Fig. 3: Importance of hydrologic processes for streamflow simulations.
From: The sensitivity of simulated streamflow to individual hydrologic processes across North America
The variance-weighted total Sobol' sensitivity \(S{T}_{i}^{w}\) regarding simulated streamflow for the nine hydrologic model components, i.e.: (a) infiltration, (b) quickflow, (c) evaporation, (d) snow balance, (e) convolution of surface runoff, (f) potential melt, (g) percolation, (h) rain-snow partitioning, and (i) precipitation correction. The results are shown for the 3316 basins that have a calibrated Nash-Sutcliffe efficiency equal to or better than 0.5 in the calibration period (January 1991 to December 2010). The two additional processes of Baseflow and Convolution (delayed runoff) are analyzed but not displayed as all basins show a sensitivity of less than 0.01 (lower limit of colorbar). Please note that the colorbar is logarithmic in order to allow for a better distinction of small sensitivity estimates. Green colors indicate a large importance of the respective process on streamflow simulations while pink colors indicate a weak impact. The location of the dots in each panel marks the location of the outlet of each basin, which coincides with a streamflow gauging station. The histograms of sensitivity values are added as an inset to each map. The histogram bins are the same as used for the colorbar.
