Hello I am Jean-François Semblat and my research mainly focuses on earthquakes and seismic waves. Some of my colleagues investigate the dynamic behaviour of the entire Earth as it may vibrate at very very low frequencies as shown on these videos. In my research, I am focused on the Earth's crust , where the tectonic plates may generate earthquakes. The rupture process at the source radiates seismic waves along various paths into the geological layers. These waves propagate along the three directions of space and the displacement may be defined by three different motion components. The geological structure is highly heterogeneous and, last but not least, alluvial layers (in blue in the top schematic) may strongly amplify the seismic waves near the surface. So, I will explain how the previous videos on wave propagation allow you to investigate SEISMIC waves. I will then focus on the AMPLIFICATION of seismic waves in soft geological structures which is the main target of my research. We will first consider the green layer in this schematic. It is assumed homogeneous, elastic and isotropic. As shown by Poisson in 1828, the wave equation may be split into two parts. Poisson has been student and professor at Ecole Polytechnique. He proposed uncoupled wave equations leading to uncoupled solutions called body waves. The first one is a scalar wave equation. So we get a single motion component and a first wave velocity V_p . It is equal to the sqrt[( lambda+2 mu ) / rho], where lambda and mu are the Lamé elastic parameters. The second one is a vectorial wave equation. We thus get two different motion components and a second wave velocity V_s. It is equal to the sqrt( mu / rho ). Consequently, V_s is always less than V_p! Considering these body waves, how does the ground move during the wave passage? For P-waves, the ground motion is only along direction X which is also the direction of propagation. That is why P-waves are also called longitudinal waves. As you can see, the ground is successively in a compressive and tensile state. For S-waves, the second wave type, the ground motion is always perpendicular to the direction of propagation. As shown in this video, it may be along the Z axis and the waves are thus called SV-waves since the motion is vertical. The ground is subjected to shear in the XZ plane. For S-waves , the ground motion may also be along the Y axis. The motion is then horizontal and we get so-called SH-waves. The ground is subjected to shear in the XY plane. Let us move to real quakes! And actual seismic waves!! As you can see for this Magnitude 4.6 earthquake recorded in the city of Nice, the amplitude of the acceleration changes very quickly with time. In the first part of the accelerogram, the fastest waves may be identified as P-waves. In the second part of the signal, S-waves appear. Their velocity is less but their amplitude is much larger! In my research, I investigate seismic waves propagating in complex geological structures such as the case of Mexico city. In this figure, you can see that the city is about 400 km far from the Pacific coast. The 1985 earthquake occurred along the coast and struck Mexico city strongly. The seismic waves were recorded at different distances from the coast. Very close to the epicenter, the first accelerogram reached a maximum value of 150 cm/s^2. Much farther, the second station showed a significant amplitude decrease with a maximum amplitude of 18 cm/s^2. When reaching the city of Mexico , the acceleration increased significantly. Finally, at the centre of the city, the acceleration reached 170 cm/s^2 which was even larger than at the first station! This phenomenon is due to the amplification of seismic waves in the thick soft sediments located underneath the city of Mexico. Since the amplification of seismic waves mainly occurs at alluvial sites, it is thus called seismic site effects. In my field of research, we may directly estimate the amplification of seismic waves in simple cases. If we consider a single elastic layer on an elastic half-space, we may compare the ground motion at point A to that at point B. In the frequency domain, I usually perform such a comparison by computing the transfer function of the wave in the layer, T(omega), as the ratio between u_A and u_B. The expression of the transfer function is very easy to obtain and is expressed as the inverse of cos of k_z1, the vertical wavenumber, multiplied by h the thickness of the layer. Let us plot the variations of the transfer function with respect to frequency. As you can see, T(omega) reaches very large values at some specific frequencies. It means that, for such frequencies, the ground displacement at point A may be much larger than that at point B. It is due to the resonance of the surficial elastic layer. The first resonance frequency is proportional to the wave velocity in the layer (V_s in this case) and inversely proportional to the layer thickness h. The factor 4 corresponds to the fact that we first get a quarter wavelength resonance of the layer! Well, how do we analyze site effects in complex geological configurations? With my colleagues of the University of Grenoble, in the French Alps, we investigate seismic waves propagating in the deep alpine valley where this city is located. The valley has a complex structure and an irregular Y-shaped geometry. As shown by the seismic recordings, the ground motion displayed in blue at the rock site is much smaller than that observed within the valley, plotted in red. It may be explained by the amplification of the seismic waves in the deep soft deposits. The geometry of the valley also influences the amplification process. To calculate the amplified ground motion in such deep and complex valleys, we thus need numerical methods. In this movie, we shall see the seismic ground motion due to the fault located at the top right of this figure. The seismic waves are first radiated from the fault and then propagate into the valley. The seismic waves are amplified due to the velocity contrast in the geological structure. They are also reflected at the valley edges and thus trapped in the alluvial deposit. In engineering seismology, we need such ground motion maps to characterize the seismic excitation on structures. As you can see, it may be much larger at some specific locations! To summarize, my research activities on seismic waves. I discriminate between various wave types : P-waves and S-waves which are body waves. Surface waves are also a very important issue. I simulate seismic wave propagation in complex geological structures in order to estimate the ground motion amplification at the surface. Finally, one key target of my research in earthquake engineering is to avoid the coincidence of the resonance of the soil and the resonance of the structure! It is a way to design safer buildings and perhaps even earthquake proof cities!!