Impact of the intensity assessment on the macroseismic parameters of earthquakes in Italy

SUMMARY Macroseismic intensity classifies the ground shaking at a locality by comparing the observed effects on humans and buildings with the scenarios characterizing each intensity degree according to a macroseismic scale. This practice may involve uncertainty in assessing intensity degree due to several factors. These uncertainties propagate to subsequent elaborations, such as the parameters of pre-instrumental earthquakes determined from macroseismic data. In Italy, more than 60% of the earthquakes in CPTI15, the Italian Parametric Earthquake Catalogue covering the period 1000–2020, rely on intensity data collected in DBMI15, the Italian Macroseismic Database. Their parameters are estimated with the ‘Boxer’ software, which determines the location and magnitude starting from their macroseismic intensity distributions. In this work, we explore the potential impact of possible inaccurate intensity assessments at a single site on macroseismic parameters (i.e. locations and magnitudes) of Italian earthquakes. We select 1108 earthquakes with at least 10 intensity data points from CPTI15 that occurred in the period 1279–2020. For each event, we simulate more than 100 sets of intensity distributions, for a total of 138.327 simulations, by varying the intensities at the sites of ±1 with a half-degree step starting from the intensity of DBMI15. Each simulated distribution is then parametrized using the same approach adopted by CPTI15 (i.e. Boxer), and the results are compared with the macroseismic epicentre and magnitude of CPTI15. The resulting parameters from all the simulated distributions are coherent with those provided by CPTI15. Locations estimates are within 5 km from CPTI15’s for 55% of the cases, and within 10 km for 83%. The magnitudes of 68 % of simulations are within ± 0.2 units of difference from the CPTI15 magnitudes and 87% within ± 0.3 units, similar to the statistical error of instrumental magnitude estimation. Moreover, we treat uncertain intensity values (i.e. 6–7) as equally representative of either the lowest intensity level (i.e. 6–7 as 6) and the highest intensity level (i.e. 6–7 as 7), and we analyse their impact on the parameters. The differences with the CPTI15 magnitude are not significant for either analysis, with more than 97.7% of the earthquakes falling within ± 0.3 units of difference.