Comparison of the performance of the extreme value type 1 (EV1) and the rayleigh distribution
The suitability of the Extreme Value Type I (EV1) and the Rayleigh distributions for flood frequency analysis is discussed. The distributions were applied to the annual maximum flood series generated by first-order autoregressive (AR(1)) process of length 90 for 1000 years. Here two methods of param...
Saved in:
| 主要作者: | |
|---|---|
| 格式: | Thesis |
| 出版: |
2010
|
| 主題: | |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 總結: | The suitability of the Extreme Value Type I (EV1) and the Rayleigh distributions for flood frequency analysis is discussed. The distributions were applied to the annual maximum flood series generated by first-order autoregressive (AR(1)) process of length 90 for 1000 years. Here two methods of parameter estimation namely the maximum likelihood and the probability-weighted moments estimators were used. The probability of exceedance of the 90th and the 99th percentiles were also considered. It was found that the EV1 distribution fits the data well compared to the Rayleigh distribution. The risks estimates calculated using the EV1 distribution agrees with the empirical distribution. |
|---|