Interpolation of rainfall through polynomial regression in the Marche Region (Central Italy)

Notwithstanding its small size (less than 10,000 km2), because of its varied topography, ranging from the Apennines Range (up to more than 2000 m amsl) to coastal environments, the Marche Region (the Adriatic side of Central Italy), is characterized by many different types of climate. In this region there are no fully satisfactory models to interpolate and generalize rainfall data from the 111 available meteorological recording stations; however, in this study an innovative way to interpret data linking precipitation to many topographic parameters is introduced. Based on those considerations, statistical analyses were carried out on rainfall historical series in order to assess significantly variations during the last 60 years and to create a model capable of explaining rainfall distribution based on geographical and topographic parameters. The model highlighted a significant decrease of rainfall from 1961–1990 to 1991–2016, over the whole period, in the hilly and mountainous sectors (100–200 mm), while closer to the coast the difference is slight (about 0–100 mm). The new model also highlights the presence of some outliers in the rainfall values, which may lead to a better comprehension of climatic dynamics in this area.