 |
| Titel |
Uncertainty of the hydrological response to climate change conditions; 605 basins, 3 hydrological models, 5 climate models, 5 hydrological variables |
| VerfasserIn |
Lieke Melsen, Naoki Mizukami, Andrew Newman, Martyn Clark, Adriaan Teuling |
| Konferenz |
EGU General Assembly 2016
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 18 (2016) |
| Datensatznummer |
250126664
|
| Publikation (Nr.) |
 EGU/EGU2016-6420.pdf |
|
|
|
|
|
| Zusammenfassung |
| Many studies investigated the effect of a changing climate on the hydrological response of a
catchment and uncertainty of the effect coming from hydrologic modelling (e.g., forcing,
hydrologic model structures, and parameters). However, most past studies used only a single
or a small number of catchments. To go beyond the case-study, and to assess the uncertainty
involved in modelling the hydrological impact of climate change more comprehensively, we
studied 605 basins over a wide range of climate regimes throughout the contiguous USA. We
used three different widely-used hydrological models (VIC, HBV, SAC), which we forced
with five distinct climate model outputs. The hydrological models have been run for a base
period (1986-2008) for which observations were available, and for a future period
(2070-2099). Instead of calibrating each hydrological model for each basin, the model has
been run with a parameter sample (varying from 1600 to 1900 samples dependent on
the number of free parameters in the model). Five hydrological states and fluxes
were stored; discharge, evapotranspiration, soil moisture, SWE and snow melt,
and 15 different metrics and signatures have been obtained for each model run.
With the results, we conduct a sensitivity analysis over the change in signatures
from the future period compared to the base period. In this way, we can identify
the parameters that are responsible for certain projected changes, and identify the
processes responsible for this change. By using three different models, in which VIC
is most distinctive in including explicit vegetation parameters, we can compare
different process representations and the effect on the projected hydrological change. |
|
|
|
|
|
|