The impact of land use and climate change on late Holocene and future suspended sediment yield of the Meuse catchment

Abstract

In this study we investigate the relative importance of changes in land use and climate on suspended sediment yield (SY) on millennial timescales in the Meuse basin. We use a spatially distributed soil erosion and sediment delivery model (WATEM/SEDEM) to simulate SY in three time-periods: 4000–3000 BP (minimal anthropogenic influence); 1000–2000 AD (includes land use and climate change); and the 21st Century. Changes in climate are based on climate model output (ECBilt-CLIO-VECODE). For the 21st Century the model is forced according to two emission scenarios of the Intergovernmental Panel on Climate Change (IPCC), namely the SRES scenarios A2 and B1. These scenarios lie towards the higher and lower end of the full IPCC scenario range respectively. For 4000–3000 BP the basin is assumed to be almost fully forested; for 1000–2000 AD land use is reconstructed using CORINE data, historical sources, and land use modelling; and for the 21st Century land use is based on the European land use change project EURURALIS. Whilst rainfall erosivity increases by only 3% between 4000–3000 BP and 1000–2000 AD, SY increases from ca. 92 000 Mg a − 1 to ca. 306 000 Mg a − 1 . This model prediction is in agreement with the limited regional multi-proxy data available. Our simulations show that almost all of this increase is due to the conversion of forest to agricultural land. Over the period 1000–1900 AD, SY shows a significant increasing trend, with a peak of ca. 388 000 Mg a − 1 in the 19th Century (due to continuing deforestation). In the 20th Century, reforestation and rapid urbanisation result in a decrease to ca. 281 000 Mg a − 1 . Sensitivity analyses show that although land use change acts as the primary control on long-term changes in SY, the sensitivity of SY to changes in climate increases as the percentage of deforested land increases. For the 21st Century the results are highly sensitive to the scenarios used. Due to relatively large increases in rainfall erosivity, SY increases by 12% compared to the 20th Century according to scenario A2, or by 8% according to B1. However, the associated land use change scenarios cause decreases in SY by 26% (A2) and 46% (B1). The net effect is thus a decrease of SY. This study highlights the potentially significant efficacy of land use planning as a tool to mitigate the negative effects of soil erosion and sediment delivery to rivers.