Rapidly expanding desert populations have significantly altered surface microclimatic conditions. These modifications are evident for much of the year, since desert areas are frequently affected by stable, clear, calm weather. Past empirical analyses of urban effects for Phoenix and Tucson, Arizona, have highlighted the large magnitude of the "urban heat island effect" with minimum temperatures in summer changing by over 5degreesC from the 1940s to present. Less understood are links of the built environment to surface microclimate within the urban area. This paper presents field data taken for a short sample period in summer in Tempe, Arizona, from an "urban canyon" environment, an asphalt lot, and a nearby irrigated park. A Surface Heat Island Model (SHIM) compares favorably with these field data. The model is a "force-restore" scheme to estimate post-sundown surface cooling at night - the time of most evident development of urban heat islands. The model incorporates thermal, morphological, and geometric features of the urban area that promote considerable change of local microclimates. A sensitivity analysis of key variables yields insights into challenges to ameliorate heat islands. Key are use of low thermal admittance materials, considerations of building height/width ratios, sky view factors, and judicious use of evaporative surface water cooling. For urban ecological studies, the use of modeling together with neighborhood scale monitoring is encouraged to assist in unraveling the intra-urban variance of microclimates and effects on ecosystem processes.