Lunar internal structure modeling using Apollo seismic travel time data and the latest selenodetic data from GRAIL and LLR

Abstract:

We tried to constrain lunar internal structure by combining the Apollo seismic travel time data and selenodetic data including those from GRAIL and LLR. We used a seven-layer model consisting of crust, upper mantle, middle mantle, lower mantle, low-velocity layer (LVL), fluid outer core and solid inner core. The model is constrained by the following observations; three selenodetic data of mass, mean moment of inertia, tidal Love number k₂ from [1], and Apollo seismic travel time data from [2]. Markov Chain Monte Carlo method is used to infer the model parameters. We collected 140 million samples from 10 chains. Viscosity is not taken into account in this calculation. Mean crustal thickness of 46 ± 4 km is estimated by fitting a normal distribution curve to the posterior distribution, which is to be compared with the previous estimate of 34 - 43 km [3]. Major part of crustal densities is sampled between 2500 and 2600 kg/m³, which is consistent with the value of 2550 kg/m³ reported by [3]. In general, seismic wave velocities in the mantle are consistent with the previous estimates [4][5]. The ranges of size and density of the outer core which satisfy the observation are relatively wide and it is difficult to tightly constrain them. Strong correlation between outer core size and LVL thickness is observed. The smaller outer core should be accompanied by thick LVL and vice versa. When we take into account the upper bound of the fluid core size of 400 km which is predicted by magnetic observation [6], the thickness of the LVL is at least about 100 km. The S-wave velocity within this low-velocity layer is estimated to be less than about 3 km/s. The effect of low viscosity [7] may change the estimate of the S-wave velocity in the LVL. The inner core radius is expected to be smaller than 280 km. The lunar displacement Love number is predicted to be h₂= 0.0423 ± 0.0004.

References

[1] Williams et al. (2014), JGR, doi:10.1002/2013JE004559

[2] Lognonne et al. (2003), EPSL, 211, 27-44

[3] Wieczorek et al. (2012), Science, doi:10.1126/science.1231530

[4] Weber et al. (2011), Science, 331, 309-312, doi:10.1126/science.1199375

[5] Garcia et al. (2011), PEPI, doi:10.1016/j.pepi.2011.06.015

[6] Shimizu et al. (2013), Icarus, doi:10.1016/j.icarus.2012.10.029

[7] Harada et al. (2014), Nature geoscience, doi:10.1038/NGEO2211