Premium
Demasking the integrated information of discharge: Advancing sensitivity analysis to consider different hydrological components and their rates of change
Author(s) -
Guse Björn,
Pfannerstill Matthias,
Gafurov Abror,
Fohrer Nicola,
Gupta Hoshin
Publication year - 2016
Publication title -
water resources research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.863
H-Index - 217
eISSN - 1944-7973
pISSN - 0043-1397
DOI - 10.1002/2016wr018894
Subject(s) - sensitivity (control systems) , environmental science , variable (mathematics) , dominance (genetics) , statistics , biological system , mathematics , chemistry , engineering , gene , mathematical analysis , biochemistry , electronic engineering , biology
Abstract Discharge as an integrated representation of all hydrological processes is the most common response variable used in sensitivity analyses. However, due to overlaying effects of all hydrological processes, the sensitivity signal of certain parameters to discharge can be masked. A more informative form of sensitivity analysis can be achieved by investigating how parameter sensitivities are related to individual modeled hydrological components. In our study, the TEDPAS (TEmporal Dynamics of PArameter Sensitivity) methodology is used to calculate daily sensitivities to modeled hydrological components and to detect temporal variations in dominant parameters. As a further enhancement to consider both magnitude and dynamics, temporal variations in parameter dominance are analyzed, both for magnitudes and rates of change of hydrological components. For this purpose, regime curves for parameter sensitivities are constructed. The results demonstrate that sensitivities of parameters increase when using the corresponding hydrological component instead of discharge as response variable. For each hydrological component, seasonal patterns of parameter dominance are detected using both magnitude and rate of change as response variable. Major differences are detected for certain capacity parameters, which are less pronounced using rates of change. Overall, we show that disentangling the diagnostic information hidden in the integrated signal of discharge can lead to a more informative signal regarding the sensitivity of hydrological components. Such advancements in sensitivity analysis can lead to a better understanding of how model parameters control the individual hydrological components in time.