There is growing public awareness of the role that low-reflectance (e.g., dark), impervious surfaces play in generating excess urban heat and stormwater runoff, and attention is turning to alternative systems that mitigate these effects and provide other benefits. We have analyzed temperature data collected from three different roof surface treatments at the Con Edison “Learning Center” (TLC) in Long Island City (LIC), Queens, New York. The three test roofs include (1) a green roof (4-inch depth modular sedum system),1 (2) a nearly black Ethylene-Propylene-Diene-Monomer (EPDM) membrane, and (3) a high-reflectance white EPDM membrane. From the first year of temperature data we report the following key findings.

The Con Edison LIC area is showing a relatively strong urban heat island signal that correlates with low vegetation abundance in the area.

The black roof daily temperature extremes are extraordinary in all seasons. However average seasonal black roof temperatures are still within 10.8°F (6°C) of either the green or white roofs.

White roof membrane temperature peaks are on average 30oF (17oC) cooler than black in summer. Green roof membrane temperature peaks are on average 60oF (33oC) cooler than black in summer. These peak reductions support claims for prolonged roof service life on white and green roofs.

With respect to direct atmospheric heating and urban heat island mitigation, the new white and green roofs are performing comparably. The data suggest that the young green roof is also improving its temperature control relative to the white surface.

The white roof surface is not showing any “winter heat penalty” relative to the black roof.

The average winter heat loss rate on the green roof (-4.16 W/m2) was 34% lower than under the black roof (-6.5 W/m2). The summer heat gain rate on the green roof (+0.41 W/m2) was 84% lower than under the black roof (+2.57 W/m2).

These building energy heat flow reductions refer only to the vertical heat flow through the roof insulation layers and not to any other building heat flows or energy requirements.

Although the percentage heat flow reductions on the green roof are significant, the realized energy savings are still modest. We estimate that the 1,000 m2 (10,764 sq ft) green roof on the Con Ed building is saving roughly $400/yr in heating costs and $250/yr in cooling costs. If this area had been a white roof instead, we estimate that cooling savings would have been $200/yr.