Characterizing Spatial Patterns and Trends of Groundwater Quality and Contamination Plumes in Southern Ontario using Geostatistical Modeling

Abstract:

Adequate knowledge of water quality in agricultural, industrial or residential sites is crucial for proper decision making. Furthermore, it is highly valuable to understand the trends and evolution of spatial plumes from groundwater water from a sustainable water health management perspective. The analysis of trends in spatio-temporal data from groundwater networks, especially from irregular spatial and temporal resolutions, poses a special challenge. The present study’s objective was to quantify the trends and contamination plumes of groundwater contaminants from 32 point wells in agricultural areas of London, Ontario, Canada from 2002–2015. This area encompasses more than 7,296 km². This study presents contamination trends and plumes of the spatio-temporal distribution of groundwater ammonium (NH4 (+)), nitrate (NO3 (-)), and phosphate (PO₄^3-) using a non-parametric regression technique known as penalised splines (P-splines) in the Groundwater Spatiotemporal Data Analysis Tool (GWSDAT). Rather than using only spatial methods that consider only individual sampling points, the GWSDAT utilized relevant information from temporal components. For the study period, the concentrations of these elements varies from 0.04–4.08 mg/l, 0.04–13.46 mg/l and 0.01–0.06 mg/l for NH4 (+), NO3 (-), and PO43- respectively. These spatio-temporal geostatistical results show high groundwater NH4 (+), NO₃(-), and PO4^3- in the residential, agricultural and residential areas, respectively. Also, the estimated groundwater flow direction and magnitude were estimated and showed a south-western direction. This study demonstrates the importance of monitoring wells and areas where nutrients with above-thresholds (standards) in groundwater discharge can contribute to eutrophic conditions in open freshwater.