Fig 1. Hourly temperature data (Farenheit) over the period from March 12th to April 4th, 2022. Data was collected at Ware Hall (blue line), the Pritzker Science Center’s white roof (purple line), the Athletic and Convocation Center (red line), and the Pritzker Science Center’s green roof (green line). The Ware Hall and ACC sensors were both placed on the black rubber sections of those roofs.

Data from Figure 3 resembles our initial hypothesis that green and white roofs will retain less heat than black roofs, resulting in a higher percent decrease from the maximum daytime temperature to the minimum nighttime temperature. There was no significant difference between any of the groups, but there were varying levels of statistical insignificance. A t-test performed for Ware and the ACC yielded a p-value of 0.8077, which makes sense as they share the same roof material and color. However, a t-test for Ware and the Pritzker's green roof produced a p-value of 0.2074, signaling that the data sets are more different.

The original hypothesis stemmed from understandings of albedo, the amount of light that a surface reflects. Lighter colors have higher albedos, which means that they will reflect most of the sun's rays and therefore will absorb less heat. It was therefore predicted that Pritzker's white and green roofs would absorb less overall heat than the black roofs of the ACC and Ware. Given this information, we believe that the differences between the roofs will be more dramatic (and potentially statistically different) during the summer months. The sun's rays are more direct and intense during the summer, so differences in how much of this sunlight is reflected will make more of a difference than during March and April when we took data.

Figure 2, which shows the daily maximum light intensity at each location, serves as a form of control for the experiment. The light intensities vary slightly, but there is no significant difference between any of the buildings except for Ware and Pritzker's green roof. Therefore, the buildings all received roughly the same amount of light. However, multiple sources of uncertainty exist within the study. Because we could not put the sensors on the roofs ourselves, we could not control their exact placement or which direction they faced. Additionally, the buildings are all located in different places, have different surrounding structures and vegetation, and all face different directions. When recreating similar studies, placing different roof materials and colors in the same space would be beneficial to reduce uncertainty. Completing the study in the hotter summer months would also provide a better sense of how each roof holds and releases heat.

In future studies, taking temperature and light data within urban settings such as Dorchester would provide us with more data on the urban heat island effect in a realistic city setting.