1961–1990 high‐resolution monthly precipitation climatologies for Italy

• DOI: 10.1002/joc.5217
• Publication Date: 2017-07-26T09:02:42Z
• Authors: A. Crespi; M. Brunetti; G. Lentini; M. Maugeri
• Keywords: Orography; Elevation (ballistics); Interpolation; Inverse distance weighting

ABSTRACT High‐resolution monthly precipitation climatologies for Italy are presented. They based on 1961–1990 normals obtained from a quality‐controlled dataset of 6134 stations covering the Italian territory and part Northern neighbouring regions. The computed by means two interpolation methods modelling precipitation‐elevation relationship at local level, more precisely weighted linear regression (LWLR) kriging (RK) performed. For both methods, optimisations also applied in order to improve model performance. Model results compared with those provided other widely used which do not consider elevation distribution: ordinary inverse distance weighting. Even though all four models produce quite reasonable results, LWLR RK show best agreement observed station leave‐one‐out‐estimated mean absolute errors ranging 5.1 mm (July) 11 (November) models. Their better performances even clearer when specific clusters (e.g. high‐elevation sites) considered. provide very similar grid point they some peculiar features. In particular, is found have extrapolation ability sites data density high enough, while robust performing over areas complex orography scarce coverage, where may unrealistic values. However, prediction intervals, provides straightforward approach quantify uncertainty any study domain, helps identify mainly affected instability. high‐resolution exhibit heterogeneous seasonal‐dependent distribution throughout domain allow main climatic zones Italy.