Regional Frequency Analysis of Maximum 24-Hour Precipitation in Arid Lands Using The L-Moments Approach (Case Study: Yazd Province)

Document Type : Original Article

Authors

1 Ph. D of Watershed Management Engineering, Faculty of Natural Resources and Desert Studies, Yazd University, Yazd, Iran.

2 Proffesor, Faculty of Natural Resources and Desert Studies, Yazd University, Yazd, Iran.

Abstract

Estimation of the maximum 24-hour precipitation is one of the important parameters in irrigation and drainage projects, design of hydraulic structures, and watershed management. This rainfall data is used to calculate short-term rainfall and design flood. At-site and regional frequency analysis of the maximum 24-hour precipitation climate projects are essential and can be a positive and efficient step for analyzing the storms and productive floods. In this research, the L-moments approach is used for regional frequency analysis of the maximum 24-hour precipitation of 18 stations in Yazd province. The Ward clustering algorithm is used to identify homogeneous region. The selected stations are classified into two groups, and each region is tested using Hosking and Wallis Homogeneity critter. After ensuring the homogeneity of the regions, by using the goodness of fit test, Generalized Extreme Value distribution is suggested as the best fit distribution in both regions. The parameters of the distribution were calculated by the L-moments method for both regional and at-site approach. The performance of the two approaches was compared using RMSE and RRMSE statistic indices. Comparison of the regional approach with the at-site approach showed that both estimates at short return periods (less than 10 years) are not much different and computational errors are also less. However, there is so much difference at longer return periods. It can be stated that the regional approach in the first and second homogeneity regions (80% and 70%, respectively) is more appropriate rather than the at-site approach for maximum 24-hour precipitation frequency analysis.

Keywords

Desert Management
Volume 8, Issue 15 - Serial Number 15
12 Articles
September 2020
Pages 37-52

Files

History

  • Receive Date: 29 November 2019
  • Revise Date: 15 January 2020
  • Accept Date: 16 January 2020

Share

How to cite

Statistics