Predictive models using “cheap and easy” field measurements: Can they fill a gap in planning, monitoring, and implementing fecal sludge management solutions?

The characteristics of fecal sludge delivered to treatment plants are highly variable. Adapting process operations accordingly is challenging due a lack analytical capacity for characterization and monitoring at many plants. Cost-efficient simple field measurements such as photographs probe readings could be proxies control parameters that normally require laboratory analysis. To investigate this, we evaluated questionnaire data, expert assessments, collected from 421 onsite containments. This data served inputs models varying complexity. Random forest linear regression were able predict physical-chemical including total solids (TS) ammonium (NH4+-N) concentrations, solid-liquid separation performance settling efficiency filtration time (R2 0.51-0.66) based on image analysis (sludge color, supernatant texture) (conductivity (EC) pH). Supernatant color was the best predictor time, EC NH4+-N, texture TS. Predictive have potential applied real-time if database developed validated in other cities. Simple decision tree single classifier containment type can also used make predictions about citywide planning, where lower degree accuracy required.