Enhancing Water Quality Assessment in Mula and Mutha Rivers Through Linear Interpolation and the River Watch Monitoring Platform

Abstract

: Monthly river water quality dashboards often have missing station records, which makes reporting and month-to-month comparison difficult. It is not clear whether simple linear interpolation can fill short gaps without creating too much error under time-respecting tests. A station-by-station workflow was applied to 8 stations on the Mula, Mutha, and Mula-Mutha rivers from January 2017 to December 2023. The method used strict bracketing start and end values, no extrapolation beyond observed data, and pseudo-missing masks in 2022 to 2023. On masked test points, linear interpolation achieved an MAE of 0.142 with 95% CI [0.129, 0.156] and improved pooled coverage from 0.944 to 0.973. Accuracy is limited to the numeric-only subset because token-coded and censored coliform values were excluded from error metrics, and findings are limited to the monitored stations and time window. The study specifies an easy to check and verify imputation and labelling protocol to support routine monthly water quality reporting for municipal monitoring teams and WASH practitioners.