Related Experiment Videos
On the properties of stochastic intermittency in rainfall processes
A Molini1, Barbera P La, L G Lanza
1Department of Environmental Engineering, University of Genova, Italy. molini@diam.unige.it
Summary
This study models rainfall events using geometric and statistical analysis of rain intermittency. It reveals an invariant probability distribution for no-rain periods within events, aiding rainfall prediction.
Area of Science:
- Hydrology
- Meteorology
- Statistical analysis of time series
Background:
- Rainfall intermittency presents challenges in hydrological and meteorological modeling.
- Understanding the statistical properties of rain and no-rain periods is crucial for accurate rainfall event characterization.
Purpose of the Study:
- To propose a mixed approach for modeling rainfall events.
- To analyze the geometrical and statistical properties of rain intermittency.
- To investigate the predictability of rainfall based on no-rain period distributions.
Main Methods:
- Analysis of geometrical and statistical properties of rain intermittency.
- Binary decomposition of the rain signal.
- Geometrical filtering to separate internal intermittency from inter-arrival processes.
- Analysis of no-rain periods distribution within rainfall events.
Main Results:
- An invariant probability distribution for no-rain periods within rainfall events was identified.
- This distribution was found to be consistent across different aggregation levels.
- The identified distribution shows satisfactory agreement with typical extreme value distributions.
Conclusions:
- The proposed mixed approach effectively models rainfall events by analyzing intermittency and no-rain periods.
- The invariant probability distribution of no-rain periods offers insights into rainfall predictability.
- Findings contribute to a better understanding of rainfall event dynamics and statistical properties.