Impacts of Future Deforestation and Climate Change on the Hydrology of the Amazon Basin: a Multi-Model Analysis with a New Set of Land-Cover Change Scenarios

Matthieu Guimberteau,Philippe Ciais,Agnes Ducharne,Juan Pablo Boisier,Ana Paula Dutra Aguiar,Hester Biemans,Hannes De Deurwaerder,David Galbraith,Bart Kruijt,Fanny Langerwisch,German Poveda,Anja Rammig,Daniel Andres Rodriguez,Graciela Tejada,Kirsten Thonicke,Celso Von Randow,Rita C. S. Von Randow,Ke Zhang,Hans Verbeeck

DOI: https://doi.org/10.5194/hess-21-1455-2017

IF: 6.3

2017-01-01

Hydrology and Earth System Sciences

Abstract:Deforestation in Amazon is expected to decrease evapotranspiration (ET) and to increase soil moisture and river discharge under prevailing energy-limited conditions. The magnitude and sign of the response of ET to deforestation depend both on the magnitude and regional patterns of land-cover change (LCC), as well as on climate change and CO2 levels. On the one hand, elevated CO2 decreases leaf-scale transpiration, but this effect could be offset by increased foliar area density. Using three regional LCC scenarios specifically established for the Brazilian and Bolivian Amazon, we investigate the impacts of climate change and deforestation on the surface hydrology of the Amazon Basin for this century, taking 2009 as a reference. For each LCC scenario, three land surface models (LSMs), LPJmL-DGVM, INLAND-DGVM and ORCHIDEE, are forced by bias-corrected climate simulated by three general circulation models (GCMs) of the IPCC 4th Assessment Report (AR4). On average, over the Amazon Basin with no deforestation, the GCM results indicate a temperature increase of 3.3 °C by 2100 which drives up the evaporative demand, whereby precipitation increases by 8.5 %, with a large uncertainty across GCMs. In the case of no deforestation, we found that ET and runoff increase by 5.0 and 14 %, respectively. However, in south-east Amazonia, precipitation decreases by 10 % at the end of the dry season and the three LSMs produce a 6 % decrease of ET, which is less than precipitation, so that runoff decreases by 22 %. For instance, the minimum river discharge of the Rio Tapajós is reduced by 31 % in 2100. To study the additional effect of deforestation, we prescribed to the LSMs three contrasted LCC scenarios, with a forest decline going from 7 to 34 % over this century. All three scenarios partly offset the climate-induced increase of ET, and runoff increases over the entire Amazon. In the south-east, however, deforestation amplifies the decrease of ET at the end of dry season, leading to a large increase of runoff (up to +27 % in the extreme deforestation case), offsetting the negative effect of climate change, thus balancing the decrease of low flows in the Rio Tapajós. These projections are associated with large uncertainties, which we attribute separately to the differences in LSMs, GCMs and to the uncertain range of deforestation. At the subcatchment scale, the uncertainty range on ET changes is shown to first depend on GCMs, while the uncertainty of runoff projections is predominantly induced by LSM structural differences. By contrast, we found that the uncertainty in both ET and runoff changes attributable to uncertain future deforestation is low.