place, time and action, can be circumvented or deviated by a new interpretation of the mountain playground. One example is the striking contrast that can be observed by the expansion and interconnection of large ski areas, and the micro-scale of space in which the new sports are practiced by young snow surfers such as the snowparks” (Bourdeau 2009).
6
Daidola’s hypothesis is that artificial snow is less satisfying and more dangerous than natural snow. Because of artificial snow, the current ski style has become repetitive, expensive and boring, and at the same time too easy to learn and too fast. On the contrary ski and snowboard magazines mainly propose powdery snow and exciting experiences. Moreover artificial snow is responsible for the vicious circle of huge investments of the last years.
7
set o f eight tourist pro files which is rich enough to take into account (1) the actual winter tourists of Auronzo, (2) the ac tual winter tourist visiting Auronzo’s main competitors, and (3) the potential winter tourist o f to morrow (table 1).
T a b l e 1 . N a r r a t i v e s d e s c r i b i n g t h e t o u r i s t p r o f i l e s
T raditional Sk i I n tens ive ( TSI ). M ain ly m ot iv at e d by t he p ra ct ic e o f d ow nh il l s k ii n g ( 3 / 4 ) an d s no w boa r d in g ( 1/ 4) . P r i ma r i ly in te re s te d a bo ut t he s k i- li fts a nd t ra il s. A pp reci ate s g a str ono my . Not sa ti sfied by th e sc arce prese nce of bars and p ub s. Ha s a mid dl e- l ow prof il e of expense ( 11 0 E uro s/d a y) . Ha s o ne o r m ore wi nt e r holid a ys e very ye ar , ofte n in t he same des t in at i on . T he bo ok in g pro ce ss i s o ft e n i nt er me d ia ted a n d t he t ime u ni t o f r e fe re nce i s 7 d ay s. B o ok s h is h ol id a y s in Fe b ruary and M a rc h. T rav e ls in c om pa ny wit h o ne o r m ore frie n ds or wit h t h e fa mily .
S ki P art -T ime (S P T). M ai nl y m ot i vat e d by t he n at ur al e n vi ron me n t a n d b y t he p r a c t ic e of d ow n hi ll ski i n g ( 2/ 3) a nd sn ow b oa r d i n g ( 1 / 3) , w it h a soft e r ap pro a c h. T he tre nd towa rd s sno w bo ar d i s mo re ev ide nt . In te re ste d a bou t sk i - li ft s and t ra il s, b ut a l so a b ou t ot he r w i nte r sp ort s. Dedi ca t es t o t radi t i ona l sk i ac t i vit ie s no t mo re t ha n ha l f o f t he day . Oft e n n ov i ce s, m a in ly w i t h fa m il y. No t i nt eres t e d in g a st ro no my. N ot sa t i sfi e d by t he sc a rce pre se nce of c hi l dre n fac i li t ie s. H as a lo w p rofi l e of ex pe nse ( 10 0 Eu ro s/day) . V i sit s t h e sam e de st i nat io n, oft e n fo r t he C h rist ma s ho li day s. T he t ime u ni t o f re fe re nce is 7 d ay s. Te nd to b oo k d ir ect l y t o t he ho spi t a l it y stru ct ure. B o ok in g is l ess p red ict a bl e and more subj e cte d t o w ea t he r c on d i t i on s a nd f or e c a st s.
S p o r t y A lt e r n a ti v e. Ma in l y mo t i va te d by t h e na t ura l e n vi ron me nt a n d t he prac ti ce of a l te rnat ive wi nt er s por t s. O fte n i n g ro u p a n d on a d a y t rip . Oft e n adu l t or e l de r. H ig hl y de pe nd e nt on weat her c on di t i ons a nd fo reca st s. Has t h e l ower p rofil e of e x pe nse ( un de r 10 0 E ur os/ day ). T hi s t y po lo gy i s co mp ose d of t w o pro fi le s: t h e p ra ct it io ne rs of c ross- coun t ry ski in g (SA X ) and th e practitione rs of bac kcount ry sk ii n g a nd sk i to uri ng ( SAW ).
I dl e (I D). Ma in ly m ot i v at e d by t he n atu r a l e n vir on me nt . Sc a rc e ly i nt e r e s t e d b y any i nt en sive spo rt act i vi t ie s. A p pre c ia t es ea sy tre k ki n g i n t h e sno w, c hi ld re n faci li t i e s a n d wel l ne ss a nd s pa se rvi ce s. No t sat i sfi e d by t he sca rce pr ese nc e b a rs an d p ub s and ot he r sport fac ilities. Often fe ma les, acc o mp an y ing skie rs. Often with fa mi ly . Te nd s t o bo ok fo r t he C h ris t ma s hol i da y s. Ha s a mi dd l e- hi gh p rofi l e of e xpe nse ( 13 0 E u ros/ da y).
