Comparative assessment of water quality in the Vam Thuat River using WQI and FCE methods

Abstract

Comparative applications of the conventional Water Quality Index (WQI) and Fuzzy Comprehensive Evaluation (FCE) remain limited for small, densely urbanized rivers, where differences in mathematical aggregation may lead to substantially different pollution classifications. This study compared WQI and FCE using the same water quality dataset collected at 15 sites along the 5.5 km An Phu Dong reach of the Vam Thuat River during the rainy season in October 2024 and the dry season in March 2025. Eight physicochemical and microbiological parameters were included in the assessment. Although pH remained within the national Class B range, dissolved oxygen decreased to 0.84 mg/L, while COD, N–NH 4 3− , P–PO 4 3− , and total coliforms reached 71.11 mg/L, 17.18 mg/L, 2.17 mg/L, and 8,900 MPN/100 mL, respectively. Dry season WQI values ranged from 33 to 84, classifying 13.3% of the sites as good, 60.0% as moderate, and 26.7% as poor. Rainy season WQI values ranged from 36 to 72, with 66.7% of the sites classified as moderate and 33.3% as poor. In contrast, FCE assigned 53.3% and 26.7% of the sites to Grades IV and V, respectively, during the dry season, and 86.7% and 13.3% to Grades IV and V during the rainy season. Both methods identified progressive deterioration toward the downstream reach, with VT14 and VT15 consistently representing the most severely affected sites. The difference in classification severity arose because WQI compressed parameter specific subindices into a single scalar value, whereas FCE retained graded memberships and applied sample specific weights. Under the adopted weighting scheme, FCE provided a more conservative classification, rather than evidence of inherently greater accuracy. Their combined application supports routine communication, boundary-sensitive screening, and targeted management of highly impacted urban river reaches.