River responses to climate change – lessons from ancient river records, modern river discharge analyses and experiments

Here we discuss three aspects of river responses to climate change, based on sedimentological data from ancient river systems, modern river discharge analyses and experiments.

Firstly, data from the sedimentary record of rivers show that sustained Froude supercritical flow conditions occurred in some rivers, where most geomorphological work was done under supercritical flow conditions. This evidence includes meso- (bedform) and macro-scale (bar-scale) structures, and has been linked to so called variable discharge rivers that occur in certain hydroclimates. Experiments, as well as modern river observations indicate that sand and even coarser sediment is transported in suspension in such rivers, and that sediment transport capacity is elevated, and the transport rates faster than the bedform migration rates. This raises questions about whether very different values of landscape diffusivity, sedimentation rate and sediment input rate variables need to be used for rivers from different hydroclimates to calculate landscape equilibrium, reaction and response timescales.

Secondly, analyses of modern river discharge show that only rivers in certain hydroclimates experience negative feedback loops due to frequent occurrence of low-magnitude and rare occurrence of high magnitude floods. Instead, rivers prone to supercritical flow occur in hydroclimates that promote high flood magnitudes and extremely low low-flow magnitudes where flows below the threshold for sediment motion are unable to rework high-magnitude event deposits and feedback loops are cumulative and positive. This shows that also river feedback mechanisms respond to climate change.

Thirdly, such rivers have been documented to occur in some Paleocene-Eocene successions that formed during multiple global warming events, including the Paleocene-Eocene Thermal Maximum (PETM) - a short interval of extreme temperatures related to the largest carbon release of the Cenozoic Era. Surprisingly, rivers shifted from subcritical-flow-dominated to supercritical-flow-dominated during initial temperature increase in Paleocene and did not shift back until after the Early Eocene warm period, despite the largest amplitude of temperature increase at the PETM and decrease post PETM. This indicates a significant, and perhaps threshold-driven change in river sensitivity to climate change.

We conclude that in some rivers, related to hydroclimates with high precipitation variability, such as in monsoon zone, and sub-humid to arid subtropics, climate signal propagation capacity and its effects may be underestimated, and that the sensitivity of river response to climate change is likely nonlinear and dependent on hydroclimate type.