Search for: All records

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

This research examines how individual preferences for the major functions of stream restoration processes are associated with flood prevention and risk mitigation in Johnson Creek of Portland, Oregon, USA. We first reviewed a set of results from an analytical hierarchy process (AHP) model to rank the major stream restoration functions and compared citizens' preferences for “flood prevention” using ordinary least squares regression. Our results show that the perceptions and interests of citizens may be centred on the inconvenience of everyday life arising from the previous flood events. Residents in the highly urbanized downstream regions showed a higher sensitivity to flooding than those living in the upper regions of the watershed. Community participation and annual incomes are positively related to flood risk perception in more developed downstream regions, while ecological or development goals associated with property protection are positively associated with higher flood risk perception in the less developed upper regions. Our findings of citizen perceptions can be adopted to help local government leaders and households mitigate flood risk while also achieving multiple benefits from stream restoration projects.

 

Over the next century, model projections suggest that river run‑off in the Pacific Northwest will increase during the winter season and that sea‐level rise (SLR) may exceed a meter. To investigate the resulting changes in flood hazard, we numerically model the February 1996 and January 1923 floods (the largest and third‐largest Willamette River floods since 1900) under present and potential future run‐off and sea level scenarios. First, we reproduce the actual February 1996 flood to within a root‐mean‐square error of 0.05 m (N = 7) for peak water levels. Next, we run scenarios in which three SLR scenarios (0, 0.6, and 1.5 m) are combined with two river run‐off scenarios (0% and 10% run‐off increase). Then the slightly larger 1923 flood scenario is run, but with modern (higher than historical) Columbia River flow. The results indicate that a 10% increase in river run‐off increased the1996 flood magnitude by 0.78 m, while 1923 flow increases flood magnitude by 0.82 m. Overall, the type and magnitude of future flood hazards vary with reach. The Portland/Vancouver Metropolitan area is most sensitive to changes in run‐off, with a smaller change of ~0.2–0.26 m per meter of SLR. By contrast, coastal regions are quite sensitive to amplified sea level and exhibit nonlinear responses based on changes to river slope and tides. Between the fluvial region and the estuary, a region of compound flood hazard exists that is sensitive to changes in river discharge, sea level, tides, and storm surge.