A0539
Title: Extreme precipitation teleconnection analysis using asymptomatic dependence structure and clustering algorithms
Authors: Reza Nosratpour - RMIT University (Australia)
Laleh Tafakori - RMIT University (Australia) [presenting]
Mali Abdollahian - RMIT University (Australia)
Abstract: Extreme precipitation events result in significant social and economic losses worldwide each year, underscoring the need for a comprehensive understanding of their mechanisms to improve disaster preparedness and mitigation strategies. The focus is on the spatiotemporal dependency analysis of extreme precipitation, which is important for understanding the complex behavior of these events. The interconnected spatiotemporal dynamics of extreme precipitation are modeled on a continental scale by employing asymptotic dependence analysis and unsupervised clustering algorithms. In this model, the network of extremal coefficient output will be the input of the k-means clustering algorithms to build the structure of the teleconnections of the extreme precipitation events. Remote sensing observations play an important role by offering global coverage that enables the study of precipitation patterns, predictability, and dependency across land and ocean. Using remote sensing data allows fully covering the study area, enhancing the accuracy of predictive models and the breadth of the insights. The results of this research have the potential to refine precipitation forecasts and develop effective mitigation strategies for extreme precipitation events.
