Comparative analysis of spatial interpolation methods in the Mediterranean area: application to temperature in Sicily

An exhaustive comparison among different spatial interpolation algorithms was carried out in order to derive annual and monthly air temperature maps for Sicily (Italy). Deterministic, data-driven, and geostatistic algorithms were used, in some cases adding the elevation information and other physiographic variables to improve the performance of interpolation techniques and the reconstruction of the air temperature field. The dataset is given by air temperature data coming from 84 stations spread in the Sicily island. The interpolation algorithms were optimized by using a subset of the available dataset, while the remaining subset was used to validate the results in terms of accuracy and bias of the estimates. Validation results indicate that univariate methods, which neglect the information from physiographic variables, significantly entail the largest errors, while performances improve when such parameters are taken into account. The best results at annual scale have been obtained using the the ordinary kriging of residuals from linear regression and from artificial neural network algorithm, while, at monthly scale, a Fourier-series algorithm has been used to downscale mean annual temperature to reproduce monthly values in the annual cycle