Empirically Derived Sensitivity of Vegetation to Climate Across the Globe

Abstract

To predict the response of vegetation to climate change, we must understand the physiological processes controlling productivity across large spatial scales, encompassing global climate space. To date there is not a fully empirical map of vegetation sensitivity to climate at the global scale. We use the response of satellite-based greenness (from Normalized Difference Vegetation Index) to inter-annual climate variations in surface air temperature (from ERA-Interim) and precipitation (from Global Precipitation Climatology Project) to derive the sensitivity of vegetation to temperature to infer mechanisms of climate constraint on vegetation productivity across the globe as represented by greenness. We focus on how the sensitivity of vegetation to temperature varies across climate space, finding that it is modulated by a balance of resources. The majority of grid cells in simultaneously warm (above ~14 \degree C) and dry (below ~1000 mm/year rainfall) conditions have negative vegetation sensitivity to temperature (browner in warm years) while at places with cooler temperatures the vegetation sensitivity is generally positive (greener in warm years). The mean annual temperature boundary between positive and negative sensitivities changes by 9 degrees C depending on how much rainfall a place receives. At very high rainfall levels (beyond 3000 mm/year), even the hottest vegetated places on Earth have positive sensitivity to mean annual temperature. The positive temperature sensitivity of these warm wet ecosystems suggests that water allows buffering against damaging maximum temperatures and that these ecosystems may actually benefit from near term warming on the scale of inter-annual variations of temperature.