Statistical modeling of annual maximum river flow in Sabah
This study aimed is to model the annual maximum river flow in several sites in Sabah with small sample sizes using the Generalised Extreme Value (GEV) distribution. Previous studies had shown that the standard method of maximum likelihood estimates produced poor estimations of GEV parameters and qua...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | en en |
| Published: |
2017
|
| Subjects: | |
| Online Access: | https://eprints.ums.edu.my/id/eprint/43455/1/24%20PAGES.pdf https://eprints.ums.edu.my/id/eprint/43455/2/FULLTEXT.pdf https://eprints.ums.edu.my/id/eprint/43455/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This study aimed is to model the annual maximum river flow in several sites in Sabah with small sample sizes using the Generalised Extreme Value (GEV) distribution. Previous studies had shown that the standard method of maximum likelihood estimates produced poor estimations of GEV parameters and quantiles for small sample sizes. The penalized maximum likelihood estimation was method was implemented as an alternative method to improve the inference over the standard method and retain model flexibility. The results of annual maximum flow modeling in Sabah using maximum likelihood estimation (MLE) and penalized maximum likelihood estimation (PMLE) are illustrated in the form of graphs for comparative purposes. Results show the implementation of PMLE had the same effect on the GEV parameter estimates as suggested by previous studies. In this study, GEV distribution was fitted independently to model data of river flow at each sites to avoid extreme value complex modeling. Since this approach violated the condition of spatial analysis, the adjusted standard error was considered to rectify the wrong assumption of marginal approach. This resulted in an appropriate corrected variance of the generalized extreme value parameters. The study also found many of the rivers in this study expected to exceed the maximum level once every 100 years. |
|---|