Estimating Missing Daily Maximum and Minimum Temperatures

Abstract

Seven methods for estimating maximum and minimum temperatures were developed from the literature and other sources. These techniques include correlative and additive procedures based on the relationships between stations, as well as procedures based on within-station temperature relations. Selection of the most appropriate technique will depend on the ultimate purpose for which the data are to be used, the size of the gaps in the weather record, and the availability of data from other stations to include in the analysis. Within-and between-station methods were compared by looking at their relative abilities to predict “pseudo” missing data items from two groups of weather stations in northern and central Idaho. Between-station regression techniques generated significantly smaller errors when compared to the remaining methods. Application of one of the methods to stations in British Columbia that contain large gaps in weather records was also described.