Marsquake locations and 1-D seismic models for Mars from InSight data

Abstract

We present inversions for the structure of Mars using the first Martian seismic record collected by the InSight lander. We identified and used arrival times of direct, multiples, and depth phases of body waves, for 17 marsquakes to constrain the quake locations and the one-dimensional average interior structure of Mars. We found the marsquake hypocenters to be shallower than 40 km depth, most of them being located in the Cerberus Fossae graben system, which could be a source of marsquakes. Our results show a significant velocity jump between the upper and the lower part of the crust, interpreted as the transition between intrusive and extrusive rocks. The lower crust makes up a significant fraction of the crust, with seismic velocities compatible with those of mafic to ultramafic rocks. Additional constraints on the crustal thickness from previous seismic analyses, combined with modeling relying on gravity and topography measurements, yield constraints on the present-day thermochemical state of Mars and on its long-term history. Our most constrained inversion results indicate a present-day surface heat flux of 22 ± 1 mW/m 2 , a relatively hot mantle (potential temperature: 1740 ± 90 K) and a thick lithosphere (540 ± 120 km), associated with a lithospheric thermal gradient of 1.9 ± 0.3 K/km. These results are compatible with recent seismic studies using a reduced data set and different inversion approaches, confirming that Mars' potential mantle temperature was initially relatively cold (1780 ± 50 K) compared to that of its present-day state, and that its crust contains 10-12 times more heat-producing elements than the primitive mantle. Plain Language SummaryThe seismic recordings from the InSight mission have proven that Mars is an active planet. Among the several 100s of detected marsquakes, 17 have a sufficient quality to constrain the internal structure of Mars. We found that most of these marsquakes occurred at depths shallower than 40 km, and are located in the Cerberus Fossae region. There are faults in this area, which could be a source of quakes. An important finding is that as on Earth, the crust is made of two types of rocks formed when hot molten material is cooling, quickly near the surface, and slowly in depth because temperature under the planet's surface is higher. Combining our seismic data with other independent geophysical measurements, we are able to reconstruct the thermal history of Mars. Our results indicate that Mars has a relatively hot mantle, and that the uppermost mantle temperature was initially colder than at the present.

Low Frequency Marsquakes and Where to Find Them: Automated Event Back Azimuth Determination Using a Multi-Body Wave Polarisation Analysis Approach

“…The authors present their own phase arrival times and do not use the MQS Catalog V8 phase picks, so the results are independent. We find that our results agree well, with the largest deviation for S0474a (21 • (Drilleau et al, 2022) versus 97 • (this paper)). For 4 events (S0154a, S0474a, S0484b, S0918a) our method delivers a significantly reduced uncertainty, increasing the value of the result for tectonic interpretation.…”

“…In addition we compare our results to those from Stähler et al ( 2021) SM 4, currently under revision in Drilleau et al (2022), who estimate back azimuths for a set of events based on Marsquake Catalog Version 8 (V8). On top of the method described above, the instantaneous phase coherence between vertical and horizontal components as a function of azimuth.…”

“…However, the SV-wave component should produce a weak maximum of both correlation coefficient and phase coherence along the correct back azimuth. Drilleau et al (2022) analyse the signals in different time windows around their phase picks as well as a noise window. They use fixed frequency ranges for the analysis, 0.4-1.0 Hz for P-waves and 0.3-1.0 Hz for S-waves, owing to S-waves' lower frequency content.…”

“…MQS uncertainties are described in Clinton et al (2021). Results from Drilleau et al (2022) are given in 'Drilleau (2022)'. SNR is calculated from the P-wave and indicated noise window in the same frequency band where the back azimuth is estimated.…”

“…Events in italics have less clear results and are more speculative. We also show results from Drilleau et al (2022). The authors present their own phase arrival times and do not use the MQS Catalog V8 phase picks, so the results are independent.…”