ASSESSMENT OF WATER QUALITY VARIATIONS IN SAN JUAN RIVER USING GIS AND MULTIVARIATE STATISTICAL TECHNIQUES

Abstract

Investigations of water quality variations need to consider the temporal aspect as well as the spatial aspect to better understand processes and identify control factors and facilitate the formulation of potentially effective measures to improve water quality. The water quality of San Juan River and its tributaries was assessed over a one-year period through monthly water sampling and 24-hour in situ measurements at 13 primary stations, bi-weekly in situ measurements at 30 stations and synoptic spatial water quality surveys along the rivers and creeks. Parameters monitored include turbidity, suspended solids, dissolved oxygen, various nutrient species, ORP, COD, BOD, coliforms and heavy metals. TSS, BOD, and COD variations showed seasonality effects: relatively high in January and February, gradually decreasing through the summer months, continually decreasing further during the rainy months of June to September, and increasing in October to December. The same was observed for Total Nitrogen, Nitrate-N and Ammonia-N. Based on COD and ORP, the Balingasa/Talayan Creek and Maytunas Creek are the most degraded in water quality. Coliforms at the 13 primary stations were all above 10,000,000 MPN. ORP values were largely negative at all stations. However, at most stations, except Balingasa/ Talayan and Maytunas creeks, ORP can become zero or positively valued due to dilution by rain and runoff. Agglomerative hierarchical clustering was used to groups the primary stations into classes of varying pollution severity. The delineated subwatersheds were characterized in terms land use, population and road density evaluated using GIS zonal analysis. Based on the factor analysis, the dominance high residential areas, industries, and informal settlements aggravate water quality with increased BOD, COD, nutrients and coliforms. Results from multiple linear regression indicate that COD levels are largely due to industries and informal settlements. The use of multivariate statistical analysis enabled a better assessment of water quality variations as well as the spatial variability of factors influencing water quality.