The robustness and reliability of any climatic assessment, especially those analyzing climate variability and change, depend crucially on the soundness of data and uncertainty associated with measurements. The metrological traceability to national standards of the measurements involved in climate change is a necessary step. This work describes the first attempt to define a fully traceable procedure for the calibration of air-temperature sensors in a special chamber manufactured for this purpose. The facility was designed to calibrate automatic weather stations by comparison with embedded reference standards traceable to temperature and pressure primary standards as maintained in a national metrology institute. The results shown in this paper were obtained through a prototype of the final calibration chamber. An application example of the calibration process is described for the temperature sensor working in the Meteorological Observatory of Moncalieri (Italy) of the SocietA Meteorologica Italiana where a continuous climate record has been kept since 1865. As result, the data of the historical data series have now associated calibration uncertainties as contributions from the instrument capability, resolution, and calibration procedure adopted. Finally, the possibility of improving climatic analysis, for evaluating climatic trend, by applying calibration corrections to data recorded has been evaluated. The in situ calibration campaign and the resulting differences between the corrected and uncorrected data for the most important quantities considered in the climate analysis, the maximum, minimum, and mean temperatures, are reported. The analysis carried out on a one-year-long observation period shows that the application of the calibration function has an impact on the recorded data. The differences between corrected and uncorrected data are outside the expanded calibration uncertainty, and in a winter season, this effect is more evident.