A look at global seismicity during 1999 in the longer-term time frame

An analysis and discussion of the main features and effects of the 1999global seismicity based especially on the inferred patterns of strong(M_w 7) earthquakes using NEIC/USGS data is presented.Based on the above data and their statistical features (see also Table 1)one may state that from the point of view of occurrence rate (number/yr)the 1999 global seismicity was fairly regular, but from the moment (orenergy) release standpoint it is well under the long-term average. The spacepartition was again rather typical and time distribution was quasi-Poissonian,noteworthy, the 1999 worldwide seismicity had an anomalously high deathtoll, to be discussed later. Two main new insights are brought by ouranalysis: (i) a (mega)quiescence along the whole south American segmentof circum-Pacific earthquake belt, which was inferred and rated asanomalous while it was underway and which, in retrospect, ended with thegreat (M_w = 8.4 HRV) 2001 Arequipa (Peru) event; (ii) an intriguingmonotonous rate decrease within only magnitude class 5.0 to 5.9 duringthe analyzed period (i.e., 1990–1999).     

Reservoir-induced Seismicity in Brazil

—?A compilation of 16 cases of reservoir-induced seismicity in Brazil is presented with maximum magnitudes ranging from 1.6?M _L to 4.2?m _b . The compilation includes: location of the main epicentral area with respect to the reservoir (inside the lake, at the margin, or outside), predominant geology, and the temporal distribution of the main phase(s) of activity (initial or delayed in relation to impoundment). Data on the regional stress field for some reservoirs is also included. Four recent cases are discussed in more detail: Tucuruí, Nova Ponte, Miranda, and Serra da Mesa. A comparison with all other reservoirs deeper than 30?m and 50?m suggests that the hazard for induced-seismicity varies within Brazil: the NE part of the intracratonic Paraná basin has higher hazard as compared with the southern part of the same basin. No correlation of the induced hazard with variations in natural seismicity can be observed.     

Crustal phases in earthquake location

For crustal earthquakes of moderate or large size, the first reported phases at distant stations are usually the first predicted by the crustal model used. For smaller events, however, or for stations at larger distances, the first phase detected is often a later crustal phase of larger amplitude. This may be eitherPg which travels entirely in the upper crustal layer with a velocity of about 5.6 km/s, orP ^* with a path mainly in the lower crustal layer at a velocity of about 6.7 km/s. Many earthquake location programs do not take account of these phases, and treat their arrivals as if they were the earlier phasePn. At the International Seismological Centre we re-identify up to 200 crustal phases each month. This often results in significant improvement in position, the direct determination of depth or even the obtaining of a formal solution where none was possible before. We find that these re-identifications are needed in most continental areas of the world, and that the original crustal model of Jeffreys and Bullen still remains a useful standard.On leave from the Seismological Laboratory, Institute for Earth Physics, P. O. Box MG-2, 76900 Bucharest-Magurele, Romania