Corrigendum to “Sensitivity test and ensemble hazard assessment for tephra fallout at Campi Flegrei, Italy” [J. Vol. Geotherm. Res. 351(2018) 1–28]

Abstract

The authors found a mistake in the quantification of the ensemble model evaluated as the weighted mixture distribution of 5 input models, as described in Section 4. The error consisted in an incorrect averaging of the input models in the ensemble. The mistake affected Figs. 10 and 11, as well as the data reported in the Supplementary Electronic Material. The corrected Figs. 10 and 11 are reported here. The new results are consistent with the results already discussed in the original manuscript. More specifically, similar results are found for the central tendency statis- tics of the ensemble (mean and median), with a slight tendency toward a larger hazard especially in the very near field. More significant changes are found for the confidence intervals (16th–84th percentiles), which re- sults wider. These results further strengthen one of the main results of the orig- inal manuscript, that is that the epistemic uncertainty resulting from the ensemble model goes beyond the nominal uncertainty quantified by any single Bayesian model. The complete updated results, in the form of Shapefiles, can be found in the Supplementary Electronic Material.