Quantifying and upscaling surface and subsurface runoff and nutrient flows under climate variability

Abstract

Understanding and predicting the impacts of runoff on nutrients under different climate conditions within an unsaturated zone of soils is a fundamental challenge in hydrological research. The aim of this study is to provide new understanding of surface and subsurface hydrological controls on nutrient fluxes within mineral soils at the hillslope and catchment scales, and how these are influenced by climate variability. The study covers three nested spatial scales: the Blind Beck catchment (9.2 km2), the sub-catchment (0.09 km2) and the plot scale (2 m2) within the hillslope located in the Upper Eden basin, Cumbria, UK. The methodology combines field experiments and sampling, laboratory analysis and modelling approaches. Runoff experiments were conducted on two hillslope runoff plots to identify runoff processes and to quantify nutrient fluxes, one under perturbed (i.e. increased rainfall) and another under control plot. The SHETRAN physically-based hydrological model was then used to simulate runoff and nutrient flux for climate scenarios. These represent current and future (intensified hydrological cycle) conditions and were generated using the UKCP09 weather generator. Analyses of flow for different climate conditions within the unsaturated zone suggested: i) overland flow varied from 14% in dry conditions (before treatment) to more than 64% in the enhanced rainfall conditions of the total measured overland flow, ii) lateral subsurface flow dominates hillslope runoff during the transition period, iii) overland flow occurs in the winter during periods of frozen soil as Hortonian flow and iv) nutrients were most concentrated in the topsoil. In extreme climate conditions, saturation excess overland flow is probably a major contributor to storm flow followed by the subsurface flow. The results have shown that sensitivity of different runoff processes to different types/size of storms can support analysis of impacts of enhanced climate variability. The enhanced rainfall treatment in the overland flow reduced the DOC concentration 1.7 times, while increasing the NO3 - concentration 2.5 times. Under the enhanced rainfall treatment at the perturbed plot, C losses are lower in the overland flow (9.7 kg/ha) compared with the subsurface flow (22 kg/ha). This indicates enhanced loss of the DOC by the subsurface leaching pathway relative to losses through the overland flow. Decreases in the mean rainfall between 0.6% and 2.6% for the 2020 and ii 2050 period of the A1B emission scenario are modelled to decrease annual runoff 0.4% and 3.4 %. For the A1FI emission scenario, a decrease in rainfall between 1.2% and 3.2% is modelled to decrease annual runoff by 3.4% for the 2020 period and by 4.8% for the 2050 period.