A comparison among strategies for interpolating maximum and minimum daily air temperatures. Part I: the selection of "guiding" topographic and land cover variables

This paper explores the derivation and selection of a comprehensive set of continuous topographic and land cover-related variables to guide the interp olation of daily maximum and minimum temperatures over England and Wales, f or an entire annual cycle to a resolution of 1 km. The work draws on and up dates historical topoclimatic modeling through use of digital elevation dat a and land cover data, using the modeling capabilities of geographical info rmation systems. The influential guiding variables under a variety of domin ant weather patterns were identified and used to assist with the interpolat ion of an annual sequence of daily maxima and minima for 1976. North map co ordinate ("northing''), elevation, and coastal and urban effects were found to be particularly significant variables in explaining the variation in U. K. daily minimum temperature. Urban factors have not previously been thorou ghly investigated, despite the high density of population in England and Wa les. Analysis of the residuals from data withheld from the partial thin pla te spline interpolation suggests that the incorporation of coastal shape an d situation, land cover, and soils data might further improve the modeling of local-scale influences on maximum and minimum temperature. They also sug gest that the results achieved (rms errors of 0.8 degreesC for maxima and 1 .14 degreesC for minima) may be close to the limits of accuracies achievabl e at 1-km resolution given the density of temperature observation data and standard exposure of the observing network used.