Quantiles of the Neyman-Scott Rectangular Pulse Rainfall Model for Hydrologic Design
|
Accession number;07A0041888
|
| Title;Quantiles of the Neyman-Scott Rectangular Pulse Rainfall Model for Hydrologic Design |
| Author;
MONDONEDO CARLO ARTURO
(Kyoto Univ.)
TACHIKAWA YASUTO
(Kyoto Univ.)
TAKARA KAORU
(Kyoto Univ.)
|
Journal Title;Abstracts for Annuals. Disaster Prevention Research Institute, Kyoto University (CD-ROM)
|
Journal Code:S0431B
|
ISSN:
|
|
VOL.;NO.49;PAGE.NO.49B,83-91(2006)
|
| Figure&Table&Reference;FIG.2, TBL.4, REF.8 |
| Pub. Country;Japan |
| Language;English |
| Abstract;Following the clustered Poisson point process approach, we applied a Neyman-Scott rainfall Model (NSM) to generate synthetic extreme value rainfall depths. The main motivation for this study is on the acceptability of synthetic quantile rainfall depths as a possible decision-making aid in hydrologic design in lieu of historical records of appropriate length. Unlike previous studies, the emphasis here is to access the ability of the NSM to preserve the historical quantile rainfall depths of 1-hour and 24-hour duration such that longer NSM rainfall records, although synthetic, can be used as reliable bases of quantile events in hydraulic structure design (i.e.: impounding structures, sewer systems, etc). The historical data used in this study were obtained from 16 yearly records (1988 to 2003) of hourly and daily rainfall taken from the Kamishiiba Observatory in Japan. Following some stationarity assumptions inherent to NSM, it is necessary to obtain parameter sets for each month. Specifically, on a monthly basis, five NSM parameters were obtained by optimizing (by the Levenberg-Marquardt method) an objective function based on historical moments (namely mean, variance, covariance lag 1, etc.) and equivalent NSM moments, which were expressions in terms of the target parameters. By limiting historical information to five moments, the parameter sets obtained for several months (as of this writing) yielded adequate hourly quantiles and poor daily quantiles. Based on the framework of the NSM, it may be necessary to supply more historical moments (hence more NSM equations in the objective function) so that the five NSM parameters can capture more of the properties of the historical data. The determination of this ideal set of historical moments is the current pursuit of the authors and will appear in sequels of this paper. (author abst.) |
|
|
|
Related Articles;
|
|
