Weather and the Instrumental Record

Abstract

Climate scientists use instrumental data from numerous weather stations to develop summary measures of regional and global temperatures. The difficulties of doing this are illustrated using both hypothetical data and information from two weather monitoring stations, where one of the stations is clearly influenced by non-climate factors to a greater extent than the other. Instrumental records are available from numerous weather stations around the globe, but whose numbers and quality have varied over time, and from satellite data. However, as demonstrated using simple data, efforts to remove non-climatic factors (such as the so-called urban heat island effect) from the surface temperature reconstructions prove to be unsuccessful. Statistical analyses indicate that, since the late 1970s, some 50 % of the temperature increase in the reconstructed data is attributable to socioeconomic factors, but that this is not true of temperatures derived from satellite data. The chapter ends by examining the potential for replacing traditional crop insurance, and its inherent drawbacks (adverse selection and moral hazard), with financial weather-based derivatives.