 |
| Titel |
Development of a method of robust rain gauge network optimization based on intensity-duration-frequency results |
| VerfasserIn |
A. Chebbi, Z. K. Bargaoui, M. da Conceição Cunha |
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| ISSN |
1027-5606
|
| Digitales Dokument |
URL |
| Erschienen |
In: Hydrology and Earth System Sciences ; 17, no. 10 ; Nr. 17, no. 10 (2013-10-29), S.4259-4268 |
| Datensatznummer |
250085977
|
| Publikation (Nr.) |
 copernicus.org/hess-17-4259-2013.pdf |
|
|
|
|
|
| Zusammenfassung |
| Based on rainfall intensity-duration-frequency (IDF) curves, fitted in
several locations of a given area, a robust optimization approach is
proposed to identify the best locations to install new rain gauges. The
advantage of robust optimization is that the resulting design solutions
yield networks which behave acceptably under hydrological variability.
Robust optimization can overcome the problem of selecting representative
rainfall events when building the optimization process. This paper reports
an original approach based on Montana IDF model parameters. The latter are
assumed to be geostatistical variables, and their spatial interdependence is
taken into account through the adoption of cross-variograms in the kriging
process. The problem of optimally locating a fixed number of new monitoring
stations based on an existing rain gauge network is addressed. The objective
function is based on the mean spatial kriging variance and rainfall
variogram structure using a variance-reduction method. Hydrological
variability was taken into account by considering and implementing several
return periods to define the robust objective function. Variance
minimization is performed using a simulated annealing algorithm. In
addition, knowledge of the time horizon is needed for the computation of the
robust objective function. A short- and a long-term horizon were studied, and
optimal networks are identified for each. The method developed is applied to
north Tunisia (area = 21 000 km2). Data inputs for the
variogram analysis were IDF curves provided by the hydrological bureau and
available for 14 tipping bucket type rain gauges. The recording period was
from 1962 to 2001, depending on the station. The study concerns an imaginary
network augmentation based on the network configuration in 1973, which is a
very significant year in Tunisia because there was an exceptional regional
flood event in March 1973. This network consisted of 13 stations and did not
meet World Meteorological Organization (WMO) recommendations for the minimum
spatial density. Therefore, it is proposed to augment it by 25,
50, 100 and 160% virtually, which is the rate that would meet WMO
requirements. Results suggest that for a given augmentation robust networks
remain stable overall for the two time horizons. |
| |
|
| Teil von |
|
|
|
|
|
|
|
|