Tourism geography in Europe

Abstract

In the tourism component of SENSOR, attraction modelling is needed to predict the likely distribution of growth in tourism facilities at the subnational level. Modelling of tourism flows between countries is obtained through a demand modelling linked to a bilateral flow matrix. This paper presents analysis of tourist beds at the NUTSX level in order to allow for a geographical disaggregation of tourism loads within the country. In summary, 79% of the variation in tourism bed densities and 39% of the variation in growth through the 1990s can be explained by physio-geographical predictors in combination with GDP/capita and population. Prominent predictors of tourist attraction are the relatively 'fixed assets' of alpine areas in the region and access to the coast, but several variables also link the attraction modelling to other model outcomes from the SENSOR project. Population density, GDP/capita, urban and nature land cover are generally positively related to tourism loads, while agriculture is negatively related to tourism. Thus, the regression models presented in the paper can be used to estimate the attractiveness of regions to tourists in a way that will be sensitive to the scenarios specified in the SENSOR project. Furthermore, the regression results suggest the magnitude of a saturation tendency, implying that crowding at some destinations will gradually redistribute tourist to other regions within the country. © Springer-Verlag Berlin Heidelberg 2008.