P. F. Biagi

Very slow-moving crustal strain disturbances

Abstract A model of discrete crustal blocks, separated by a thin weak transition zone whose stress-strain relations contain derivatives of real order implying a memory mechanism in the medium, seems to give a good explanation for the low velocity values (0.03–15 cm/s) of slow moving crustal strain disturbances. Tilt and strain anomalies observed before moderate and large earthquakes are thought to be generated in the focal zone of the forthcoming earthquake, due to the spatial and temporal redistribution of the physical processes that occur during the preparation time of the earthquake. On the basis of experimental results it has been assumed that these anomalies propagate from the hypocentral area. In addition post-seismic deformations show the effects of propagation. The anomalies observed have a duration and a velocity ranging from minutes to months and from O.O5 cm/s to 10 cm/s, respectively. The velocity of migration of post-seismic deformation also lies in this range. The velocity of propagating strain disturbances is obtained as a function of the frequency and of the linear dimension of the oscillating blocks. The velocity values calculated using this model range from 0.1 cm/s to 10 cm/s, for viscosities of 10 9 –10 11 ) g cm −1 s −2+z (0.1≤ z ≤ 0.5) and blocksizes (1–4) · 10 6 cm, at frequencies of 10 −5 -10 −8 s −1 .

Anomalies in geophysical and geochemical parameters revealed on the occasion of the Paravani (M = 5.6) and Spitak (M = 6.9) earthquakes (Caucasus)

Abstract On the occasion of the Paravani (May 13, 1986) and Spitak (December 7, 1988) earthquakes, which occurred in the Caucasus, irregularities in several geophysical and geochemical parameters were revealed. In this paper, data on anomalies in the water level in deep wells, in the helium content in thermal waters, in electromagnetic signals revealed by surface and underground antennae, and in an F ( σ gt ) function based on traveltimes of P and S waves, are presented. Data were collected from sites located in the Georgian S.S.R. In the case of the Spitak earthquake, the anomalies in all the presented parameters can be attributed to the processes that preceded and accompanied the earthquake. For the Paravani earthquake, no significant precursory phase seems to have been reflected in the helium content in thermal waters, though this phase can be observed in the other parameters. In general, it would appear that the magnitude of the earthquake and the distance from the epicentre exert a significant control on the appearance of the anomalies.

Possible Creep-related Tilt Precursors Obtained in the Central Apennines (Italy) and in the Southern Caucasus (Georgia)

Daily averaged tilt component data from two sites of the Central Apennines (Italy) and of the Southern Caucasus (Georgia), respectively, revealed intermediate-term tilts as possible precursors to earthquakes (M=3.0÷4.7) which occurred in the above-mentioned seismic areas within a distance of 50 km from the sites. A good temporal correlation as well as a fair spatial correspondence between these residual tilts (with amplitude and duration of some microradians and months, respectively) and main shocks were pointed out, by removing both secular trends and seasonal thermoelastic effects from the raw tilts. An attempt was made to justify the above-mentioned results, based on the assumption that the observed intermediate-term preseismic tilts are the manifestation of aseismic creep episodes of comparable duration in the fault materials of thrust faults close to the tilt sites. The mechanism refers to a strain field slowly propagating from the preparation focal area to the tilt site, through crustal blocks separated by weak transition zones. This propagation is thought to be the cause of the local aseismic fault slip recorded by the tiltmeters. Previously, both discrete structures and strain propagation effects were revealed in the Central Apennines and are thought also to exist in the Southern Caucasus. As in the past, the rheological properties of fault materials are revealed as viscoelastic ones. In fact, creep equations obtained by applying several viscoelastic models on our data, proved to fit quite well some of the observed tilt precursors, producing viscosity and rigidity values very similar to those reported in literature.

A digital recording system of electromagnetic emissions

SummaryA new equipment able to reveal the electric component of electromagnetic emission and suitable for underground deployment has been made. The system enables to measure the absolute value of integrated voltage of signals. Data are recorded by a solid-state memory circuit. The system is performed as a plug-in modular design and it can be also used in diverse fields of application.RiassuntoÈ stata realizzata una nuova strumentazione per la rivelazione della componente elettrica delle emissioni elettromagnetiche in laboratori sotterranei. Il sistema è in grado di misurare i valori assoluti della tensione integrata dei segnali. I dati vengono registrati su un circuito di memoria a stato solido. Il sistema è stato realizzato con caratteristiche modulari e può essere utilizzato anche in altri campi di applicazione.РезюмеСконструировано новое оборудование для регистрации электрической компоненты электромагнитных излучений в подземных лабораториях. Система позволяет измерять абсолютную величину суммарного напряжения сигналов. Данные регистрируются в схеме с твердотельной памятью. Предложенная система собрана на специальных сменных модулях и может быть использована для различным прикладных задач.

