Spatial pattern assessment of river water quality: implications of reducing the number of monitoring stations and chemical parameters.

The Tamsui River basin is located in Northern Taiwan and encompasses the most metropolitan city in Taiwan, Taipei City. The Taiwan Environmental Protection Administration (EPA) has established 38 water quality monitoring stations in the Tamsui River basin and performed regular river water quality monitoring for the past two decades. Because of the limited budget of the Taiwan EPA, adjusting the monitoring program while maintaining water quality data is critical. Multivariate analysis methods, such as cluster analysis (CA), factor analysis (FA), and discriminate analysis (DA), are useful tools for the statistically spatial assessment of surface water quality. This study integrated CA, FA, and DA to evaluate the spatial variance of water quality in the metropolitan city of Taipei. Performing CA involved categorizing monitoring stations into three groups: high-, moderate-, and low-pollution areas. In addition, this categorization of monitoring stations was in agreement with that of the assessment that involved using the simple river pollution index. Four latent factors that predominantly influence the river water quality of the Tamsui River basin are assessed using FA: anthropogenic pollution, the nitrification process, seawater intrusion, and geological and weathering processes. We plotted a spatial pattern using the four latent factor scores and identified ten redundant monitoring stations near each upstream station with the same score pattern. We extracted five significant parameters by using DA: total organic carbon, total phosphorus, As, Cu, and nitrate, with spatial variance to differentiate them from the polluted condition of the group obtained by using CA. Finally, this study suggests that the Taiwan EPA can adjust the surface water-monitoring program of the Tamsui River by reducing the monitoring stations to 28 and the measured chemical parameters to five to lower monitoring costs.