E cl e ct i c ( EC ). M a i nly mo t i va te d by t he idea of ha vi n g f un by mea ns o f a set d iv e r s i f ie d act iv it ie s not r e lat e d t o t r ad i t io na l s k ii ng . I nt e r es t e d i n g a s t r ono my , w e ll ne ss an d s pa a nd sho p p in g. A pp r e c ia t e s mi nor w i n t e r a c t i vit i e s ( i . e . i ce sk a t i ng, h oc k ey, s led- do g… ). T he n a t ur al e nv iro nm en t re mai ns im po rt a nt . Ha s t he h ig he st p rofi l e of expe n se ( ove r 1 50 E uro s/da y) but t he sh ort er avera ge stay . N ot sa t i sfied b y t he sc arc e p r e se nc e of b ar s a nd p ub s a n d ot h e r sp ort s f ac i li t ie s.
C o u n t e r c u lt u re . M ai nly mo t i vat ed b y t h e de s i re of ha v in g d e e p a n d a ut h e nt ic e x pe r ie nc e s, m e r g i ng w il de r ne ss a n d p la y grou nd , n at ure a nd a d re na l i ne . M a i nl y sp ort y a nd ex pert i n t he ir di sci p li ne. C om po se d of t w o pr ofi le s: p ra ct it i one rs of fr ee style (CCP ) a nd of fre e r i de (CCW ). O f te n t ra ve ll in g w it h f r iend s p r a ct ic i ng t h e sa me act iv ity . Int ere ste d in n a t ur al snow, v ery de pe nde nt on w he ate ar c on dit i ons a nd fo rec a st s. CCP i s in t e re ste d i n ded ica t ed fac il it ie s a s sno wp a rks . CC W a p prec i at e s t he o rga n iz ati on o f t h e nat ura l e nv ir onme nt in t o off- pi st e t ra il s an d t he pre sence of smal l sk i- li fts an d se cu rity se rv ice s. B oth ar e se nsi t i ve t o e n vi ronme n ta l issu e s Ha v e a me di um - hi gh pro fi le o f ex p en se ac c ord in g t o t he serv i ce s t h at a re off er ed .
3.2 Representing Alternativ e Futures
Ev ery simulation run requires the users to mak e three choices: (1) the development strategy to be tested; (2) the societal scen ario that sets the conditional context in terms of numb er o f tourists that could be available to choose the destin ation; and (3) the climate projection, in terms of snow cover and temperature. The co mbination o f these choic es defines the future conditions, fro m 2011 to 2050, under which the tourists’ response is simulated. Th e underly ing idea th at has inspir ed the mod el is to identify the mo st robust d evelopmen t strategy. In this regard , we have defined four spatially explicit alternativ e strategies which are able to take into account various orientation towards tourism and the perception o f climate change from the local stakeholders’ point o f view.
Th e first strategy is the pursu e o f the traditional ski intensive paradigm. Indeed, one of the most familiar measu res in the struggle against snow-deficient winters is the constru ction of high co st artificial snowmak ing fac ilities (Burki et al. 2007). How ever, in this stra tegy Auronzo not only maintains the ski areas o f Mount Agudo and Misurina, but also develops two n ew ski area s with snowpark , on the basis o f two p rojects about Marzon and Da Rin valley s, which have different spatial conditions. The overall hosting cap acity remains untouched while a minor increase concerning restaurants and retailers supply is included . The second strategy emb races the vision o f Daidola [2006] and Burdeau [2009] o f an alternativ e light sk i oriented and post-mo dern development. It integrates the wilderness and the play ground concepts increasing the supply of controlled o ff-piste track s (Marzon and Da Rin valley s), cross-country itineraries
(Palus) and snow parks (Mount Agudo). These are assisted by a very limited develop ment o f the existing ski lifts and a mo re sober artificial snow mak ing behavior. Th e third is the non-snow strategy that is the well established, but o ften not self-sustaining, process of diversification and enlargemen t of tourist offer by means o f higher quality hotels, shopping, g astronomy, pubs and bars, and, mo st o f all, wellness and spa centers. This is mainly implemented in Auronzo, Misurina and Palu s villages. The ski areas remain in function without the support o f artificial snow. Finally the fatalistic strategy consists of no changes in the supply behavior, which could also be described as “business as usual”. We are planning to consid er 2 societal scenarios, one mo re conservative and one more optimistic, which will be derived interpolating historic European population dynamics and the alpine tourism fluxes. The climate p rojections will b e developed in order to represent 3 IPCC scenario s o f aerosol and GH G forcings (A2, A1B and B1), draw ing on the En semb les European project8.
