Integration of the analytic hierarchy process and Monte Carlo feature selection for a spatiotemporal weighting scheme in meteorological drought monitoring
Muhammad Hamza, Rizwan Niaz, Ibrahim Nafisah, Luca Di Persio, Abdulkareem M. Basheer, Mohammed M. A. Almazah
Abstract
There is a compelling need to integrate different drought indices values because these indices utilize various distributions for computation, resulting in different outcomes when the underlying distributions change. This inconsistency highlights the need for a unified approach to assessing drought conditions rather than relying on multiple indices. In response to this challenge, a new framework has been developed that combines existing drought measures into a single Comprehensive Drought Index (CDI). This innovative approach establishes a robust and inclusive method for comparing drought classifications by applying weights to the index values derived from multiple established indices. Validation results demonstrate a strong correlation with these existing measures, underscoring the effectiveness of this comprehensive method. The CDI offers a more efficient tool for improved drought management and response strategies by providing a more nuanced and cohesive understanding of drought complexities.