Spatial-temporal isotope patterns in European winter precipitation reflect atmospheric circulation modes over North Atlantic and Arctic Oceans

To investigate the spatial and temporal modes in European winter precipitation isotopes and their climatic controls, we performed an empirical orthogonal function (EOF) analysis on the Piso.AI dataset, a recently released machine-learning-based monthly gridded precipitation isoscape across Europe. The first two EOF modes, EOF1 and EOF2, explain 56.83 % and 12.94 % of the total variance, respectively. EOF1 shows spatially uniform changes in δD, influenced by regional variations in temperature and water vapor transport. EOF2 reveals a dipole δD pattern between northern and southern Europe, reflecting regional differences in climate and moisture sources. These spatial-temporal patterns are corroborated by results from δ¹⁸O and isotope-enabled model simulations, enhancing the robustness of our findings. The Arctic Oscillation (AO) and North Atlantic Oscillation (NAO) significantly impact the first δD mode through their modulation of European temperatures and the strength of water vapor transport from lower latitudes, with a strong correlation of 0.86 for AO. The second mode captures unique climate variability independent of NAO and AO. These findings enhance understanding of atmospheric circulation and water vapor transport processes that control isotope changes across Europe. Furthermore, they offer insights into paleoclimate reconstruction, highlighting the potential for using spatially distributed isotopic records to reconstruct past AO activities. Additionally, they suggest the possibility of developing new climate reconstructions by leveraging the differences in isotopic signatures between northern and southern Europe.