Pieter J.M. de Laat¹⁾ and Issa Al-Nsour²⁾

¹⁾ Corresponding author

UNESCO-IHE Institute for Water Education, P.O. Box 3015, 2601 DA Delft, The Netherlands

Tel.: +31 15 2151826, Fax: +31 15 2122921

E-mail: p.delaat@unesco-ihe.org

²⁾ Ministry of Water and Irrigation, Amman, Jordan

E-mail: Essa_Al-Nosour@mwi.gov.jo

Abstract: For the design of small retention dams in the desert area of Jordan, rainfall and runoff were measured in a catchment of 37 km². In a period of two years eight runoff events were recorded from which a relation could be derived between rainfall intensity and runoff coefficient. Statistical analysis of time series of almost 60 years of daily rainfall values of stations in the vicinity of the catchment showed a significant downward trend of the annual daily extreme precipitation. In this study extreme rainfall events for various return periods were estimated from the last 30 years, which time series is stationary. The relation between rainfall intensity and runoff coefficient was then used the estimate peak runoff from extreme rainfall events.

A similar study was carried out in the 1990s in the same region based on a ten year shorter time series of rainfall data and in the absence of runoff measurements. Peak runoff was then estimated with the use of the Soil Conservation Service - Curve Number (CN) method from the extreme rainfall data derived from the shorter time series. Large differences in the predicted extreme discharges were found between the two methods. The estimated peak runoff for a 50-year event appears almost seven times larger than found in this study. It is shown that the difference may predominantly be attributed to the use of a non-stationary time series, which is caused by a variation or change in the climate.

Key words: climate variability/change, extreme discharge, Curve Number method, statistical analysis