The simulation of global change scenarios required downscaling
methods. Although the three methods tested here gave similar
results, they may have shown biases that remained unknown. This
might be a problem as the watershed was located in a very heterogeneous
grid cell of the global climatic models, a result of the
contrasted topography.