Abstract
The present study focuses on quantifying the impact of the choice of
spatio-temporal resolution and hydrology models on the projection of
extreme flow and their link to the catchment size. We use two
process-based distributed hydrology models forced with a large-ensemble
regional climate model (50-member ClimEx dataset) over the 1990-2100
period at different spatio-temporal scales. The extreme summer-fall flow
corresponding with each spatio-temporal resolution was extracted by
pooling the members together and computing the empirical cumulative
distribution function. The results show that by refining the time-step
from daily to sub-daily, the summer-fall extreme flow projected over the
future period exceeds that of the reference period for the small but not
large catchments. By increasing the catchment size, the hydrology
model’s contribution to the variability of extreme flow increases.
Moreover, the choice of spatial resolution affects the extreme flow’s
trend in terms of magnitude, significance, and direction. But no pattern
regarding the catchment size and spatial discretization variations
exists.