The Role of Rainfall Dynamics in Influencing River Water Quality

Rainfall signals provide new potential to evaluate patterns of river water quality change. Water quality dynamics are often analysed by concentration-discharge behaviour; however, important system behaviour (e.g. sediment delivery) occurs from rainfall not necessarily resulting in high discharge.
Rainfall variability influences hydrological processes and solute exports. During dry periods, solutes accumulate in catchments and are remobilized during rainfall events. Changes in timing and magnitude of (extreme) drought and rainfall events can trigger shifts in solute exports from catchments. Hence, understanding the relationships between rainfall variability and water quality is necessary for interpreting long term changes and robust estimates of future water quality. However, these dynamics are poorly captured by process-based water quality models which limits the applicability of these models outside of calibration conditions (e.g. as required in climate change studies).

The project aims at a better understanding of the atmospheric controls on (changes in) water quality dynamics as a prerequisite for water quality assessments. The objectives are: (1) Identify controls of rainfall characteristics (intensity, duration, depth, within-storm variability and dry periods between events) and pre-event conditions on responses in solute concentrations in rivers (and proxies such as turbidity and electrical conductivity) using long term time-series of different catchments. (2) Explore temporal and spatial scales and grouping behaviour of catchments with respect to the influence of rainfall dynamics on river water quality. (3) Develop and test a model framework linking water quality responses to drivers differently across catchment groups (utilizing JHI spatial data resources) and application to evaluate water quality change and WFD status change interactions with shorter-range weather change and long-term climate scenarios.

The methods applied include time series analysis of hydrometeorological variables (with focus on rainfall temporal patterns) and water quality determinands, multivariate statistics to group similar rain and drought events (e.g. hierarchical clustering, random forests) and to derive catchment grouping (e.g. hierarchical clustering) and model development. The project will benefit from access to long term national records as well as high frequency records from various research catchments across Scotland held by the supervisory team.