We built three contrasted future LULC scenarios from 2020 to 2085 with the CLUMPY model (Mazy and Longaretti, 2022). The CLUMPY model is an innovative model of land use and land cover change comprising a calibration-estimation module separate from a non-biased allocation module. It is calibrated by using time series of past LULC maps (Mazy and Longaretti, 2022). The model then calculates transition probabilities for each LULC class according to relevant spatial explanatory variables. Next, the model can produce maps of future LULC distributions according to information it learned during the calibration-estimation phase. This model has the benefits of being easy to use, proposing nonbiased allocation methods and producing scenarios of future LULC change either by adjusting manually the matrix of LULC transitions probabilities (used for the Conservation and Tourism scenarios) or by training the model on specific areas of the past time series (only used for the Conservation scenario).

The scenarios were based on the “Montagne 2040” report that explore potential future for specific alpine region and their associated narratives (Claveranne, 2013). The “Business As Usual” scenario (BAU) extended the 2006-2016 observed trends. The “Conservation” scenario assumed: the expansion of protected areas, reduced support for agriculture and livestock farming, and LULC transitions, i.e., from grassland to shrub or forest, and from shrubland to forest (aka landscape encroachment). The “Tourism” scenario assumed the development of ski resorts and mass tourism in the valley, growing urbanization at low and middle altitude, and associated LULC transitions from grassland or shrubland to urban (see Supplementary Material D). Scenarios of LULC change only considered broad categories of LULC: urban, forest, shrub and grassland.