At-site and regional frequency analysis of extreme rainfall modelling in Peninsular Malaysia
The objective of this study is to determine the appropriate probability distribution and to estimate the rainfall quantiles for monthly maxima daily rainfall data from the year 2005 to 2019 for 30 rain gauge stations in Peninsular Malaysia based on at-site and regional hydro- logical frequency analy...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Published: |
UniMAP Press
2022
|
| Online Access: | http://psasir.upm.edu.my/id/eprint/100485/ https://amci.unimap.edu.my/2022-ready-issue-2/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The objective of this study is to determine the appropriate probability distribution and to estimate the rainfall quantiles for monthly maxima daily rainfall data from the year 2005 to 2019 for 30 rain gauge stations in Peninsular Malaysia based on at-site and regional hydro- logical frequency analysis. Five three-parameters probability distributions were considered in this study i.e generalized extreme value (GEV), generalized Pareto (GPA), generalized logistic (GLO), generalized normal (GNO) and Pearson Type III (PE3). Cluster analysis based on Ward’s method was used to identify the homogeneous region which is further con- firmed by discordancy and heterogeneity measures. The L-moment method of estimation is used to estimate the parameter of a model. The L-moment ratio diagrams and Monte Carlo simulation based on Z DIST were used to assess the goodness of the fitted model. Results obtained by traditional at-site frequency analysis are compared with those obtained by regional frequency analysis. The results showed that the best probability distribution for monthly maxima daily rainfall data at each station and the ones of corresponding homogeneous regions obtained by regional frequency analysis were not necessarily consistent. Although the optimal probability distribution may vary according to the stations, in most cases, data for most of the stations are found to follow the generalized logistic distribution while for the regional study, rainfall data for most of the regions are well fitted by the generalized extreme value distribution. Meanwhile, the uncertainty due to quantile estimates for at-site and regional data is considerably low for less than 100 years return period but high for more than 100 years. |
|---|