The impact of land-use change on the sensitivity of terrestrial productivity to precipitation variability: a modelling approach

Abstract. Larger climate variability and more frequent extreme events (e.g. droughts) are expected to occur. Hence, assessing the sensitivity (response) of terrestrial net primary productivity (NPP) to climate variability is crucial for future estimations of terrestrial carbon sequestration. We hypothesize that land-use change (LUC) can modify this sensitivity, defined here as the anomaly correlation between annual NPP and precipitation in the growing season. To assess this changing sensitivity, we use the LPJmL model driven by observed climate variability and two different land-use sets.

We find that the effect of LUC is mostly extended in semi-arid grasslands: sensitivity increases with conversion into rainfed crops, but it does not change for managed pasture expansion. Reforestation in temperate regions generally decreases sensitivity: trees are able to respond later to water deficit due to their deep root system. The expansion of irrigation in northern India reduces sensitivity since water availability is not a limiting factor for crop growth. Although we are able to identify spatial patterns of changes in NPP-precipitation sensitivity, the significance of our results focusing at isolation of the land-use effect is limited.