International audienc

The InSight lander will deliver geophysical instruments to Mars in 2018, including seismometers installed directly on the surface (Seismic Experiment for Interior Structure, SEIS). Routine operations will be split into two services, the Mars Structure Service (MSS) and Marsquake Service (MQS), which will be responsible, respectively, for defining the structure models and seismicity catalogs from the mission. The MSS will deliver a series of products before the landing, during the operations, and finally to the Planetary Data System (PDS) archive. Prior to the mission, we assembled a suite of a priori models of Mars, based on estimates of bulk composition and thermal profiles. Initial models during the mission will rely on modeling surface waves and impact-generated body waves independent of prior knowledge of structure. Later modeling will include simultaneous inversion of seismic observations for source and structural parameters. We use Bayesian inversion techniques to obtain robust probability distribution functions of interior structure parameters. Shallow structure will be characterized using the hammering of the heatflow probe mole, as well as measurements of surface wave ellipticity. Crustal scale structure will be constrained by measurements of receiver function and broadband Rayleigh wave ellipticity measurements. Core interacting body wave phases should be observable above modeled martian noise levels, allowing us to constrain deep structure. Normal modes of Mars should also be observable and can be used to estimate the globally averaged 1D structure, while combination with results from the InSight radio science mission and orbital observations will allow for constraint of deeper structure

[1] Despite the tight constraints put by seismology on the elastic properties of the Earth's lower mantle, its mineralogical composition and thermal state remain poorly known because the interpretation of seismic measurements suffers from the trade-off between temperature, iron content, and mineralogical composition. In order to overcome this difficulty, we complement seismic data with electromagnetic induction data. The latter data are mostly sensitive to temperature and iron content, while densities and acoustic speeds mostly constrain the mineralogy. A 0.5 log unit increase in electrical conductivity can be caused either by a 400 K increase of the temperature or by an increase of iron content from 10% to 12.5%. Acoustic velocity is only marginally sensitive to temperature but it increases by 0.8 km s À1 on average as the perovskite fraction increases from 50% to 100%. Olsen's (1999) apparent resistivities in the period range [15 days, 11 years], and Preliminary reference Earth model (PREM) densities and acoustic speeds are jointly inverted in the depth range [800 km, 2600 km] by using a Monte Carlo Markov Chain method. Given the uncertainties on these data, estimates of perovskite fraction are well constrained over the whole depth range, but information on temperature and iron content is only obtained for depths less than 2000 km, corresponding to the penetration depth of the long-period electromagnetic field. All parameter values are determined with an uncertainty better than 15-20% at the 1s confidence level. The temperature in the uppermost lower mantle (i.e

No abstract