The 18 September 1692 Earthquake in the Belgian Ardenne and Its Aftershocks - Historical Seismology: Interdisciplinary Studies of Past and Recent Earthquakes - page 215

the Northern Rhine area. They calculated those parameters for the 1692 Verviers
earthquakes by using the data from Alexandre and Kupper (1997). They obtained a
magnitude M L = 7 by applying the relationship

between M L given by the Bensberg network and M W (Reamer and Hinzen 2004).
Their location is in the locality Esneux, some 30 km to the southwest of Verviers.
Both of our epicenters are within their 80% confidence limit.
Gasperini, Bernardini, Valensise and Boschi (1999) presented a method that
assesses the location, the physical dimensions and the source orientation of large
historical earthquakes by the use of macroseismic intensity data. They applied in
a systematic way this program called BOXER to all M

6 . 8, which corresponds to M w = 5 .

5 earthquakes having
occurred in the central and southern Apennines in Italy during the past four cen-
turies. The method computes the macroseismic epicentre and the magnitude as well
as the azimuth, the length and the width (down dip) of the box representing the
seismogenic fault having generated the earthquake We will not take into account
the latter parameters because we consider that our macroseismic data are unable to
furnish such information.
As intensity attenuation with distance, magnitude-epicentral intensity and
magnitude-macroseismic radii relationships are different in northwestern Europe
compared to those in Italy, it was necessary to modify the parameters of these laws in
the BOXER program. We approximated the Ambraseys (1985) laws relating magni-
tude to isoseismal areas by a mathematical formulation which can be used by BOXER.
We considered an equation of the type: M

> 5 .

b i log 2 A i ; where A i is the area (ex-
pressed in km 2 ) of the isoseismal of intensity I and a i ,

= a i +

b i are coefficients depending
of the considered intensity. The values of these coefficients are indicated in Table 2.
To calculate the magnitude of the 1692 earthquake with BOXER, two datasets
have been considered (Table 1). The first (Imin) considers only the minimal intensity
evaluation for localities for which a range of intensity values is given. The second
one (Imax) considers the maximal values of the intensity ranges. Localities with a
single value are included in the both datasets with this single value. We calculated
also the magnitude with attenuation valid for Italy for comparison. Magnitude has
also been determined by using the Ambraseys (1985) and Levret et al. (1994) atten-
uation laws calculated on the three isoseismals (intensities V, VI and VII) traced on
Fig. 5. All these results are presented in Table 3.
Attenuation laws generally take into account the earthquake focal depth. Thus,
the focal depth is a parameter that should be determined using intensity data. We
consider that for historical earthquakes, the reliability of macroseismic data is not
sufficient to allow depth determination. Thus, we used the Ambraseys and Levret

Ta b l e 2 Coefficients u sed in the Boxer program to approx imate Ambraseys laws
Intensity a i b i

IV 2.306 0.119

V 2.926 0.119

VI 3.546 0.119

VII 4.166 0.119