4. MODEL DESCRIPTION
Th is concise model description loosely follows the ODD protocol (Grimm et al. 2006). We also provide the UML class, sequence and activities diagrams (the first two as appendices and the rest in
the project website: http://www .dse.unive.it/clim/
climalptour.htm).
4.1 Mo del Overview
Th e model purpose is to an aly ze alternative winter development strategies for the case study simu lating the tourists’ response under different climate scenarios.
En tities and state va riables a re visually presented in the U ML
class diagram (Appendix A). Th e main entities are the tourists, the tourism facilities and the meta-econo mic agents. Th e tourists are divided in eight pro files: traditional sk i-intensive (TSI), ski part-time (SPT), sporty alternative cross-country (SAX), sporty alternative wilderness (SA W), idle (ID), eclectic (EC), counter-culture wilderness (CCW) and counter-counter-culture playground (CCP). A ccording to the p rofile they b elong to (v ariable “ty pe”), ag ents have different preferences and behavior with regards to the tourism facilities. The tourism facilities, which in turn co mpose the destination supply structure, are divided in eight types: four are snow related (facilities dedicated to downhill skiing, snowparks, cross-country skiing and o ff-piste skiing) and four are non-snow related (acco mmodations, re staurants, retailers, and other facilities including those for kids, wellness and spa, and other sports). The meta-economic agents are the accountants o f the tourism facilities. One meta-economic agent is in charge o f man aging one type of tourism facility. They keep track o f the investment required to put in place their facilities, d efined by the development strategy to be an aly zed , and of their money flow. Th e development strategies, which represent possible orientations o f the local stak eholders, are set exogenously by the model.
Concerning the scales, both the snow and th e non-snow related facilities are located in a spatial grid o f approximately 10.000 cells which represent the destination. Each cell represents a 150 x
150 meters square and contains the actual geographical in formation of the area. N ine spatially distributed reference points have b een identified in order to represent the different snow conditions under alternative climate scenarios. A simu lation is compo sed of 40 cy cles, which are the winter seasons from 2011 to 2050. Each cycle consists o f 126 days (time steps) that
represent the 18 week s fro m the 1s t o f Decemb er to the 6t h o f
April. Summing up, every simulation takes into account 720 weeks and 5040 time steps.
Process overview and scheduling are captured in the UML
sequence diagram (Appendix B), which shows the sequence o f opera tions p erformed by each class. Ev ery operation is then furth er describ ed as activity diagram. Initially, the spatial units (patches) update their attributes and configure the tourism facilities pre sence as per selected develop ment strategy . Each o f the meta-econo mic ag ent is assigned with the investment needed to meet that con figuration. The reference points read snow cover and temperature from the climate data which describe the selected climate projection. Th ey also perform three kinds o f forecasts concerning snow cover, at short and med iu m term, and snow security , a t short term. Th en, the snow facilities check those foreca sts and store the in formation that they will subsequently pass to the tourists. Downhill and snowparks can decide to produce artificial snow. A fte r th at, the tourists, whose total amount is taken by th e societal scenario, can check the destination in order to become v isitors, if their require ments are met. According to their behavior, they can check the forecasts and go to the p lanning ph ase, in which they decide their day of arrival. If th ey are in the destination th ey enter into a loop o f
opera tions which describ e the use of the tourism facilities. Then, ev ery facility can check its own users and pass the information to the patches that can visually describe the tourists d ensity on the grid. The meta-econo mic agent calculates the return associated to the facilities u se and update their balance sheet. Finally , the tourists in their last day of vacation calculate their overall satisfaction. If this is n egative they exit fro m th e simulation, if it is positive they remain among the potential visitors and can plan a furth er vacation. Those that eliminated are substituted by other tourist agents, but they are also accounted for in the statistics about dissatisfaction.
Input data is provided in various forms. The sp atial units are
georeferentiated, based on GIS (geographical in formation system) lay ers concerning elevation, slope and radiation. The reference points are provided with climate data in form o f time series o f snow cover and temp erature, a ccording to the selected climate projection. Finally, the societa l data provides the total number o f tourist agents available for each o f the 40 cy cles during every simulation.
