In modelling exercises there are different sources of uncertainty, including data uncertainty, model uncertainty, and scenario uncertainty. For the model predictions of Scots pine development under climate change in the Prades area using the Gotilwa+ model a sensitivity analysis was performed on the input variables of the model, more particularly on model parameters related to photosynthesis, stomatal condition, and on input variables related to thermal inertia, tree structure and soil. Each tested variable was increased and decreased with 20% of its initial value, and the effect after a modeling period of 50 and 100 years was evaluated for the output variables aboveground biomass (t/ha), basal area (m2/ha), GPP (t.ha-1.year-1), NPP (t.ha-1.year-1), WUE (mmol/mol), transpiration (mm/year), autotrophic respiration (t.ha-1.year-1), and heterotrophic respiration (t.ha-1.year-1).
Concerning sensitivity to photosynthesis parameters, all output variables in particular biomass, GPP, NPP, WUE, and respiration were sensitive to changes in the maximum rate of electron transport Jmax, while not any of the output variables appeared sensitive for changes in maximum rate of carboxylation VCmax. This might be because VCmax is not light limited in the Mediterranean area, in contrast to other more northern latitudes. Concerning sensitivity to parameters determining stomatal conductance, all variables, particularly biomass and WUE were sensitive to changes in the constant of Leuning LEC, and only little sensitive to other parameters. Concerning sensitivity to temperature variables most output variables were very sensitive to maximum temperature, less to minimum temperature, and little or not to thermal inertia. As to the sensitivity to tree structure variables and allometrics, the output variables are most sensitive for changes in the respiring fraction of sapwood and for the amount of mobile carbon. As to the soil variables, the output is mainly sensitive for field capacity and soil depth.