Extreme heat exposure as a result of the interactions between heat waves and urban heat islands can lead to resident fatalities. Heat exposure is already the leading environmental cause of death, and with global climate change and growing urban populations, risk of heat exposure fatality is expected to become more severe, especially among marginalized groups. These conditions make the spatial heterogeneity of land surface temperatures particularly relevant to urban planners, who need to consider equitable, safe environmental conditions for all residents. Within cities however, there can be much variation in surface air temperatures, following physical heterogeneity of development and vegetation, which have interactive effects on urban canyon-scale microclimate. Yet data gaps in key parameters of physical heterogeneity result in uncertainty in urban canyon scale air temperature variability. In this study, we use a three-dimensional computational fluid dynamics model, ENVI-met, to conduct a global sensitivity analysis of physical morphology and vegetation parameters (e.g. aspect ratio, orientation toward prevailing winds and green coverage) in idealized urban street canyons to quantify both parameter uncertainty and scenario variability. Our findings demonstrate although vegetation parameters have a great influence on modifying canyon’s surface temperature, their cooling effect strongly depend on canyon’s physical morphology conditions. The results of those sensitivity analysis can be used to develop a policy-relevant typology of street canyons that can be applied in an adaptive planning framework as one component of population risk quantification.