 |
| Titel |
Understanding signatures in hydrological calibration - A Bayesian perspective |
| VerfasserIn |
Dmitri Kavetski, Fabrizio Fenicia, Peter Reichert, Carlo Albert |
| Konferenz |
EGU General Assembly 2017
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 19 (2017) |
| Datensatznummer |
250141658
|
| Publikation (Nr.) |
 EGU/EGU2017-5192.pdf |
|
|
|
|
|
| Zusammenfassung |
| Calibration and prediction using hydrological models has received tremendous attention in the literature. Calibration based on streamflow signatures, such as flow duration curves, is of particular interest - it offers fascinating opportunities to capture hydrological characteristics of interest and to undertake calibration in data-sparse conditions. Despite its clear appeal, signature calibration requires careful development and implementation to produce meaningful results, especially if reliable uncertainty estimates are desired.
This talk provides a Bayesian perspective on hydrological calibration using streamflow signatures, and its implementation using Approximate Bayesian Computation (ABC) algorithms. Following a brief theoretical expose, including the relationship to traditional calibration, we provide a series of case studies that elucidate the advantages and limitations of signature calibration under a variety of scenarios. |
|
|
|
|
|
|