While there exists extensive assessment of urban heat, we observe myriad methods for describing thermal distribution, factors that mediate temperatures, and potential impacts on urban populations. In addition, the limited spatial and temporal resolution of satellite-derived heat measurements may limit the capacity of decision makers to take effective actions for reducing mortalities in vulnerable populations whose locations require highly-refined measurements. Needed are high resolution spatial and temporal information for urban heat. In this study, we ask three questions: (1) how do urban heat islands vary throughout the day? (2) what statistical methods best explain the presence of temperatures at sub-meter spatial scales; and (3) what landscape features help to explain variation in urban heat islands? Using vehicle-based temperature measurements at three periods of the day in the Pacific Northwest city of Portland, Oregon (USA), we incorporate LiDAR-derived datasets, and evaluate three statistical techniques for modeling and predicting variation in temperatures during a heat wave. Our results indicate that the random forest technique best predicts temperatures, and that the evening model best explains the variation in temperature. The results suggest that ground-based measurements provide high levels of accuracy for describing the distribution of urban heat, its temporal variation, and specific locations where targeted interventions with communities can reduce mortalities from heat events.