GNGTS 2016 - Atti del 35° Convegno Nazionale

286 GNGTS 2016 S essione 2.1 a reduction of the underestimates of larger peak values, which, being those affected by larger errors, give a major contribution to the rmsl values. This result can be related to the higher rate of attenuation with distance predicted by the equations obtained in this study using a two-stage procedure, in comparison to those predicted by previous equations, obtained with a single- stage regression. Such a difference is consistent with what expected as consequence of the bias affecting single-stage procedure results for the lack of separation between the influence of source and propagation properties, in presence of a correlation between event magnitudes and recording distances in the regression dataset. Fig. 3 shows also that estimate errors have a distribution significantly deviating from log- normality, as also confirmed by the calculation of the parameter LH of Scherbaum et al. (2004), which, for the best equations obtained in this study is quite lower (34% and 37% for PHA and PHV, respectively) than the value of 50% expected for log-normality. This should be taken into account in hazard estimates, in that, dealing with the ground motion prediction uncertainties, the assumption of a homogeneous log-normal error distribution, independent on ground motion amplitude, could distort hazard evaluation. One question that could be raised with regard to the comparative test results is that, having used the validation dataset to optimize the functional form, this dataset could be assimilate to the base for a sort of “epistemic” regression, aimed at improving the agreement between observations and predictions by modifying the equation functional form instead of its coefficients. Thus, the better results obtained in comparison to previously published equations might just reflect a casual better fit of the selected equations to the validation dataset. Therefore, a confirmation that the equations proposed in this study actually improve prediction effectiveness will require future tests on completely new data. References Akkar, S., Bommer, J.J., 2007a: Empirical Prediction Equations for Peak Ground Velocity Derived from Strong- Motion Records from Europe and the Middle East . Bull. Seismol. Soc. Am., 97 , 511-530. Akkar, S., Bommer, J.J., 2007b: Prediction of elastic displacement response spectra in Europe and the Middle East . Earthquake Engin. Struct. Dyn., 36 ,1275–1301. Danciu, L., Tselentis, G-A., 2007: Engineering Ground-Motion Parameters Attenuation Relationships for Greece . Bull. Seismol. Soc. Am., 97 (1B), 162–183. Fukushima, Y., Tanaka, T., 1990: A new attenuation relation for peak horizontal acceleration of strong earthquake ground motion in Japan . Bull. Seismol. Soc. Am., 80 , 757-783. Joyner, W.B., Boore, D.M., 1993: Methods for regression analysis of strong-motion data . Bull. Seismol. Soc. Am., 83 , 469–487. Sandikkaya, M. A., Akkar, S., Bard P.Y., 2013: A Nonlinear Site-Amplification Model for the Next Pan-European Ground-Motion Prediction Equations . Bull. Seismol. Soc. Am., 103 , 19–32. Scherbaum, F., Cotton, F., Smit, P., 2004: On the use of response spectral-reference data for the selection and ranking of ground motion models for seismic-hazard analysis in regions of moderate seismicity: The case of rock motion . Bull. Seismol. Soc. Am., 94, 2164–2185. Further Developments on Correlations Between Strong Earthquakes and NOAA Electron Bursts from Space C. Fidani Central Italy Electromagnetic Network, Fermo, Italy and SARA electronic instruments, Perugia, Italy During the last twenty years many strong earthquakes have occurred worldwide which have produced many deaths and massive damage costing billions of euros and dollars even though several of these affected areas had had earthquake-resistant buildings. These events are useful for testing different physical parameters from space, with the goal of trying some deviations