No preseismic evidence from hydrogeochemical parameters on the occasion of the April 29, 1991, Georgian earthquake, Caucasus

Abstract The behaviour of helium content in thermal waters and of water level in deep wells was investigated on the occasion of the strong (M = 6.9) Georgian earthquake which occurred in the Caucasus on April 29, 1991. Data were collected from a measurement network located in Georgia. Just one significant preseismic variation appeared in the water level which seemed to be connected with the Georgian earthquake. In other wells, only some co-seismic and post-seismic effects are visible. No statistically viable variation was noted in the helium content. In contrast, clear anomalies in these parameters were observed on the occasion of the 1988 Spitak earthquake with the same magnitude which occurred in the Caucasus. We prepose that the location of the epicentre with respect to the measurement network was the reason for this disagreement. We therefore suggest the introduction of a coefficient c into relationships that give the size of the premonitory area as a function of the magnitude of the forthcoming earthquake. This coefficient, ranging from 0 to 1, can take into account the extent to which the focal zone belongs to different tectonic structures and the extent to which the measurement network coincides with these structures.

Tilt measurements and seismicity in Central Italy over a period of approximately three years

Tilt data obtained from July 1983-January 1986 at three sites in the Central Apennines (Italy) have been examined. A filtering of raw recordings indicated the existence of residual anomalies. Attempting a correlation between such anomalies and regional seismicity (3.0 </ magnitudeM </ 5.4) has allowed us to interpret the anomalies as possible preseismic variations. The results seem to confirm the existence of a relation between precursor time T, earthquake magnitude M and epicentral distance R. Finally, the presence of a time delay between the onsets of the anomalies detected at different sites for the same earthquake gives further credit to the possibility of the existence of their propagation.

Ground Tilt anomalies accompanying the main earthquakes occurred in the central apennines (Italy) during the period 1986–1989

SummaryA four-year set of tilt data (1986–1989) recorded at three sites of the Central Apennines (Italy) and which in the past pointed out quite good raw tilt responses to seismicity, only with moderate thermoelastic effects, have been investigated to search for the possible existence of tilt anomalies accompanying the main earthquakes of the region. It has been observed that, even if fairly good tilt responses to principal evants seem to appear, correlations of similar or better quality are obtained when comparing the same tilts with pressure and mainly with temperature. However, when taking into account earthquakes with magnitudesM≥3.0 and more strictE (strain) andL (source dimension) threshold values (the last two parameters being used to evaluate tilt field at a distanceR), only one event (M=3.5;R=5.1 km) has been selected and tectonic from non-tectonic contributions have been pointed out clearly at one tilt site. The quoted earthquake showed the bestE andR/L values (E=10−6;R/L=2.4) in the whole period and it was preceded by a clear tilt anomaly with amplitude and duration of about 5 μrad and 2 months, respectively. Since the tilt signal in this period is not affected by any non-tectonic effect (maximum correlation coefficient with temperaturer=0.2) the quoted anomaly has been considered as a probable precursor of the above-mentioned earthquake it preceds. Residual, tilt vector oscillations along the direction-including the epicentral area of this event, were also observed. On the contrary, all the remaining anomalous tilts proved to be well justified with pressure and/or temperature variations (r values up to 0.8).

Transmission and reflection of Rayleigh waves at a thin low velocity vertical layer: A laboratory and theoretical study

A three-dimensional ultrasonic modelling was used to study the Rayleigh waves reflected from a thin low velocity vertical layer and those transmitted through the same zone. The reflection and transmission coefficients of the Rayleigh waves have been calculated by an approximate method based on the Green function technique. Nonrigid contact boundary conditions were used for the low velocity layer modelling. A rather good correspondence between the theoretical and experimental absolute values of the reflection and the transmission coefficients was obtained. Some discrepancies between experimental and theoretical results, in particular with regard to phase shifts, may be explained by a certain inadequacy of the theoretical model to the low velocity layer investigated by the ultrasonic modelling.

Ground tilt variations detected in the central Apennines (Italy) in the period 1986-1989 and their correlation with seismicity

SummaryGround tilt measurements carried out at three tilt sites of the Central Apennines (Italy) revealed a quite good correspondence with seismicity (M≤4.9) which took place within a distance of 100 km from the sites and over a period of four years (1986–1989). The best correlation is obtained when the tilt signals are delayed by about 1.5 months with respect to the seismic one (number of earthquakes per month), suggesting that seismicity is predictable to a certain extent from the local tilt response. Correlations of similar quality showed also that tiltmeters located at different sites seem to react in a different way to events distributed along the two main active fault zones (namely, the Apennine and anti-Apennine ones), indicating an anisotropic response of the medium. To explain how the seismicity, selected along the Apennine and anti-Apennine directions, give rise to different tilt responses at tilt sites located at a relative small distance, the behaviour of the geological structures and of the geodynamic processes of the investigation area have been taken into account.

Sismogenesi e basse velocità crostali negli Appennini

Sulla base di evidenze sismologiche e del movimento crostale della zona centro-appenninica abbiamo proposto un modello per spiegare ľandamento del massimo sforzo di taglio con la profondita e ľesistenza di basse velocita di propagazione di perturbazioni elastiche (delľordine di 1 cm/s) attraverso blocchi crostali. Il campo di stress considerato e quello causato dal carico topografico e dalle corrispondenti masse di bassa densita che causano le anomalie negative di Bouguer osservate nella regione. Il massimo sforzo di taglio, dovuto allo spostamento relativo orizzontale tra carico topografico e anomalie negative, puo raggiungere il valore di circa 200 bar.