However, with only one seismometer this is a very challenging task. In contrast to a seismic network on earth with numerous stations, additional reference points that constrain the origin of the signals registered is missing. In preparation of the mission, scientists with experience in earthquake location and characterisation problems are invited to share their knowledge by participating in a blind test.

Based on a synthetic dataset, we aim to improve the planned single station location methods to be used in the routine analysis of the martian dataset. The waveforms provided in the blind test mimic both the streams of data that will be available from InSight, as well as the expected level of tectonic and impact seismicity and the noise conditions on Mars. The test is 'blind' in the sense that the actual event catalogue is not provided to any participants, and the structural model used to generate the seismograms has been selected from a suite of 14 candidate models. In course of the blind test, we hope to learn from the methodologies proposed by the community. The blind test is described in detail in a recent publication in SRL (Clinton et al., 2017, Preparing for InSight: an invitation to participate in a blind test for Martian seismicity, Seism. Res. Letters, doi: 10.1785/0220170094).

The test officially opened on 1 August 2017 and registration will close on 1 October 2017. All participants must provide their event catalogues by 1 February 2018.  A website where interested parties can register and access seismic waveforms is available at http://blindtest.mars.ethz.ch.

All groups who contribute a catalogue to the blind test will be invited to be co-authors on a paper summarising methods and performance with respect to the true event catalogue that will then be released.

We encourage scientists and students, in teams of any size, to investigate the blind test dataset and we look forward to your participation!

The earthquake on Friday morning also triggered a tsunami warning. Tsunami waves of up to one metre high have been measured along the Mexican coast so far (www.tsunami.gov). However, there may well have been even larger waves at places not equipped with measuring instruments.

The Cocos Plate slides under the North American Plate off the Mexican coast, and the relative movement of the two plates generates considerable tension that is repeatedly discharged in the form of earthquakes. Initial assessments have classified this latest quake as a slip on a steeply dipping fracture surface (USGS, https://earthquake.usgs.gov/). Combined with the depth of the earthquake, this indicates that the earthquake did not occur directly on the boundary between the Cocos and North American Plates, but at a greater depth within the Cocos Plate.

The Mexico earthquake was registered by the SED's monitoring stations in Switzerland too, with the result that a minor, 1.6-magnitude quake in Göschenen (Canton of Uri) at 7.46 a.m. was automatically recorded as having a magnitude of 3.0. A manual appraisal soon identified this as a false alarm and the magnitude was corrected to 1.6. The automatically-calculated estimate was too high due to waves from the Mexico quake, which were still being registered in Switzerland one hour after the earthquake hit Central America. The Göschenen earthquake happened at the same time as the waves reached Switzerland, so the readings for it were excessively high.

Image on the left: Intensity map for the Mexico earthquake, showing a magnitude of 8.1 (USGS). Shaking reached a maximum intensity of VIII to IX on the Mexican coast. However, the earthquake also caused strong tremors in inland Mexico and in Guatemala. Image on the right: Surface waves from the magnitude-8.1 Mexico earthquake superimposed onto the waveforms for the magnitude-1.6 Göschenen earthquake. The Göschenen earthquake was automatically assigned the wrong magnitude (magnitude 3) as a result.

After a first preliminary analysis, the event was attributed a seismic magnitude of 6.3 (Mb). This makes it one magnitude higher than North Korea's previous nuclear test, conducted on 9 September 2016. The seismic data cannot tell us whether the recorded explosion was actually caused by a nuclear test or whether it involved the detonation of a hydrogen bomb or was the result of a conventional nuclear test.

The figure shows the seismograms of the powerful explosion recorded by the DAVOX station.