 |
| Titel |
A surrogate modelling framework for the optimal deployment of check dams in erosion-prone areas |
| VerfasserIn |
Debasish Pal, Honglei Tang, Stefano Galelli, Qihua Ran |
| Konferenz |
EGU General Assembly 2017
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 19 (2017) |
| Datensatznummer |
250146968
|
| Publikation (Nr.) |
 EGU/EGU2017-11043.pdf |
|
|
|
|
|
| Zusammenfassung |
| Despite the great progresses made in the last decades, the control of soil erosion still remains
a key challenge for land-use planning. The nonlinear interactions between hydrologic and
morphologic processes and increase in extreme rainfall events predicted with climatic change
create new areas of concern and make the problem unresolved. Spatially distributed models
are a useful tool for modelling such processes and assessing the effect of large-scale
engineering measures, but their computational requests prevent the resolution of problems
requiring several model evaluations—sensitivity analysis or optimization, for instance. In
this study, we tackle this problem by developing a surrogate modelling framework
for the optimal deployment of check dams. The framework combines a spatially
distributed model (WaTEM/SEDEM), a multi-objective evolutionary algorithm and
artificial neural networks as surrogate model. We test the framework on Shejiagou
catchment—a 14 km2 area located in the Loess Plateau, China—where we optimize check
dam locations by maximizing the mass of sediments retained in the catchment and
minimizing the total number of dams. Preliminary results show that the performance of
the existing check dam system could be improved by changing the dam locations. |
|
|
|
|
|
|