This paper investigates the impact of pavement temperature on deflection measurements from embedded geophones and accelerometers under real-world heavy-vehicle traffic on the E16 motorway in Borlänge, Sweden. Heavy-vehicle passages were analyzed, covering a wide range of axle configurations, gross vehicle weights, speeds, and pavement temperatures from about 1 °C to more than 24 °C. Processed vibration velocity and acceleration signals were used to derive load-normalized surface deflections, which were related to near-surface pavement temperature using statistical and regression analyses. Linear and exponential temperature–deflection models were calibrated to normalize deflections to a data-driven reference temperature of +8 °C. Both models significantly reduced temperature-induced variability, with the exponential formulation providing the most reliable correction, consistent with the nonlinear stiffness–temperature behavior of asphalt concrete. Temperature-corrected deflections showed markedly improved agreement with multilayer elastic responses computed using ERAPave, with coefficients of determination up to 0.86 and normalized RMSE values near 12–17%. The results demonstrate that embedded geophones and accelerometers, when combined with appropriate temperature-normalization models, can provide accurate, traffic-speed, and year-round pavement deflection monitoring in cold and seasonally variable climates.