Identifying climate and human impact trends in streamflow : A case study in Uruguay

Abstract

Land use change is an important driver of trends in streamflow. However, the effects are often difficult to disentangle from climate effects. The aim of this paper is to demonstrate that trends in streamflow can be identified by analysing residuals of rainfall-runoff simulations using a Generalized Additive Mixed Model. This assumes that the rainfall-runoff model removes the average climate forcing from streamflow. The case study involves the Santa Lucía river (Uruguay), the GR4J rainfall-runoff model, three nested catchments ranging from 690 to 4900 km2 and 35 years of observations (1981–2016). Two exogenous variables were considered to influence the streamflow. Using satellite data, growth in forest cover was identified, while the growth in water licenses was obtained from the water authority. Depending on the catchment, effects of land use change differ, with the largest catchment most impacted by afforestation, while the middle size catchment was more influenced by the growth in water licenses.