Given the regional surface network of the Umbria region, a mountainous area located in central Italy, the observed hourly temperature time series from 2010 to 2017 have been analyzed applying  basic and extended quality control (QC) procedures following the World Meteorological Organization (WMO) standards. The validation procedure consisted of automatic QC, producing validated data with metadata subsequently recorded in NetCDF format. After these controls, data have been manually checked and an extended procedure has been applied to reconstruct the temperature time series for missing data. The spatiotemporal method used to reconstruct the data has been linear interpolation for 1-hour gaps, and the Empirical Orthogonal Function algorithm (EOF) for the 2-hour and longer gaps. The introduction of a complete and homogeneous data set of hourly reanalysis ERA5 (ECMWF) allowed the reconstruction of the longest gaps with statistical and physical consistency. The final product of this study is a continuous station time series of hourly temperatures that will be available to the public by the end of 2020; a daily version of the original time series is already available at the regional website (https://annali.regione.umbria.it/).Keywords --- Agrometeorology, Quality control, Missing data, ERA5, Empirical Orthogonal Function