Importance of High-frequency Chemistry for Resolving Hot Moments in Headwaters: A Combined Optical Sensor and Time-series Modelling Approach

Abstract:

There are several studies that clearly demonstrate that storm periods are hot moments for biogeochemical flux in headwaters (e.g. Robson et al 1993 J Hydrol: 291-310). Despite this few models have been developed that can adequately capture fast (e.g. sub-hourly) temporal dynamics through these hot moments, and an equally small proportion of monitoring stations with high-frequency observations that have been analysed systematically. We are able to show that inadequate temporal sampling of hot moments (‘aliasing’) leads to subsequent misinterpretation of biogeochemical and hydrological processes. We have addressed this through continuous deployment of state-of-the-art field sensors producing sub-hourly river data where hot moment dynamics are extracted directly using new algorithms within the CAPTAIN toolbox, following Jones, Chappell & Tych (2014 Environ Sci Technol: 13289-97). These algorithms are capable of extracting the river transport parameters associated with reactive solutes, notably carbon. We demonstrate that most of the dynamics within these biogeochemical time-series relate to individual rainstorms (the hot moments), with the shape of the storm-based response being strongly moderated by the hydrological pathways. With parsimonious modelling using CAPTAIN we are able to show that these shapes are distorted by under-sampling of stream chemistry giving rise to different interpretations of the biogeochemical pathways and associated hydrological pathways. This work provides clear evidence that researchers and practitioners should implement high-frequency water quality monitoring and associated time-series analysis.