Important legal notice
   
Contact   |   Search   
RTD info logoMagazine on European Research N° 43 - November 2004   
Top
 HOME
 TABLE OF CONTENTS
 EDITORIAL
 Goodbye, Mr Busquin
 Reducing congestion on the inflammation highways
 Childhood diets
 strengthening European research
 Neighbourhood science 
 A growing concern
 Capturing distant worlds on film
 COMMUNICATING SCIENCE
 IN BRIEF
 PUBLICATIONS
 AGENDA
 CALLS FOR PROPOSALS

Download pdf de en fr


LOCAL SEISMIC ACTIVITY
Title  Dealing with the local ‘site effects’ of an earthquake

Deep geophysical investigations to understand plate tectonics are essential to the evaluation of seismic risks. At the same time, they are generally of little use in forecasting the real damage an earthquake, sometimes of average magnitude, can cause on the Earth’s surface. There is always going to be a specific ‘site effect’ that is dependent on low-depth geology. Over the past decade, evaluations of this effect have been central to preventive earthquake research, focusing in particular on the phenomenon of seismic noise. The Sesame project has made a major contribution to our knowledge in this field.

The violent earthquake, measuring 8.1 on the Richter scale, that hit Mexico City in 1985 had its epicentre about 350km away, in the Michoacan region. Paradoxically, this area lying closest to the origin of the earthquake experienced much less destruction. The same scenario was repeated in San Francisco in 1989: the epicentre of the earthquake that measured 7 on the Richter scale was no more than 100 km from the city, but even at the very heart of the Californian bay, the extent of the tremors varied in intensity and the destructive force from site to site.

Every significant earthquake holds its own surprises in terms of the relative intensity at specific sites, such as occurred in Athens (EL) in 1999 and San Giulano di Puglia (IT) in 2002. These disparities – known as ‘site effects’ – have long been the subject of careful study by seismologists and are known to reflect the soil geology and topology of the immediate location. These geoseismological studies have progressively built up a valuable body of knowledge. We now know that solid geological formations (rocks) transmit seismic movements without transforming them, while sedimentary sandy or clay soils – that are much less rigid – serve to amplify the very destructive surface waves. Alluvial valleys are particularly sensitive in this respect. The same is true of sites where layers of a different nature are either superposed or adjacent. In-depth research has shown that such configurations can give rise to phenomena of seismic wave trapping, creating a resonating effect as in a violin, amplifying the intensity and the duration. Relief can play a similar role, although generally to a lesser degree. A wave crossing a mountainous region will concentrate its energy in the peaks that are consequently subject to more intense vibratory movements.

From expost to exante
Knowledge acquired of site effects analysed after the event naturally raises the question of whether or not the earthquakes could have been anticipated. If so, we would have a valuable preventive tool to be applied to town planning. It is advisable to carry out seismic microzoning operations before expanding towns, for example, and to adapt locally the minimum paraseismic construction standards to take account of subsoil configurations. This is particularly pertinent for the siting of engineering works such as viaducts, tunnels and dams, or risk industries, such as nuclear power plants and chemicals factories.

In terms of methodology, obtaining a reliable estimate of the site effect at a given location is a complex affair. A qualitative knowledge of geological formations and their geometry is far from sufficient, as it is also necessary to predict quantitatively how these formations are going to behave in response to real waves. Digital modelling is a promising tool but requires precise geotechnical and geophysical surveying, the cost of which our society seems – paradoxically – increasingly less prepared to meet. This makes it essential to explore other, less-costly surveying techniques that are able to provide reliable information either on these parameters or directly on the site effects themselves.

In areas with a very high seismic activity, it is sometimes possible to observe the ground response to low magnitude tremors, but such sources remain random. In the vast majority of regions with a dormant seismicity, there is no observable seismic activity of any significance for years or even decades at a time. Also, in urban areas, an added difficulty is posed by what is known as ‘seismic noise’. Although insignificant on the Richter scale, this is produced by human activities (earthworks, drilling, road and rail traffic, etc.) or certain natural phenomena (waves breaking on the coastline, avalanches in mountainous areas, etc.).  

An ingenious engineer
In 1989, the Japanese seismologist Nakamura nevertheless revolutionised the world of applied seismology. He was the first to suggest that this seismic noise, possibly caused by artificial surface vibrations, could be analysed to significant effect in regard to the site effect of a given location. His empirical methodology, known as the H/V technique, established that the spectral ratio between the vertical and horizontal components of noise recorded at a given site make it possible to establish its capacity for a frequential response to seismic waves. This sparked a surge of scientific interest in seismic noise analyses over the past decade. Other Japanese research teams have also shown that the ‘dense network’ or ‘array’ recording of ambient vibrations could also be a very effective tool in extracting quantitative parameters on the subsoil structure, in particular the speed at which seismic waves spread, and to a great depth.

