Code Repository for Investigating Anthropogenic Influences on Arctic Sea-Ice Extent and AMOC Decoupling

This collection of R and Python scripts provides a computational framework for analyzing timeseries relationships among Arctic seaice extent, the Atlantic Meridional Overturning Circulation (AMOC), and ocean–atmosphere variability. The codes include procedures for data preprocessing, timeseries imputation using Kalman filtering, and ensemble reconstruction of seaice extent from NetCDF datasets. Several modeling approaches are implemented, including neural networks, random forest analysis, and ant colony optimization to improve model training and evaluate the relative importance of climatic variables. In addition, statistical analyses such as crosscorrelation, detrended composites, and wavelet coherence are applied to examine temporal relationships between seaice variability, sea surface temperature, and AMOC fluctuations. Visualization routines are also included to map Arctic observation stations and illustrate key patterns in the data. Together, these scripts provide a practical workflow for investigating longterm Arctic climate variability and the potential influence of external forcing on seaice–ocean interactions.