Reanalysis-based mesoscale wind maps for the design of structures and infrastructures with an application to Italy

Synoptic extreme winds are traditionally mapped at the lower bound of the countrywide macroscale resolution (hundreds of km) on the basis of time series measured at land anemometric stations, while the assessment of the design wind speed at the construction site is entrusted to the designer within the so-called return criterion. Coarse, uneven distribution of the stations, uncertainties in their setup, measurement errors, challenging subjective evaluation of the exposure roughness, inconsistencies among national wind provisions are some of the critical issues affecting the in force map-and-return approach. This study is intended to test an alternative approach to directly assess the wind hazard at the lower bound of the meso-$\gamma$ scale resolution (about 2 km) around a construction site. The approach is grounded on data issued from a weather forecast computational model, its reanalysis by means of assimilated remote sensing observations, and possibly its downscaling. Three different reanalysis/downscaling models are adopted. The resulting wind maps over the Italian Country are critically compared with measurements at 21 stations. The errors made by each model are assessed for current and extreme wind speed with different return periods. Finally, a reanalysis-based engineering approach to design wind speed is presented by proposing model correction factors.