Most o f the operations make use o f submodels in form o f simple algorith ms and logical tests which a re presented in detail in the activity diagrams. Mo st o f the para meters used in the activity diagra ms are calibrated on the case study, by means o f field surv ey s. The rest are retrieved fro m the literature.
Th e model in itialization represents the destination winter tourism conditions in 2011 concerning the actual amount, type and spatial con figuration of the tourism facilitie s.
4.2 ABM Desig n Concepts
Emergence. Once the boundary conditions of possible futures are set by the 3 choices on the scenarios for any model run, then the performance of the destination is a phenomenon that emerges fro m the tourists’ behavior. This is numerically expressed in terms of tourist attracted and money flow produced by each facility type (see updateUsers and updateBalance activity diagrams) and visualized on the spatial grid at e ach time step in terms o f tourists’ density (see checkOccupation activity diagram).
Objectives. The tourists can choose whether to go or not to the destination, according to their pre ferences about the de stination supply and to th e weather forecasts (see checkDestin ation and doForecast activity diagrams).
Prediction. Tourists’ expectations are based on the facilities av ailable and their sp atial con figuration in th e destination and on the snow cover and security forecasts.
Sensing. The tourists are fully aware of the destination’s facilities and spatial attributes b efore the vacation, but they perceive the environ mental conditions o f the facilities they use only in loco .
Learning. Each day of their vacation the tourists calculate their satisfaction which depends on the effectiv e environmental conditions en countered in terms o f snow cover and tourists’ density , so that their availability to a subsequent vacation depends on the me mo ries of th e previous one (see calcSatisfaction activity diagram).
Adaptation. The tourists adapt by not visiting Auronzo again once their satisfaction goes negative, in favor o f other
co mpeting d estinations (see updateSatisfaction activity diagram).
Interaction. The tourists directly receive stimuli, through sensing, fro m the destination’s facilitie s and spatial attributes, which affect their eligibility to book their vacation. They also indirectly interact among each other with negativ e effects on their respective satisfaction, over certain den sity thresholds. Stocasticity is used to reproduce variability in the various
ag ents’ decision processes, primarily by mean s o f normal distributions (see the activity diagrams).
Observation. The model collects data on the economic perfo rmance o f the eight types o f facilities and on the spatially distributed tourist fruition o f the destination.
5. D ISCUSSION AND CONCLUSION
Our work is still in progress but it can already provide so me interesting insights on the modeling of climate change and tourism.
First o f all, the scale o f analysis and the level of detail represent a significant imp rovement in the climate change research, which is normally performed at a much higher level o f spatial and temporal aggregation. This scale perfectly fits the crucial so cioecono mic dynamics o f lo cal adaptation that h ave to be investigated .
Second, to our knowledge this is one o f the first attemp ts to formalize the supply stru cture o f a winter tourism d estination in classes by mean s o f UML. The result is simple and clear but is able to represent co mplex interactions, for alternative development strategies, in a spatially explicit way . This is
extremely valuable giving our objective o f assessing robustness and flexibility rather than finding optima l solutions. This conceptualization can ev entually become a generic ontology , if it will b e proved to fit the application to other case studies.
Th ird, the simu lation focuses on the feed-backs o f the de mand side of tourism, which is often missing in the existing tourism mo dels. However, the tourists are the ultimate judges o f a destination adaptation strategy , because they will decide the winners and losers o f the future. This is why we regarded as fundamental to focus on their agency in an heterogeneous way, drawing fro m d isciplines such as custo mer behavior, which could only be done by means o f an the agent-based approach.
Fourth, the model develop ment itself is a novelty because we integrated some o f the ma in practices o f the ABM co mmu nity which are n ever found together but can mutually benefit each other.
Th e next steps of our work include the implemen tation o f the U ML model into a software for computer simulation. AuronzoWin Sim will then be tested and refined, befo re being used in a particip atory context with groups o f local stakeholders, in two ways. Initially, it will support the collective discussion on possible adaptation and develop ment strategie s, and the criteria to assess their robustne ss. Secondly , it will incorporate the strategic adjustments proposed by the stakeholders.
REFERENCES
AuronzoWin Sim Model Website, http://w ww.dse.unive.it/clim/
climalptour.htm
Boch, G., J. Rumbaugh, I. Jacobson, Th e unified modeling language user guide, Addison-Wesley , 1999.
Bourdeau , P., Mountain tourism in a climate of change. In Alpine sp ace - man & environ ment, vol. 7: Global Change and Sustainable Develop ment in Mountain Regions, Innsbruck University Press, Innsbru ck, Austria, 2009.
