NIST–NRC Comparison of Total Immersion Liquid-in-Glass Thermometers

Abstract

The use of liquid-in-glass (LIG) thermometers is described in many documentary standards in the fields of environmental testing, material testing, and material transfer. Many national metrology institutes, including the National Institute of Standards and Technology (NIST) and the National Research Council of Canada (NRC), list calibration services for these thermometers among the Calibration Measurement Capabilities of Appendix C of the BIPM Key Comparison Database. NIST and NRC arranged a bilateral comparison of a set of total-immersion ASTM-type LIG thermometers to validate their uncertainty claims. Two each of ASTM thermometer types 62C through 69C were calibrated at NIST and at NRC at four temperatures distributed over the range appropriate to each thermometer, in addition to the ice point. Collectively, the thermometers span a temperature range of −38 • C to 305 • C. In total, 160 measurements (80 pairs) comprise the comparison data set. Pair-wise differences (T NIST -T NRC ) were formed for each thermometer at each temperature. For 8 of the 80 pairs (10 %), the differences exceed the k = 2 combined uncertainties. These results support the claimed capabilities of NIST and NRC for the calibration of LIG thermometers.

LiG Metrology, Correlated Error, and the Integrity of the Global Surface Air-Temperature Record

“…Prior to year 2000, LiG thermometer calibrations at NIST employed visualization by eye, which is the standard method used to obtain LiG thermometer readings at meteorological stations. The uncertainty attending a LiG thermometer reading taken by eye (visual repeatability) is 1/4 of a scale division [101,105]. For a calibrated 1 • C/division mercury or alcohol (spirit) meteorological LiG thermometer, the true air temperature is taken to be somewhere within ±0.125 • C of the measurement as read by eye.…”