Conclusions and some comments on the calculation of uncertainty when constructing a temperature scale

Abstract

This paper is a continuation of [1], in which we considered a procedure for calculating uncertainties when constructing the International Temperature Scale of 1990 (ITS-90) using standard platinum resistance thermometers. This procedure follows directly from [2]. The present paper supplements [1] as regards problems of covariance between the resistance of a platinum resistance thermometer when measuring temperature and its resistance at fixed points. Some statements were also formulated in [3–5]. Since the international document of the BIPM [6] has already been accepted in this area, the procedure for calculating uncertainty when analyzing their individual sources is already being used. In this paper, we show that it is possible to take into account any covariance between these resistances. The use of the proposed procedure gives results that are similar to the procedure based on orthogonal interpolation functions [6–10]. General Uncertainty of the Result of a Measurement of Temperature by a Calibrated Platinum Resistance Thermometer. To calculate the temperatures T when they are measured using a platinum resistance thermometer in accordance with the ITS-90, the following inversion equation holds [11]: