مهديه پهلوان

Climate Pattern Analysis in Southwestern Iran an‎d Adjacent Regions Using ERA5 an‎d Integrated Data Mining Approaches

Southwestern Iran an‎d its neighboring regions are highly sensitive to climate variability, which directly affects agriculture, water resources, an‎d disaster management. eva‎luating these dynamics is challenging due to the limited spatial coverage of observations an‎d the complex interplay between regional an‎d large-scale drivers. In this study, multi-decadal ERA5 records of temperature, precipitation, an‎d circulation indices were preprocessed an‎d analyzed using a combination of statistical an‎d data mining approaches to identify an‎d classify dominant climate patterns across the region. The analytical framework included principal component analysis, clustering, an‎d classification methods such as ran‎dom forests an‎d support vector machines, with model skill assessed through established statistical metrics. The results reveal distinct spatial–temporal regimes: precipitation variability aligns with large-scale circulation, while temperature extremes persist locally. Integrated frameworks consistently yielded more stable an‎d reliable classifications than individual models. Methodologically, this study highlights the strength of coupling reanalysis data with integrated data mining for robust regional climate diagnostics. Practically, the results provide actionable insights for drought preparedness, agricultural planning, an‎d adaptation to climate change in one of the Middle East’s most vulnerable regions.

Climate variability , ERA5 reanalysis , Data mining approaches , Precipitation forecasting , Persian Gulf SST , Southwestern Iran