Dynamics of fluvial hydro-sedimentological, nutrient, particulate organic matter and effective particle size responses during the U.K. extreme wet winter of 2019–2020

The floc size distribution of suspended sediment is a critical driver for in-channel sedimentation and sediment-associated contaminant and nutrient transfer and fate in river catchments. Real-time, in situ, floc size characterisation is possible using available technology but, to date, limited high resolution floc data have been published for fluvial systems draining upland extensive grassland catchments. To that end, suspended sediment floc size distribution and turbidity were characterised at 15-minute intervals using Laser In-Situ Scattering and Transmissometry (LISST) diffraction and a YSI turbidity sonde for six storm events in the upper River Taw (15 km(2)) catchment in SW England. Maximum event discharges (Q) ranged between 4.3 and 20.0 m(3) s(−1), with clockwise hysteretic responses (HI = 0.18–0.48) of total suspended solid concentrations (TSS) and Q. The sediment flushing index was highest in the early autumn (0.93) and storm event TSS fluxes varied from 0.04 to 2.9 t km(−2). This suggests a change in sources or composition of sediment during higher Q and highly variable patterns of sediment flux from event-to-event. The proportion of particulate organic matter (POM) to TSS was highly variable (5–89%) and did not increase with Q, indicating POM source limitation. The fine-grained tail (D(10) and D(16)) of the floc size distributions decreased during hydrograph rising limbs, with the finest floc sizes associated with the highest TN and TP concentrations at peak Q. The results suggest that dynamic interactions between wet ground and extreme rainfall events can flush significant amounts of sediment from the relatively undisturbed extensive grassland upland catchment. We strongly encourage a sensors-based approach to reveal the spatio-temporal complexity of floc size and associated pollutant export during high Q generated by extreme rainfall since this can help to elucidate processes and mechanisms and generate high-resolution data for water quality modelling without significant user intervention.