Tree-ring intra-annual stable isotopes (delta C-13 and delta O-18) are powerful tools for revealing plant ecophysiological responses to climatic extremes. We analyzed interannual and fine-scale intra-annual variability of tree-ring delta C-13 and delta O-18 in Chinese red pine (Pinus massoniana) from southeastern China to explore environmental drivers and potential trade-offs between the main physiological controls. We show that wet season relative humidity (May-October RH) drove interannual variability of delta O-18 and intra-annual variability of tree-ring delta O-18. It also drove intra-annual variability of tree-ring delta C-13, whereas interannual variability was mainly controlled by February-May temperature and September-October RH. Furthermore, intra-annual tree-ring delta O-18 variability was larger during wet years compared with dry years, whereas delta C-13 variability was lower during wet years compared with dry years. As a result of these differences in intra-annual variability amplitude, process-based models (we used the Roden model for delta O-18 and the Farquhar model for delta C-13) captured the intra-annual delta O-18 pattern better in wet years compared with dry years, whereas intra-annual delta C-13 pattern was better simulated in dry years compared with wet years. This result suggests a potential asymmetric bias in process-based models in capturing the interplay of the different mechanistic processes (i.e., isotopic source and leaf-level enrichment) operating in dry versus wet years. We therefore propose an intra-annual conceptual model considering a dynamic trade-off between the isotopic source and leaf-level enrichment in different tree-ring parts to understand how climate and ecophysiological processes drive intra-annual tree-ring stable isotopic variability under humid climate conditions.

Changes in snow precipitation at high latitudes can significantly affect permafrost thermal conditions and thaw depth, potentially exposing more carbon-laden soil to microbial decomposition. A fully coupled process-based surface/subsurface thermal hydrology model with surface energy balance is used to analyze the impact of intra-annual variability in snow on permafrost thermal regime and the active layer thickness. In the four numerical scenarios considered, simulations were forced by the same meteorological data, except the snow precipitation, which was systematically altered to change timing of snowfall. The scenarios represent subtle shifts in snow timing, but the snow onset/melt days, the end of winter snowpack depth, and total annual snow precipitation are unchanged among scenarios. The simulations show that small shifts in the timing of snow accumulation can have significant effects on subsurface thermal conditions leading to active layer deepening and even talik formation when snowfall arrives earlier in the winter. The shifts in snow timing have a stronger impact on wetter regions, especially soil underneath small ponds, as compared to drained regions. This study highlights the importance of understanding potential changes in winter precipitation patterns for reliable projections of active-layer thickness in a changing Arctic climate.