Tsunamis are waves triggered by the vertical deformation of the ocean bottom, caused by submarine earthquakes or landslides. They lead to waves crossing the oceans at high speed (around 800 km/h), and a potentially enormous quantity of water flooding the coasts when these waves come to shore. Theoretically, sea level anomalies observed by altimetry should reflect these waves. However, observation is difficult, since the additional height is one of the signals of ocean variability. Studying the differences between the few altimetric observations and the tsunami propagation models should enable the scientific community to enhance their understanding of such phenomenon and to fine-tune the models. It is clear that only a multidisciplinary, multi-technique study can grasp all the forces at work here (geophysical, hydrodynamic, energetic etc.).

Altimeter Sea Level Anomalies account for many different ocean signals such as large-scale and mesoscale ocean variability. These signals considerably limit our ability to detect tsunami waves, or can at least significantly modify the observed characteristics. Most of these signals can be removed, however, using an ocean variability mapping technique. Note that this is possible only because at the time of the December 2006 tsunami, we had very good space/time sampling of the ocean with four altimeters (Jason-1, Topex/Poseidon, Envisat, GFO). Such a configuration is required to describe the ocean's mesoscale variability and thus to extract the signals generated by the tsunami from the background ocean variability signals.

It must be noted that satellite altimetry is not sufficient for the early detection and warning of tsunamis. Even with a four altimeter configuration (as it was the case during the 26 December 2004 Indian Ocean tsunami), the probability of observing a tsunami just after it is triggered remains low [Okal et al., 1999]. This also poses some specific data acquisition and processing issues (data processing time would hardly be compatible with the time required to issue an alert). The unique contribution of satellite altimetry is to better understand and improve the modelling of tsunami propagation and dissipation. In particular, reported observations from the 26 December 2004 Indian Ocean tsunami have been used to refine the initial displacement conditions due to the earthquake, so that observations match model outputs [Ablain, 2006].

The 2004 Indian Ocean tsunami

Until the Indian Ocean tsunami on 26 December 2004, tsunami observations by satellite altimeters had been relatively insignificant. Studies carried out in the past [Okal et al.,1999] show that TOPEX was the only altimeter to detect a tsunami caused by an earthquake in Nicaragua in 1992. The signal was not clearly observed because of its weak amplitude, close to 8 cm, and the great ocean variability in this area. The probability of a satellite altimeter observing a tsunami is low because it requires that the satellite overflies the tsunami wave almost immediately after it originates, due to the tsunami's great propagation speed (about 800 km/h in an ocean 5,000m deep). Tsunami signals in the open ocean are also quite weak.

Ground track for Jason-1 (top left) and Envisat (bottom left) overlaid by the CEA wave propagation simulation at the time of the satellite's passage. The area corresponding to the tsunami's front is circled. Sea level anomalies measured by Jason-1 and Envisat compared to the CEA simulation (top and bottom right). (Credits CEA and CLS).

The 2011 Japan Tsunami

On 11th March 2011, a 9.0- magnitude earthquake off Japan's north-eastern coast generated a huge wave which, breaking on the shore, devastated everything in its path. Three altimetry satellites (Jason-1, Envisat and Jason-2) observed the front wave over the Pacific Ocean between 5 and 22 hours after the earthquake. Envisat detected the first wave front 5 hours and 25 min after the earthquake with a trough-to-crest amplitude of more to 30 cm. Jason-1 measured a larger amplitude (>60 cm), 7 hours and 30 min after the earthquake.

Top: Ground tracks Jason-1 pass #147, at 7:30 hours after the quake (left) and Envisat pass #419 at 5:25 hours after the quake (right), overlaid by a wave propagation simulation at the time of the satellite's passage.
Bottom: Model tsunami (black lines) with the satellite altimetry data along the passes, respectively, Jason-1 (left, in red) and Envisat (right, in pink). Credits T.Y Song (2012, Geoph.Research Lett.).

Further information: