Assessment of daily rainfall data recorded by two different networks in Piedmont (North-West Italy)

Rainfall data are essential for various hydrological applications related to water resource management, power production, irrigation, flood control, forecasting and validation of remotely sensed data from space platforms. Studying and analysing extreme rain events, dry and wet periods and trends can help in planning and manage the effects of the climate change. The availability of daily precipitation data is necessary to make good climatic analysis and to better understand extreme events. One of the main problem of rainfall measurement is his quite small representativeness for large area. For this reason, is necessary have a high density of rain gauges spread in whole the study area. In Piedmont are available several networks managed by public and local authorities. In this study, we will analyse the correlation of two independent automatic climate networks, to give a novel contribution to the analysis of rainfall daily data in the Piedmont region. The dataset comes from the meteorological stations of the Regional Agency for Environmental Protection (ARPA) and the Agro-Meteorological network (RAM). The comparison was performed using couple of daily pluviometric parallel measurements, of ARPA (reference serie) and RAM serie. The selection was based on five parameters: the overlapping period, the difference in elevation, the distance, the aspect and their characteristic. Once defined them for each couple the free and open source script CoRain written in R language was used. This script uses an innovative analysis approach to compare two parallel rain series (with an overlapping period). CoRain combines a set of well-known statistical tools and highlight the overestimations and underestimations due to rain gauges. The precipitation were split in different classes of intensity calculated on the ARPA series, and also the number of events for each class have been observed. This methodology allowed to highlight the percentage of precipitation events that can be considered equal between the pairs of series and, at the same time, to underline the type of events that can induce the greater difference between the two stations. The comparison between the two networks, highlight that the major difference between the precipitation records is related to the type of instrument. In fact, was found a systematic inhomogeneity in the number of rainy days, defined as days withmore than 1mm precipitation. About the extreme events, an overestimation of the ARPA stations is observed, while they tend to underestimate weak events of rainfall. This project shows the importance of comparing different meteorological networks in the study area to better understand extreme events of rainfall.