Urban expansion is a significant driver of near-surface temperature increases in modified landscapes. While research often focuses on the effect of urban environments on downstream non-urban locations, thermal advection from upstream urban development on existing urban environments remains understudied. We analyze historical observations for rapidly and non-rapidly expanding cities in the southwestern US to isolate the primary driver responsible for urban canopy layer warming. Results indicate increasing minimum temperatures (Tmin) of 0.0835 degrees C/year (0.0802 degrees C/year) and decreasing diurnal cycles of -0.0506 degrees C/year (-0.0664 degrees C/year) for Phoenix, AZ (Las Vegas, NV). In contrast, non-rapidly expanding cities show negligible Tmin increases of 0.0005 degrees C/year (0.021 degrees C/year) and negligible diurnal cycle changes of +0.0035 degrees C/year and -0.00073 degrees C/year for Williams (AZ) and Flagstaff (AZ), respectively. To fully quantify the hypothesized first-order effect of thermal advection in regulating canopy layer warming, we examine the role of upstream urban expansion on observed temperature changes at Phoenix' Sky Harbor International Airport using the Weather Research and Forecasting (WRF) model and idealized (semi-Lagrangian) simulations. Two sets of WRF monthly summertime simulations (June 2002 and June 2019) are conducted, with multiple scenarios whereby the urban extent around Sky Harbor Airport is modified. Consistent with multi-decadal observations, WRF simulations identify upstream urban expansion as the primary driver of observed near-surface temperature changes. Semi-Lagrangian simulations corroborate WRF results, confirming the role of thermal advection through increasing Tmin and decreasing diurnal cycles. Our semi-Lagrangian model offers a computationally efficient tool to predict future minimum temperature trends due to continued urban expansion, with potential application for semi-arid cities globally.