Abstract. Past seismic events worldwide demonstrated that damage and death toll depend on both the strong ground motion (i.e., source effects) and the local site effects. The variability of earthquake ground motion distribution is caused by local stratigraphic and/or topographic setting and buried morphologies, that can give rise to amplification and resonances with respect to the ground motion expected at the reference site. Therefore, local site conditions can affect an area with damage related to the full collapse or loss in functionality of facilities, roads, pipelines, and other lifelines. To this concern, the near real time prediction of damage pattern over large areas is a crucial issue to support the rescue and operational interventions. A machine learning approach was adopted to produce ground motion prediction maps considering both stratigraphic and morphological conditions. A set of about 16'000 accelometric data and about 46'000 geological and geophysical data were retrieved from Italian and European databases. The intensity measures of interest were estimated based on 9 input proxies. The adopted machine learning regression model (i.e., Gaussian Process Regression) allows to improve both the precision and the accuracy in the estimation of the intensity measures with respect to the available near real time predictions methods (i.e., Ground Motion Prediction Equation and shaking maps). In addition, maps with a 50 × 50 m resolution were generated providing a ground motion variability in agreement with the results of advanced numerical simulations based on detailed sub-soil models. The variability at short distances (hundreds of meters) was demonstrated to be responsible for 30–40 % of the total variability of the predicted IM maps, making it desirable that seismic hazard maps also consider short-scale effects.
The Predictive Relationship between Earthquake Intensity and Tweets Rate for Real‐Time Ground‐Motion Estimation
