Urban stormwater runoff pollutant loadings: GIS land use classification vs. sample-based predictions

Area of Science:

• Environmental Science
• Urban Hydrology
• Water Quality Management

Background:

• Global urbanization increases stormwater volume and degrades water quality, impacting aquatic ecosystems.
• Characterizing stormwater is challenging due to high variability and numerous outfalls.
• Modeling and monitoring are essential for estimating and validating pollutant loadings.

Purpose of the Study:

• To determine and compare stormwater pollutant loadings predicted by land-use classification (desktop study) versus actual outfall monitoring.
• To assess the accuracy of predictive models against empirical data for key pollutants.
• To inform strategies for managing urban stormwater impacts on the South Saskatchewan River.

Main Methods:

• Conducted a 'desktop' study using eight land-use classifications to predict pollutant loadings.
• Implemented an 'outfall' sampling regime for a 'monitoring' study to measure actual pollutant loadings.
• Analyzed total suspended solids (TSS), chemical oxygen demand (COD), metals, and polycyclic aromatic hydrocarbons in seven Saskatoon catchment areas.

Main Results:

• The Preston Crossing catchment area was identified as a major source of predicted annual loadings (e.g., 550,000 kg TSS, 265,000 kg COD).
• Actual sampled loadings for TSS and COD were 362,700 kg and 652,700 kg, respectively.
• Discrepancies between predicted and actual loadings ranged from 29% to 156% for individual pollutants and 48% to 130% for summed pollutants per catchment.

Conclusions:

• Both predictive modeling and direct monitoring are crucial for accurately assessing stormwater pollutant loadings.
• Significant differences between predicted and actual loadings underscore the need for robust validation methods.
• Effective stormwater management and the development of treatment technologies require comprehensive outfall assessment and monitoring.