Effects of Antecedent Precipitation Amount and COVID-19 Lockdown on Water Quality along an Urban Gradient

Water quality is affected by multiple spatial and temporal factors, including the surrounding land characteristics, human activities, antecedent precipitation amounts. However, identifying relationships between water spatially temporally varying environmental variables with a machine learning technique in heterogeneous urban landscape has been understudied. We explore how seasonal variable amounts other small-scale affect E. coli, total suspended solids (TSS), nitrogen-nitrate, orthophosphate, lead, zinc concentrations Portland, Oregon, USA. Mann–Whitney tests were used to detect differences seasons COVID-19 periods. Spearman’s rank correlation analysis was identify relationship explanatory variables. A Random Forest (RF) model predict using as inputs. The performance of RF compared that ordinary least squares (OLS). identified statistically significant all pollutant (except TSS) wet dry seasons. Nitrate only display reductions median (from 1.5 mg/L 1.04 mg/L) during lockdown period, likely associated reduced traffic volumes. highest coefficients one-day zinc, TSS concentrations. Road length positively coli zinc. best predicts orthophosphate (R2 = 0.58), followed 0.54) nitrate 0.46). most difficult had RMSE, MAE, MAPE values. Overall, outperformed OLS, evaluated MAPE, R2. an effective approach modeling both categorical COVID data along continuous rain streams. Implementing optimization techniques can further improve model’s allow researchers use for modeling.