Alexandria Engineering Journal (May 2024)
Analysis of recent decade rainfall data with new exponential-exponential distribution: Inference and applications
Abstract
This study introduces the New Exponential-Exponential Distribution (NEED) within the broader New Exponential-Generating (NE-G) family of distributions, targeting enhancements in rainfall data analysis. The significance of this research lies in addressing the need for sophisticated statistical models to accurately capture the complex variability of rainfall patterns, which are critical for effective environmental planning and disaster management. Aiming to refine rainfall data modeling, we employ the NEED model, emphasizing its application across diverse climatic conditions. Our methodology encompasses a comprehensive evaluation of seven distinct parameter estimation techniques, with a particular focus on the Anderson-Darling and maximum product spacing methods. These were selected based on their performance in minimizing bias and mean square error, assessed through a rigorous Monte-Carlo simulation study. Additionally, the study utilizes rainfall data from various geographical regions to validate the model's efficacy. The major conclusion of our investigation is the demonstrable superiority of NEED over traditional models in fitting rainfall data, as evidenced by its enhanced predictive accuracy. This outcome not only contributes to the theoretical advancements in statistical meteorology but also offers practical methodologies for improved weather forecasting. The integration of NEED with contemporary machine learning algorithms further suggests potential for groundbreaking applications in climate science and water resource management.
