My research focuses on reconstructing past climate variability through paleoclimate data assimilation by optimally fusing climate models with diverse natural proxy records, such as tree rings, corals, and ice cores. I work on the processing and integration of multi-proxy data, with an emphasis on developing adaptive algorithms to accommodate proxies with varying temporal resolutions. I also focus on improving proxy-climate interpretation and systematically evaluating how both statistical and process-based proxy modeling influence assimilation outcomes. By directly assimilating water isotope information, my research aims to provide novel constraints on regional hydroclimate dynamics that are not accessible through traditional assimilation targets.

Within the EUROpest project, I contribute to producing high-resolution temperature and precipitation reconstructions for 1300-1800 CE across Europe to support assessments of pest outbreak risks under past climate conditions. This involves curating and screening natural proxy records and historical documentary evidence, extracting reliable climate information at their true temporal resolution, and combining advanced AI techniques with data assimilation to generate high-quality reconstructions. The resulting reconstructions are then validated against existing products derived from various approaches and compared with newly developed high-resolution paleoclimate simulations to demonstrate their robustness and reliability.