Seismotectonic Framework of the 1997-1998 Umbria-Marche (Central Italy) Earthquakes P. Galli Servizio Sismico Nazionale F. Galadini Istituto di Ricerca per la Tettonica Recente THE SEPTEMBER 1997-APRIL 1998 UMBRIA- MARCHE SEQUENCE At 0:33 and 9:40 GMT on September 26, 1997 two earth- quakes ofM w = 5.5 and M w = 6, respectively, struck a large area of the central Apennines, between the Umbria and Marche regions, while on October 14, 1997 a third earth- quake (Mw = 5.6) occurred a few kilometers to the south, frightening people in the city of Rome, 120 km away from the epicenter (Figure 1). The earthquakes caused twelve fatalities and 140 injured (the 0:33 event encouraged the entire population to sleep outdoors, thus reducing the number of victims that the stronger event of the morning would have caused), and it destroyed or heavily damaged about a thousand buildings, among these a hundred ancient churches and palaces. The historical downtowns of Nocera Umbra, Fabriano, Foligno, and tens of smaller villages were seriously hit, some of them being completely destroyed (Collecurti, Cesi, Isola, Molina, etc.). On 26 September, 1997 (9:40 GMT event) part of the roof of the basilica of St. Francis of Assisi (20 km from the epicenter), decorated with beautiful frescoes by Giotto, col- lapsed. The live images of the destruction and the deaths of four technicians (who were checking the damage of the earthquake of that night) made these earthquakes tragically famous ones throughout much of the world. Generally speaking, most of the damage was due to the poor quality of building materials used in construction (rounded river stones with scarce mortar, unreinforced mas- sive stone masonry, and only a few reinforced concrete build- ings), to the old age of the houses (mostly 15th-19th centuries), and to the heavy reinforced concrete roofs of many buildings (several of them restored after the 1979 M, = 5.9 Norcia earthquake) that, like a ram, pushed away the surrounding walls. The seismic sequence continued until April 1998 (although M s = 3 events still are occurring monthly), and damage was experienced for many months. The total damage area is a 60 km elongated zone between the towns of Fabriano and Norcia, while the area of most severe damage (I > VIII MCS; Mercalli-Cancani-Sie- berg scale) is approximately 40 km long, 5-8 km wide, and elongated in a northwest-southeast direction approximately between the villages of Gualdo Tadino and Preci (Camassi et aL, 1997). Figure 2 shows the damage distribution (until April 1998) in terms of the MCS scale, together with the epi- centers of the main shocks. In Figure 1 the maximum dam- age areas relative to the main shocks are reported; they represent a "snapshot" of each single event, as they do not contain the cumulative damage due to the entire seismic sequence. The seismogenic structure ruptured progressively first toward the south, from the Cesi-Collecurti basin zone (Sep- tember 3, M L 4.4; September 26 at 0:33 GMT, M w 5.7; Fig- ure 1); then toward the north, from the Annifo plain and Nocera Umbra area (September 26 at 9:40, M w 6 and 9:47, M L 4.7; Figure 1); and then again southward in the Sellano- Mevale zone (October 12, M w 5.2 and October 14, M w 5.7; Figure 1). Two other events with Mw = 5.3 and 5.4, respec- tively, occurred in the Colfiorito basin on October 3 and 6. During October November 1997 the area south of the Nera River Valley (Preci-Campi basins) also was affected by a seismic swarm (Ms < 5; Figure 1). After a period of relative quiescence, between the end of March and the beginning of April 1998, two other strong earthquakes occurred in the northern sector of the active structure (the Gualdo Tadino area: March 26, M L = 5.6, depth 40-50 km and April 3, M e = 5.3; Figure 1). With the exception of the March 26 event, all the seis- micity is confined to the upper crust (depth < 10 km), most being located in the first 6 km below the surface. SURFICIAL BREAKSAND RELATIONSHIP WITH THE DEEPSTRUCTURE Although earthquake-induced sliding of debris was observed in the September 26 epicentral area along the existing bed- rock fault scarps that border both the Colfiorito and the Cesi-Collecurti basins, no clear surface faulting evidence has been recognized for the first two main shocks of 0:33 and 9:40 (Galli et al., 1997; Basili et aL, 1998). The aftershocks' Seismological Research Letters Volume 70, Number4 July/August 1999 417