Over a period of three and a half years, beginning in 2001, more than 80 researchers from 14 geophysics institutions worked on the vast Sesame project (Site effects assessment using ambient excitations). Their aim was to develop a better command of the theoretical and methodological bases for the use of seismic noise to produce valid estimates of the site effect. “The main attraction of this approach is its ease of use at a very low cost,” stresses Pierre-Yves Bard of the Laboratoire de géophysique interne et tectonophysique (LGIT) in Grenoble (FR). “However, it is necessary to make sure that this cost benefit is coupled with reliable results. There isn’t a scientific consensus yet on the validity of their physical basis or real confidence in the conclusions regarding site effects. The major drawback of these methods, and especially the H/V technique, is that they were developed in a purely empirical manner. They raise high hopes but, in the early stages, too little research was carried out to clarify their physical foundations. The danger is that the current enthusiasm for these techniques will have a distorting effect and lead to completely false results.”

The Sesame project first sought to elucidate the quite vague concept of seismic noise analysed at a given site. The first objective was to determine the physical characteristics of the noise sources. Are they local or distant? At what depth are they produced? And what are their temporal characteristics? A great number of site tests were carried out (in Basle in Switzerland, Colfiorito in Italy, Grenoble in France, Liège and Uccle in Belgium, and Thessaloniki in Greece) to measure the ambient noise and to compare this with the results obtained by digital modelling that artificially reproduces the batteries of vibrators located at many points in the locations studied. A very large number of digital experiments were also carried out on simple and perfectly controlled structures to test the capacity of these methods (H/V and array) to identify pertinent characteristics of the subsoil or site effects.

Simulation of seismic noise in the Colfiorito basin (IT)

Simulation of seismic noise in the Colfiorito basin (IT).(a) Map of local ID resonance frequencies at the site, calculated on the basis of the Colfiorito geophysical model. The lower the frequency, the greater the depth of the sediment resting on the rocky substrata.(b) Map of H/V noise frequency peaks obtained from the analysis of seismic noise simulations. The high degree of correspondence between these two maps shows that the H/V noise ratio method permits the correct imaging of variations in the thickness of covering sediments.(c) Relative percentage difference between local ID resonance frequencies obtained from a geophysical model and H/V noise peak frequencies.(d) Comparison between the local ID resonance frequencies obtained from a geophysical model and H/V noise peak frequencies. In the vast majority of cases, the two frequencies show a 20% similarity.

(a) Map of local ID resonance frequencies at the site, calculated on the basis of the Colfiorito geophysical model. The lower the frequency, the greater the depth of the sediment resting on the rocky substrata.

(b) Map of H/V noise frequency peaks obtained from the analysis of seismic noise simulations. The high degree of correspondence between these two maps shows that the H/V noise ratio method permits the correct imaging of variations in the thickness of covering sediments.

(c) Relative percentage difference between local ID resonance frequencies obtained from a geophysical model and H/V noise peak frequencies.

(d) Comparison between the local ID resonance frequencies obtained from a geophysical model and H/V noise peak frequencies. In the vast majority of cases, the correlation is excellent.

Disseminating knowledge

Software programmes and user guidelines have been developed for both these methods of analysis. In particular, significant improvements were made to the ‘array’ technique that, although very promising for measuring the propagation of surface waves, is not widespread in Europe. For the H/V technique, a practical guide for making and interpreting measurements was compiled, plus supporting multi-platform software. This could ultimately help ‘standardise’ this method and control its use. A use that could then become routine in the framework of paraseismic regulations.

The conclusions of the Sesame project were presented and discussed at the 13th World Conference on Seismic Engineering, held in Vancouver in August 2004. “We do not claim to have succeeded in answering all the complex questions posed by the use of seismic noise,” concludes Pierre-Yves Bard. “But I believe we have made an important contribution to the immense research effort carried out on this subject on every continent.”


Printable version

Features 1 2 3 4 5 6
  The united science of seismology
  Seismic warnings: the Icelandic laboratory
  Dealing with the local ‘site effects’ of an earthquake
  Understanding the furies of the Earth
  An educated awareness of risk
  The mechanics of tectonics


  TO FIND OUT MORE  
 
  •  

      CONTACTS  
     
  • Pierre-Yves Bard – LGIT - Grenoble (FR)
  •  


       
      Top
    Features 1 2 3 4 5 6

    TO FIND OUT MORE

    CONTACTS