 |
| Titel |
Propagation of hydro-meteorological uncertainty in a model cascade framework to inundation prediction |
| VerfasserIn |
J. P. Rodríguez-Rincón, A. Pedrozo-Acuña, J. A. Breña-Naranjo |
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| ISSN |
1027-5606
|
| Digitales Dokument |
URL |
| Erschienen |
In: Hydrology and Earth System Sciences ; 19, no. 7 ; Nr. 19, no. 7 (2015-07-01), S.2981-2998 |
| Datensatznummer |
250120756
|
| Publikation (Nr.) |
 copernicus.org/hess-19-2981-2015.pdf |
|
|
|
|
|
| Zusammenfassung |
| This investigation aims to study the propagation of meteorological
uncertainty within a cascade modelling approach to flood prediction. The
methodology was comprised of a numerical weather prediction (NWP) model, a
distributed rainfall–runoff model and a 2-D hydrodynamic model. The
uncertainty evaluation was carried out at the meteorological and hydrological
levels of the model chain, which enabled the investigation of how errors
that originated in the rainfall prediction interact at a catchment level and
propagate to an estimated inundation area and depth. For this, a hindcast
scenario is utilised removing non-behavioural ensemble members at each stage,
based on the fit with observed data. At the hydrodynamic level, an
uncertainty assessment was not incorporated; instead, the model was setup
following guidelines for the best possible representation of the case study.
The selected extreme event corresponds to a flood that took place in the
southeast of Mexico during November 2009, for which field data (e.g. rain
gauges; discharge) and satellite imagery were available. Uncertainty in the
meteorological model was estimated by means of a multi-physics ensemble
technique, which is designed to represent errors from our limited knowledge
of the processes generating precipitation. In the hydrological model, a
multi-response validation was implemented through the definition of six sets
of plausible parameters from past flood events. Precipitation fields from the
meteorological model were employed as input in a distributed hydrological
model, and resulting flood hydrographs were used as forcing conditions in the
2-D hydrodynamic model. The evolution of skill within the model cascade shows
a complex aggregation of errors between models, suggesting that in
valley-filling events hydro-meteorological uncertainty has a larger effect on
inundation depths than that observed in estimated flood inundation extents. |
| |
|
| Teil von |
|
|
|
|
|
|
|
|