LAUSR

Abstract Water management in the Caatinga/Atlantic forest ecotone in Brazil is critically dependent on better understanding of potential future changes in streamflow and sediment dynamics. This paper evaluates both the future impacts of land use and land cover (LULC) changes and the impacts of climate change on the streamflow and sediment yield in the Tapacura River basin in northeastern Brazil, using a novel combination of approaches. Projected climate data derived using global circulation model HadGEM2-ES were coupled to regional circulation model ETA-CPTEC/HadCM3 for two representative concentration pathways (RCP 4.5 and 8.5), with bias correction. Two future LULC scenarios were generated: (a) optimistic (current LULC), and (b) pessimistic (land use change trends continue), using the multilayer perceptron algorithm (MP). The Soil and Water Assessment Tool (SWAT) model was used to estimate future streamflow and erosion for different periods (2011–2040, 2041–2070 and 2071–2099). The SWAT model was calibrated for period of 1995–2003 and validated for 2004–2013. The results showed good accuracy in relation to R2, NSE and PBIAS for the calibration and validation of the runoff, as well as for the verification of the sediment yield. Simulations indicated significant increases in erosion for the pessimistic scenario under RCP 8.5, followed by the pessimistic scenario and RCP 4.5. Lower sediment yields occurred for the optimistic and RCP 8.5, with lower still for the optimistic and RCP 4.5. However, the latter estimates are still considerably higher than baseline conditions. Although higher flows are found for some scenarios, the increases in sediment yield have serious implications for reservoir siltation and storage reduction. Despite modeling uncertainty, the results demonstrate that the proposed methodology has promising scope to contextualize potentially significant regional hydrological changes which have implications for land and biodiversity management and the sustainability of water resources in the Caatinga/Atlantic forest ecotone.