Nitrate concentration in runoff from roofs in central PA in 2005 ranged from less than 0.5mg/L to in excess of 5mg/L  and was sometimes higher from green roofs and sometimes higher from non-green roofs (Figure 1). The average concentration of the runoff from green and non-green roofs was about 1.5 mg/L and there was no difference between green and non-green roofs (Figure 2). The total amount on nitrate in the runoff for all measured events was higher from non-green roofs than from green roofs due to retention of rain by green roofs. One event in particular, a heavy rain in October had far higher total nitrate in the runoff from non-green roofs compared to green roofs (Figure 3). The total nitrate in pounds per acre of roof was more than 4 times greater from non-green roofs than from green roofs for the storms measured.

The conclusion that green roofs reduce total nitrate runoff however must also take into account differences in nitrate in the rain. In central PA the average nitrate in the rain in 2006 was about 1.3 mg/L (NADP, Figure 5). In an area where the concentration of nitrate was less the results would most likely be different. The research roofs were 4 years old so nitrate in the initial compost added to the media was likely gone and these roofs were not fertilized for over one year before the monitoring period. Different fertilization practices or roof ages would also affect the outcome. Data from nutrient studies of established roofs suggest that 1.5 mg/L in a saturated media extract is a minimum for good plant growth so in parts of the country with clean rain it is likely that nitrate in runoff from a green roof might be expected to be higher than from a non-green roof.

Robert Berghage, David Beattie, Albert Jarrett, and Thomas O’Conner 2007. Green roof Runoff Water Quality. In Proc. Green Roofs for Healthy Cities, Minneapolis, Mn