Statistically downscaled climate projections as a support for adaptation tools

ADRIADAPT is a project inside the framework of the Italy-Croatia Interreg Cooperation Programme. Focused to contrast impacts over Adriatic coastal areas, which are particularly exposed to climate changes, it aims to provide a resilience information platform, suitable for performing vulnerability analysis and making decisions. In this work, climate projections are computed for some Emilia-Romagna coastal areas throughout a statistical downscaling scheme, based on the canonical correlation analysis between local climate indices (predictands) and large-scale fields (predictors). Firstly, the scheme has been calibrated and validated at a seasonal time scale for minimum and maximum temperature, tropical nights, heatwave duration, frost days, obtained by the ERACLITO observation gridded dataset for Emilia-Romagna, and large-scale fields (mean sea level pressure, 500hPa geopotential height, and 850hPa temperature) of the ECMWF-ERA40 and ERA-interim re-analysis data set. Calibration is performed over the 1961-1985 and 2006-2010 periods, while validation concerns the 1986-2005 period. Correlation coefficients, bias, and root mean square errors are taken as skill measures. Secondly, large-scale field data simulated by four global climate models from CMIP5 experiments (CMCC-CM, MPI ESM-MR, CNRM -CM5, Can-ESM2) in the framework of two emission scenarios (RCP4.5 and RCP8.5) has been treated as input to the statistical downscaling scheme to obtain local climate indices for the next four 20-year periods: 2021-2040, 2041-2060, 2061-2080, 2081-2100. Changes respect to the 1986-2005 period, taken as climatic reference, are evaluated. A Poor Man’s ensemble technique is applied to reduce uncertainties and give more statistical robustness to the results. The minimum and maximum temperature projections show a significant increase could be expected to occur for all seasons and both RCPs. The magnitude of changes is higher for the maximum temperature, especially during the summer season when changes up to 4°C for RCP4.5 and 8°C for RCP8.5 are expected at the end of the century. As regards extreme temperature indices, the seasonal tropical nights and heatwaves duration are projected to increase while frost days to decrease over all the four-time periods and for both emission scenarios.
This work has been performed in the framework of Italy-Croatia Interreg Cooperation Programme – ADRIADAPT Project (https://www.italy-croatia.eu/web/adriadapt/).