Higher temperature variability reduces temperature sensitivity of vegetation growth in Northern Hemisphere

Interannual air temperature variability has changed over some regions in Northern Hemisphere (NH), accompanying with climate warming. However, whether and to what extent it regulates the interannual sensitivity of vegetation growth to temperature variability (i.e., interannual temperature sensitivity)-one central issue in understanding and predicting the responses of vegetation growth to changing climate-still remains poorly quantified and understood. Here we quantify the relationships between the interannual temperature sensitivity of mean growing-season (April-October) normalized difference vegetation index (NDVI) and ecosystem model simulations of gross primary productivity (GPP), and variability in mean growing-season temperature for forest, shrub, and grass over NH. We find that higher interannual variability in mean growing-season temperature leads to consistent decrease in interannual temperature sensitivity of mean growing-season NDVI among all vegetation types but not in model simulations of GPP. Drier condition associates with similar to 130 +/- 150% further decrease in interannual temperature sensitivity of mean growing-season NDVI by temperature variability in forest and shrub. These results illustrate that varying temperature variability can significantly regulate the interannual temperature sensitivity of vegetation growth over NH, interacted with drought variability and nonlinear responses of photosynthesis to temperature. Our findings call for an improved characterization of the nonlinear effects of temperature variability on vegetation growth within global ecosystem models.