A survey of past and future changes in global river flow

Abstract

River flow is an essential component of the global water cycle and arguably the best monitored variable in land hydrology. Both anthropogenic climate change as well as direct human influences on the terrestrial environment influence river flow at local to planetary scales. Here, we survey recent scientific advances in quantifying past trends and projecting future changes in global river flow, with focus on trends in flow volumes and seasonality. Previously published evidence of past changes is complemented by an analysis of changes in river flow using in-situ observations and river discharge estimates from a global re-analysis product. Available literature on future projections is accompanied by an analysis of Global Climate Model projections routed through the global river network. Systematic patterns of past changes in river flow emerge at the regional to global scales, despite significant uncertainties and small-scale spatial variability. These uncertainties are related e.g. to differences in study periods, considered indicators of change, availability of in-situ observations, and uncertainties of model-based reconstructions. Some of the most pronounced changes in past river flow include increasing river flow in the northern high latitudes and robust decreasing trends in significant parts of central and south America, the Mediterranean region, the southern tip of Africa as well as central and southern Australia. In other world regions, available in-situ observations are sparse or there is conflicting evidence, meaning that trends are still uncertain. We further show systematic shifts in the river flow seasonality with a tendency for earlier streamflow and a dampened seasonal cycle in across many parts of the world likely to due to higher temperatures and earlier snowmelt, with later streamflow in the Mediterranean linked to changes in rainfall. Climate model driven assessments of future changes in river flow suggest that total global discharge to the oceans may increase with global warming, albeit with large regional differences in the direction of change and substantial model uncertainty. Across several studies based on different model ensembles, there is consensus for increasing river flow in the northern high latitudes and some evidence for increasing river flow in tropical Africa, the Indian subcontinent and eastern and tropical Africa. Finally, we highlight regions in which past changes in river flow are consistent with future projections, which include but are not limited to increasing river flow in northern North America and northern Europe, as well as drying tendencies in the Mediterranean, the southern tip of Africa and Southern Australia. The consistency between model projections and historical trends in these regions gives confidence to future water management decisions, while disparities between historical trends and projections highlights regions where better understanding of the processes governing past and future change in river flow will be required moving forward.