Problems associated with homogeneity testing in climate variation studies: a case study of temperature in the northern Great Plains, USA

Abstract

Global climate change is a controversial issue facing researchers and climatologists today. In order to obtain the most reliable results when studying climate change, the data being analysed must be as homogeneous as possible. A homogeneous time series is one in which trends and variations are caused only by effects of weather and macroclimate. The concept of homogeneity has been addressed by some researchers, but only by testing 'average' time series such as the means and the annuals. This paper utilizes a homogeneity test developed by Alexandersson and applies it to mean monthly maximum, minimum, and mean temperature data from 22 stations in the northern Great Plains, USA. One of these, Valentine, is a first-order station and is used as the reference station. When Valentine was adjusted for a possible inhomogeneity due to its move, it was found that Valentine's adjustments had a distinct seasonal pattern. After testing the other stations against Valentine, it was found that the position of a significant discontinuity in a station's monthly mean or annual series was not always the same in a corresponding monthly maximum and minimum series. In addition, a seasonal pattern similar to that of Valentine was found for each station's adjustment values.