Abstract:
This thesis presents a detailed meta-analytical synthesis of two recent undergraduate data sets that monitored 13 parameters in the Buriganga River system. Collectively, the original works covered8 sampling locations (five sewer outfalls and three river stations) during Summer and Autumn 2022. By harmonizing units, screening outliers, and pooling 208 individual measurements, the study aimed to benchmark water quality against national and WHO guideline values, identify critical pollutants and hotspot sites via integrated water-quality indices, and propose a statistically optimized future sampling design without incurring additional laboratory costs. Merged Water-Quality Index (WWQI) values ranged from 280 to 510, universally categorizing the system as “Very Polluted”. Cadmium (Cd) contributed 42 %–61 % of the total WWQI score, followed by turbidity, pH and electrical conductivity. Pooled Principal Component Analysis (PCA) explained 74 % of the variance with two components: “organic/inorganic pollution” (PC1) and “seasonal thermal stress” (PC2). Hierarchical clustering segregated sites into three distinct groups: (i) chronic sewer-dominated (Dhaka Uddan), (ii) mixed sewer-river (Kamrangirchar-Beribadh, Bosila), and (iii) river-dominated (Ultingonj). Power analysis indicated that monitoring only Cd, turbidity, EC, DO and BOD at four strategic sites would capture ≥90 % of existing variability, thereby reducing future analytical expenditure by ~60 %. The findings provide an evidence-based, cost-effective roadmap for regulators, BIWTA and DoE to priorities remediation efforts while minimizing redundant sampling
