Turning Routine Data into Systems Insight: Multivariate Analysis of Water Quality Dynamics in a Major Drinking Water Reservoir

Abstract

Long-term water quality datasets generated by the routine monitoring activities of public water supply utilities are often under-utilized in terms of the information they contain regarding system dynamics. This study demonstrates the insight that may be derived from thorough analysis of such data, in terms of evaluating key factors underlying spatial and temporal patterns in the water quality of major storage reservoirs. Principal component (PCA) and multivariate curve resolution-alternating least squares (MCR-ALS) analyses were applied to three consecutive years of monitoring data, comprising 22 physicochemical parameters, measured at three sampling stations within a major drinking water reservoir. Reservoir nitrogen levels were found to be most strongly influenced by urban run-off from the south-eastern catchment region, while total phosphorus levels were more closely linked to inflows to the northern end of the reservoir from the mainly agricultural neighbouring catchment. Elevations in soluble reactive phosphorus (SRP) were correlated with increases in the vertical temperature gradient of the water column at the relatively shallow northern end, suggesting possible release from sediments as a major source. SRP and thermal gradient, as opposed to absolute water temperature, were found to be the factors most closely aligned with chlorophyll-α levels in the reservoir. The analyses highlighted the catchment origins and in-storage foci of key factors driving algal productivity within the reservoir, suggesting that water quality management strategies may be further informed by investigation of sediment characteristics and interplay between physical, chemical and biological processes at the northern end of the reservoir.