The variability and change of Italian climate in the last 160 years

The awareness of the importance of data quality and homogeneity issues in the correct detection of climate change has increased rapidly in the last few years. Most of the contributions have been addressed to upper air data, however errors and inhomogeneities also concern surface ones. At surface level it is often assumed that such inhomogeneities have random distribution and that, considering a sufficiently large number of series, average records with negligible bias can be obtained. This assumption is likely to be correct if global or hemispheric averages are considered, but it may not be correct at a regional scale. The aim of the work is a rigorous reconstruction of the Italian climate for the last centuries (the longest series start in the late 1700s), with particular attention to the identification of spurious non-climatic signals introduced by changing instruments and methods in the measurement procedures. A data set of 111 precipitation series, 48 minimum and maximum temperature series and 67 mean temperature series was set up, together with the information about the station history (metadata). The records were subjected to a detailed quality control and homogenisation procedure that was extensively supported by a large metadata availability. The series were grouped by means of Principal Component Analysis and regional average records were obtained and analysed for trends. Trend analysis was performed on seasonal and annual basis by means of the progressive Mann-Kendall statistics and the progressive analysis of the linear regression coefficients. A comparison between the homogenized and the original series and the preliminary results of the analysis are presented. Particular emphasis is given to stress the importance of data homogenisation in the correct detection of long-term trends.