Effects of the spatial pattern of vegetation cover on urban warming in a desert city

Vegetation cover lowers air and surface temperatures in urban areas. What is less known, however, is the spatial pattern of vegetation and its variable influence on urban warming. Hence, we combine remote-sensing techniques with climate data to address the role of spatial patterns of vegetation in relation to air temperatures. Landsat Enhanced Thematic Mapper Plus image at 30 meter (m) resolution over the Phoenix metropolitan area acquired on April 19, 2000, was used. Multiple endmember spectral mixture analysis, an extension of the spectral mixture analysis approach, was used to quantify vegetation fractions at subpixel level. The Getis statistic was used to determine the spatial pattern of vegetation fractions (clustered, random, dispersed) at different levels of spatial scales (i.e., 11 × 11, 17 × 17, 23 × 23, 29 × 29, 35 × 35 window size). Results from this study suggest that spatial arrangements of vegetation play an important role in lowering air temperatures (i.e., maximum, minimum, mean). It was found that clustered vegetation can lower air temperatures more effectively than dispersed vegetation. The spatial pattern of clustered vegetation can lower minimum air temperatures (nighttime) more effectively than maximum air temperatures (daytime). The temperature difference of the lowering between dispersed and clustered patterns of vegetation cover for minimum air temperature is 7.44°C.