Abstract
The introduction and subsequent redesign of the HO-83 thermometer have brought the homogeneity of temperature time series from many U.S. Weather Service stations into question. The nature of these inhomogeneities was investigated using data from a 1 5-month side-by-side comparison of the old and new versions of the HO-83 thermometers.
Examination of differences between the two instruments found that the original version of the HO-83 read approximately 0.6°C warmer than the redesigned instrument. Significant changes in the differences between the two instruments were noted between winter and summer. It is suggested that, for stations with climatology similar to the ones used in this study, monthly mean temperature reported by the original version of the HO-83 be adjusted by adding −0.4°C to June, July, August, and September observations and by adding −0.7°C for the remainder of the year.
The physical basis of the problem with the old HO-83 seems to he related to beating of the instrument housing by internal heat sources coupled with inadequate ventilation. The mean differences between the two instruments were negative for all hours, in contrast with a change in sign between day and night as would be expected if the housing was heated and cooled by radiation. The effects of the wind speed and ambient temperature on the observed temperature differences are consistent with this finding.
