Environmental drivers of seasonal variation in green roof runoff water quality

- Seasonal variation in green roof runoff water quality was much higher than among-event variation.
- Most dissolved element concentrations in runoff were high in summer, positively correlated with temperature.
- Patterns implicate microbial mineralization or other temperature dependent processes.
- Nitrate leaching from the green roof increased several-fold for the growing season following application of an organic N fertilizer.
- Large temporal variation in green roof effluent water quality supports the need for long-term studies.

Green (vegetated) roofs provide many beneficial environmental services but can also pose a disservice by leaching nutrients and metals, via storm water runoff, to downstream aquatic ecosystems. Current estimates of water quality impacts rely on limited samples (snapshots in time) and may not accurately reflect the true influence of green roof ecosystems, which likely vary temporally as do natural ecosystems. Using a 46 m2 green roof in Cincinnati, OH, we analyzed runoff from >80 events over two years for pH, conductivity, and concentrations of dissolved nutrients, base cations, and metals. We related the variation in water chemistry to environmental variables including air temperature, storm event magnitude, and estimated antecedent moisture. We observed strong seasonal patterns in bioactive elements, with carbon, nitrogen, phosphorus, and base cation concentrations highest in the summer, and positively correlated with temperature. This suggests temperature-mediated processes such as microbial mineralization of organic matter, desorption or weathering, rather than plant uptake or hydrologic variation among storms, are the major controlling mechanisms for runoff water quality in this newly constructed green roof. The large temporal variation in green roof effluent water quality supports the need for long-term studies to characterize the complexity of these engineered ecosystems and their responsiveness to environmental variation.

191