The goal of this study is to improve our understanding of the interactive function of impervious and vegetation covers at different levels of the local and intra-urban spatial scales in relation to air temperatures in an urban environment. A multiple regression model was developed using impervious and vegetation fractions at different scales to predict maximum air temperature for the entire Phoenix metropolitan area in Arizona, USA. This study demonstrates that a small amount of impervious cover in a desert city can still increase maximum air temperature despite abundant vegetation cover.