Interpolation of Growing Degree-Days in Non-Homogeneous Terrain

This paper investigates the spatial interpolation of growing degree-days (a measure of heat accumulation for crops) in Southland, New Zealand. This type of climate map aids land-use management and identifies high quality climates suitable for more profitable, high value crops. The use of an automated spatial interpolation technique (ANUSPLIN) speeds the production of climate maps and makes the process objective. The results of this study indicate that the accuracy of the spline model interpolated surface is comparable with that of traditional, hand-contoured maps. However this accuracy decreases with increasing topographic complexity. The density of the data network also affects the accuracy of the interpolated maps; results indicate that a less dense network than that used is sufficient for the desired level of accuracy and scale