Burki, R., B. Abegg, and H. Elsasser, Climate change and tourism in the alpine region of Switzerland. In B. Amelung et al., Climate change and tourism: assessment and coping strategies, 165-172, Maastricht, 2007.
Camanni, E., La nuova vita d elle Alpi, Bollati Boringhieri, Torino, Italia, 2002.
Castellari, S., Climate change, impacts and adaptation strategies in the Alpine Space: so me results from th e INTERREG III B project ClimChAlp, Proceedings of the con ference on Mountains as Early Indicators of Climate Change, Padova, Italy , 17-18 April 2008.
Daidola G., Il corto circuito dello sci di massa. In World Wide Found (WWF), Alpi e turismo: trovare il punto di equilibrio, WWF Italia, February , 2006.
En semb les Projects’s Website, http://ensembles-eu.meto ffice.co m/
ESFR ClimChAlp, Extended Scientific Final Report o f the ClimChAlp project, 2008, The Climate Change Asse ssment
Etienne, M., Co mpanion modelling: a tool for dialogue and concertation in Biosphere Reserves. UNESCO-MAB, Te chnical Notes 1 :44-52, Paris, Fran ce, 2006.
Giupponi, C. and A. Sgobbi, Models and decisions support sy stems for participatory decision making in integrated water resource man age ment. In P. Koundouri, Coping with Water D eficiency, 165-186, Springer, 2008.
Grimm, V ., U. Berger, F. Bastiansen , S. Eliassen , V. Ginot, J. Giske, J. Goss-Custard , et al., A standard protocol for describing individual-based and agent-based models.
Ecological Modelling, 198(1-2): 115–126, 2006.
Intergovernmental Pan el on Climate Change (IPCC), Contribution o f Working Group I to the Fourth Assessment Report o f the IPCC, Climate Change 2007: The Phy sical Science Basis, Cambridge University Press, Camb ridge, United Kingdo m, 996 pp, 2007.
Macchiavelli, A., Alpine tourism: d evelopment contradictions and conditions for innovation, Journal of Alpine Research, 97-1,
2009. Available at http://rga.revues.org/index843.html.
Minghetti, V ., Il turista della neve in Italia: profilo, esigenze, nuovi modelli di consu mo . In Pechlaner, H., and M. Manente, Manuale del turismo montano: prospettive, camb iamenti e strategie di management, Touring Club Italiano, Milano, Italia, 2002.
Organization for Econo mic Co-operation and Development (O ECD), Climate change and the European Alps: adapting winter tourism and natural hazard management, January , 2007. Perez , P., A . D ray , D . Cheland, and J.E. Ariaz-Gonzalez , An
agent-based mod el to address coastal man agement issues in the Yucatan Peninsula, Mexico. In 18th World IMACS Congress
and MODSIM09 International Congress on Modelling and Simulation. Cairns, Australia from 13-17 July , 2009.
Provincia Autonoma d i Bolzano, ASTAT, Pro filo dei turisti in Alto Adige: anno turistico 2007/08, In fo N. 29, May , 2009.
Available at
http://www.provinz.bz.it/astat/it/mobilita-turismo/474.asp
Provincia Autono ma di Trento , Osservatorio Provinciale per il Turismo , La stagione turistica invernale 2006/07 in Trentino, December, 2007. Available at http://www.turismo. provincia .tn .it/osservatorio/pubblicazioni/pubblicazioni_report/ report_andamenti_stagionali/
Regione Veneto, online database of the department o f statistics, last access January, 2010. Available at
http://statistica.regione.veneto.it/dati_settoriali_turismo.jsp Sax, M., J. Schmude, and A. Dingeldey , Water and tourism –
simulating the tourist water u se in the Upper Danube catchment a rea. In Reducing th e vulnerabilities o f societies to water related risks at th e basin scale (Pro ceedings o f the international sy mposium o n integrated water resources man ageme nt, Bochu m, Germany , September 2006) IAHS Publ. 317, 2007. Simp son, M.C., S. Gössling, D. Scott, C.M. Hall, and E. Gladin,
Climate change adaptation and mitigation in the tourism sector: frameworks, tools and prac tices, UN EP, University of Oxford, UNWTO, WMO, Paris, France, 2008.
World Tourism Organization (UNWTO), Climate change and tourism, Proceedings of the 1 st International Con ference on Climate Change and Tourism, Djerba , Tunisia, 9-11 April, 2003.
World Wide Found (WWF), Alpi e turismo: trovare il punto di equilibrio, WWF Italia, February , 2006.
APPENDIX B. AuronzoWinSim SEQUENCE DIAGRAM
Testo WP
