Verifying the Representativeness of Water-Quality Monitoring to Manage Water Levels in the Wicheon River, South Korea

Changes in water level between the mainstems and tributaries of a river can create backflow effects that alter the representativeness of water-quality monitoring data. In South Korea, 16 multi-functional weirs intended to manage water levels were constructed on 4 major rivers as part of a restoration project. However, they are causing backwater effects in tributaries that coincide with poorer water-quality measurements at monitoring stations along these tributaries despite there being no change in upstream pollution sources. Therefore, this study developed a new methodology for verifying the representativeness of a water-quality monitoring network via spatiotemporal observations of electrical conductivity, self-organizing maps for monthly pattern analysis, locally weighted scatter plot smoothing for trend analysis, load duration curves, and numerical modeling. This approach was tested on the Wicheon River, a primary tributary of the Nakdong River, because the measured decline in water quality there has the potential to trigger major policy changes in the basin including limits on local development. The results clearly show that the monitoring station in the lower Wicheon is negatively affected by weir-derived backwater from the Nakdong, suggesting that this station needs to be moved upstream or a new station established upstream, beyond the backwater effects. Our approach was able to assess clearly the representativeness of an existing water-quality monitoring network using widely accessible data and methods, making this type of assessment applicable to many other situations around the world.