Study region: Northwest Iran is a significant agricultural area, particularly for rainfed farming. This region is not only vital for food production but also supports a large population that relies heavily on seasonal precipitation. Therefore, the ability to accurately predict precipitation is essential for improved planning in agriculture and water management. Study focus: This study aims to develop a model for predicting precipitation patterns during autumn (October–November–December) in northwest Iran. To achieve this, we employed a method known as D-Vine copula quantile regression, which facilitates the exploration of the relationship between autumn precipitation and large-scale climate signals. To identify the most significant climate indicators, we utilized a statistical measure called Kendall’s tau. Additionally, we implemented a five-fold cross-validation approach to evaluate the models’ predictive performance on data not used during training. New hydrological insights for the region: Our results indicated that two climate indices—NINO3.4 and the Southern Oscillation Index (SOI)—exhibited the strongest correlation with autumn precipitation in the region. Among the seven models we developed, the most effective one incorporated NINO3.4, SOI, and the North Atlantic Oscillation (NAO). This model yielded the most precise predictions with the least amount of error. These findings demonstrate that leveraging global climate patterns to predict regional precipitation can be a powerful tool for areas such as northwest Iran, where agriculture heavily depends on seasonal rainfall.

Keywords