Comparison of air temperature calibrations

European national metrology institutes use calibration systems of various types for calibrating thermometers in air. These were compared to each other for the first time in a project organized by the European Association of National Metrology Institutes (EURAMET). This EURAMET P1061 comparison project had two main objectives: (1) to study the equivalence of calibrations performed by different laboratories and (2) to investigate correlations between calibration methods and achievable uncertainties. The comparison was realized using a pair of 100 {\omega} platinum resistance thermometer probes connected to a digital thermometer bridge as the transfer standard. The probes had different dimensions and surface properties. The measurements covered the temperature range between -40°C and +50°C , but each laboratory chose a subrange most relevant to its scope and performed measurements at five nominal temperature points covering the subrange. To enable comparison between the laboratories, comparison reference functions were determined using weighted least-squares fitting. Various effects related to variations in heat transfer conditions were demonstrated but clear correlations to specific characteristics of calibration system were not identified. Calibrations in air and liquid agreed typically within ±0.05°C at +10°C and +80°C . Expanded uncertainties determined by the participants ranged from 0.02°C to 0.4°C and they were shown to be realistic in most cases.