Remotely sensing the cooling effects of city scale efforts to reduce urban heat island

While recent years have seen many analyses of techniques to reduce urban heat island, nearly all of these studies have either been evaluations of real small scale applications or attempts to model the effects of large scale applications. This study is an attempt to analyze a real large scale application by observing recent vegetated and reflective surfaces in LANDSAT images of Chicago, a city which has deployed a variety of heat island combative methods over the last 15 years. Results show that Chicago's new reflective surfaces since 1995 produced a noticeable impact on the citywide albedo, raising it by about 0.016, while citywide NDVI increase is around 0.007. This finding along with counts of pixels with increased albedo and NDVI suggest that the reflective strategies influenced a larger area of the city than the vegetative methods. Additionally, plots between albedo increase and corresponding LANDSAT temperature change over the test period have linear regressions with steeper slopes (-15.7) and stronger linear correlations (-0.33) than plots between NDVI increase and temperature change (-8.9 slope, -0.17 correlation). This indicates that the albedo increases produced greater LANDSAT cooling than the NDVI increases. Observation of aerial images confirmed that typical instances of efforts to increase albedo, such as reflective roofs, produced stronger LANDSAT cooling than common instances of NDVI efforts, such as green roofs, street trees and green spaces. Accordingly, the reflective strategies were likely much more effective at cooling Chicago's LANDSAT heat island and may signify a generally more effective strategy for similar cities. (C) 2011 Elsevier Ltd. All rights reserved.