Related Experiment Videos
Selecting a probability distribution for extreme rainfall series in Malaysia
M D Zalina1, M N M Desa, V T V Nguyen
1Faculty of Science, Universiti Teknologi Malaysia, Johor. zalina@mel.fs.utm.my
Summary
The Generalised Extreme Value (GEV) distribution is the most accurate for estimating Malaysia's maximum annual rainfall. This study compared eight statistical models to determine the best fit for rainfall data.
Area of Science:
- Hydrology
- Statistical modeling
- Extreme value theory
Background:
- Accurate estimation of maximum rainfall is crucial for water resource management and disaster preparedness in Malaysia.
- Several statistical distributions are available for modeling extreme hydrological events, but their suitability varies by region.
Purpose of the Study:
- To comparatively assess eight candidate probability distributions for modeling annual maximum rainfall in Malaysia.
- To identify the most appropriate distribution for reliable extreme rainfall estimation in the region.
Main Methods:
- Utilized annual maximum rainfall series (1-hour resolution) from seventeen stations in Peninsular Malaysia (23-28 years of data).
- Employed the L-moment method for parameter estimation.
- Assessed model performance using Probability Plot Correlation Coefficient, root mean squared error, relative root mean squared error, and maximum absolute deviation.
- Investigated extrapolative ability using bootstrap resampling.
Main Results:
- The Generalised Extreme Value (GEV) distribution demonstrated superior performance compared to Gamma, Generalised Normal, Generalised Pareto, Gumbel, Log Pearson Type III, Pearson Type III, and Wakeby distributions.
- GEV provided the most accurate and reliable estimates for the annual maximum rainfall series.
Conclusions:
- The Generalised Extreme Value (GEV) distribution is recommended as the most suitable model for describing annual maximum rainfall in Malaysia.
- Findings support improved hydrological risk assessment and infrastructure design in the region.