You are here

The September 16, 2015 Chile Earthquake

18 Sep 2015

A Mw8.3 earthquake occurred at UTC time 2015-09-16 22:54:33.220 (NEIC) offshore of Chile near Illapel, with NEIC epicenter location of (31.570S, 71.654W). The earthquake ruptured a portion of the subduction zone to the north of the rupture area of the 2010 Mw8.8 earthquake. The earthquake generated strong ground shaking and a few meters of tsunami run ups at the nearest coast. More than a million people were evacuated after the earthquake due to the tsunami warning, and more than ten perished in this disaster. More are feared dead, according to BBC News, in this event reminiscent of the Mw8.8 quake that struck the coast of central Chile in February 2010. The recent quake is the strongest quake to strike this year, and is a sobering reminder that this type of event is expected in the Mentawai seismic gap in Sumatra, Indonesia, at the same latitude as Singapore.

Location of the quake indicated by the star, showing proximity to nearest cities. (Image credit: USGS)

Our scientists used 31 teleseismic P-waves and 24 SH-waves to invert the kinematic rupture process of the earthquake (see figures), assuming NEIC epicenter location and W-phase focal mechanism (reference as we did for Nepal, Ji et al. 2002). We used the fault plane with dip angle of 19 degree and strike of 5 degree. Here we rotated the strike of the fault plane by 10 degree to make it agree better with the strike of the trench. The slip model is then used to generate the seafloor deformation, which is used to calculate the tsunami (see figures).


Surface projection of the slip model, the epcienter is indicated by the star and the rectangle shows the boundary of the fault plane.

Comparison of the observed (black) and modeled (red) teleseismic seismograms (in displacement). Station names are indicated to the left of the traces along with the azimuths and epicentral distances in degrees. Peak amplitude in micron of data is indicated above the end of each trace.

Moment-rate function of the rupture model.

The model predicts the tsunami wave arrivals and amplitude remarkably well.

The seafloor deformation was calculated using Okada.1985. We assumed the water instantaneously follows the seafloor deformation that initiated the tsunami waves. We used COMCOT (Cornell Multi-grid Coupled Tsunami Model, Wang and Liu. 2006; Li et al. 2012 & 2014) to simulate tsunami wave propagation. Model predictions at local DART location 32401,32402, and 32412 are compared with the DART measurements. DART data are downloaded from Preliminary results indicated that the tsunami predications at these DARTs are slightly earlier than the data, suggesting that the spatial location of the slip is slightly close to the DARTs. As the hypocenter of the earthquake has large uncertainty, and was not well determined. Thus, we’ve done some tests by moving the source further south. Our preferred source model has to be shifted the original source model by 40km further south. This shifted model explains these DART data well.Maximum tsunami wave amplitude 7 hrs after the earthquake. The red dots show the DART location, and name of the DART is represented by the white numbers.

The slip model for this event can be downloaded here