Clim-final
Download as PPTX, PDF
0 likes235 views
This document discusses using maximum likelihood estimation to fit a generalized extreme value distribution to annual maximum precipitation data from Burke Garden, VA for two time periods. The results show similar location, scale, and shape parameter estimates for both periods, indicating no significant change in the distribution of extreme precipitation events over time.
![GEV Distribution:
family of continuous probability distributions
developed within extreme value theory
unites the Gumbel, Fr辿chet and Weibull distributions
AKA: type I, II and III
F(x;亮,,両)=exp{[1+両(
モ
)]}
1
is the location parameter
> 0 the scale parameter and 両 the shape parameter.](https://image.slidesharecdn.com/4b653301-4796-4065-8616-9c790a5c60bc-161020214730/85/Clim-final-5-320.jpg)
Clim-final
- 1. Estimation of Generalized Extreme Value Distribution: Maximum-likelihood Method Case study: Precipitation Annual Maximum Series Burke Garden, VA Ako Heidari
- 2. Extreme Value Theory Branch Of Statistics Dealing With The Extreme Deviations From The Median Of Probability Distributions Try To Assess The Probability Of Events That Are More Extreme Than Any Previously Observed Based On Given Ordered Sample
- 3. Paradigm: Dealing With AMS Model Development Idealized Model Wastage
- 4. GEV Applications: Used In Many Disciplines Specially In Hydrology And Meteorology Wind Engineering Management Strategy Biomedical Data Process Thermodynamic Of Earthquake
- 5. GEV Distribution: family of continuous probability distributions developed within extreme value theory unites the Gumbel, Fr辿chet and Weibull distributions AKA: type I, II and III F(x;亮,,両)=exp{[1+両( モ )]} 1 is the location parameter > 0 the scale parameter and 両 the shape parameter.
- 8. Different Method: Method Of Moments Maximum Likelihood Likelihood Function: ; = ; Is Unique In Adaptability To Model Changing
- 9. R packages for extreme values Ismev package extReme Fetdvcommand
- 10. Functions for extreme value distributions Extends simulation, distribution, quantile and density functions univariate and multivariate data parametric extreme value distributions A. G. Stephenson. evd: Extreme Value Distributions. R News, 2(2):31-32, June 2002. URL: http://CRAN.R-project.org/doc/Rnews/ Package: evd
- 11. Function: fgev Maximum-likelihood Fitting Of The Generalized Extreme Value Distribution
- 12. Data: T-S1: 1896-1949 T-S2: 1950-2000
- 13. Results: T-S1
- 14. Results: T-S2
- 15. Results: T-S 1 Location: 1.95 scale :0.52 shape :-0.12 T-S 2 Location: 1.99 scale :0.57 shape :-0.06
- 16. Results: Standard Location: 0.080 scale :0.058 shape :0.10 Error Location: 0.086 scale :0.06 shape :0.08
- 17. Answer: 1~ 1, 裡1 2~ 2, 裡2 (1 2) (裡1 + 2)1(1 2)~ 2
- 18. Summary Question: Estimating Extreme Values For A Specific Time Series Methodology: Maximum Likelihood, Generalized Extreme Value Answer: No Significant Change