**5. The earthquakes of May 1984**

The seismic sequence which started in May 1984 consists of two main shocks, in 7 May (Ms=5.8 seismic moment is M0 =0.59x1018 Nm and depth 10 km) and in 11 May *(*Ms=5.2; M0 =0.21x1018 Nm, depth 10 km). The main shock fault plane was part of the Sangro fault, one of three major SW-dipping normal faults that take up much of the extension of this part of central Italy in Lazio and Abruzzo (figure 1a) (Westaway et al., 1989).

In order to evaluate the hydrological effects produced by the seismic sequence, we analyzed the data registered in 15 hydrometrical stations and the variations of levels of two lakes (Barrea and Scanno), all localised in the epicentral area and monitored by the Italian Hydrographic Survey.

Hydrological variations have been registered in 12 hydrometric sections. Among them, 3 hydrometric stations WSW in the epicentral area, and another station located west of the continuation toward north of the Val di Sangro fault did not show notable hydrological anomalies in the days before the event of 7 May, while during all this month flow rates were comparable with the average values calculated for the period 1979-89, even if the rainfalls from January to May 1984 have been lightly higher than the average.

Nevertheless, gas outpouring and water muddying have been noticed in the evening of 5 May, in the springs at Posta Fibreno, localized about 2 km before the hydrometric sections.

Also in the hydrometric section of Sangro river at Ateleta, the hydrological anomaly was pointed out only in the value of the average streamflow level in May 1984. Consequently, for this earthquakes we have a total of 14 sites where hydrological effects have been noticed, taking into account that in 3 sites (the Rapido and Gari rivers at Cassino, the Gari river at S. Angelo in Theodice) the positive hydrological anomalies preceded the earthquake of 7 May. Figures 7a and 7b show the anomalous behaviour in two studied sites. In Fig. 7a we show an hydrological anomaly consisting of a strong increase in the streamflow rate of the Sangro River at Barrea, not imputable to the rainfalls. Figure 7b shows the daily streamflow level of Fucino channel, "a detail" of Fig. 7a. The two sites are very near each other. It shows how during April 1984 the sharp increase of 7 May cannot be imputed to the very small rainfalls on 6 May.

Fig. 7. Increase in the streamflow rate of the Sangro River at Villetta Barrea (a) and in daily streamflow level of Fucino channel (b).
