Long Period Seismology on Europa

Fabio Cammarano, Mark Panning, Ved Lekic, Michael Manga, Barbara Romanowicz

Introduction

Seismological observations provide unparalleled capability for studying planetary interiors. While seismological studies of the Earth and, to a lesser extent, the Earth's Moon have placed strong constraints on the internal structure and dynamics of these bodies, the absence of seismic measurements on other planetary bodies has stymied analysis of their detailed structures. There are a wide variety of internal compositions and structures hinted at by recent exploration of the Jovian and Saturnian systems. In order to inform future mission design, it is important to determine what observations hold potential for answering outstanding questions concerning planetary interiors.

Physically consistent models of planetary bodies, constrained by moment of inertia measurements and by well-characterized elastic and anelastic properties of relevant minerals, make possible the study of the seismic response of planetary bodies, even when seismic measurements are not yet available.

Here, we summarize the main results of such interdisciplinary approach for the Jupiter's moon Europa. We develop a menagerie of physically consistent models of Europa (Cammarano et al., 2006) which allows us to explore which seismic measurements on Europa have the potential to answer the many outstanding questions about its structure and current thermal state (Panning et al., 2006).

Physically consistent interior models

We calculate a range of thermodynamically consistent models for the physical structure of Europa, as constrained by the satellite's mass and moment of inertia. We start with either a pyrolitic or a chondritic mantle composition and a core of either pure iron or iron plus 20% sulfur. The models completely characterize the radial seismic structure, i.e. elastic and anelastic properties, and they can be used to compute the seismic response of the planet.

The coupling between the thermal state of the ice shell and its viscosity dictates the ice-shell thickness and its seismic properties. It is likely that attenuation could be very high within the $\lq\lq$warm$''$, convective part of the ice shell. Due to the feedback between radiogenic and tidal heating, two extreme thermal profiles are possible in the mantle (see Figure 1). Strong dispersion and dissipation are expected in the hot convective mantle, while anelasticity effects will be much weaker in the case of the cold mantle.

Figure 35.1: Physically consistent models for hot (solid lines) and cold (dashed) thermal structures with a pyrolitic mantle. Purely elastic models without dissipative effects are shown in light gray. The bottom panel shows a mantle close-up of the same models.

There is a strong relationship between different thermal structures and compositions. The $\lq\lq$hot$''$ mantle may well keep temperatures high enough to be consistent with a liquid core made of iron plus light elements. In the case of the $\lq\lq$cold scenarios$''$, the possibility of a solid iron core cannot be excluded and it may even be favored. The depth of the ocean and of the core-mantle boundary are determined with high precision once we assume a composition and thermal structure. Furthermore, the depth of the ocean is not very sensitive to the core composition used.

Predicted Seismic Response

The normal modes for the determined seismic radial structures of Europa are computed with the MINOS code (Woodhouse, 1988). Seismograms for any proposed source and receiver configuration can then be modeled using normal mode summation, which models the complete broadband seismic wavefield. Given the normal mode catalogs and a predicted seismic source process, we can compute synthetic seismograms at any distance from the source. The seismograms presented here assume a M$_W=$5 (seismic moment of 3.94 x 10$^{16}$ Nm) normal faulting source, as proposed in Nimmo and Schenk (2006). For simplicity, we used a dip-slip event with a 45$^{\rm o}$ dip and 90$^{\rm o}$ rake. We computed all modes up to 0.1 Hz, and then bandpass filtered the seismograms with corner frequencies at 12 and 800 seconds period, and cut-off frequencies at 10 and 1000 seconds period.

Figure 35.2: Synthetic displacement seismograms at a distance of 25$^o$ (680 km) from the M$_W=$5 normal event. Seismograms are calculated for the low-attenuation cold chondritic model with ice shell thicknesses (from top) of 5, 20, 40, and a solid 137 km thick ice layer model. The maximum amplitude for each panel is shown to the left of the panel.

We send to Panning et al. (2006) for a detailed discussion about measurement requirements and potential for answering questions on Europa's interior. Compared to high frequency signals, that require a surface installation, long-period measurements can be acquired potentially by an orbiter. Unless an orbiter is at the correct altitude for geosynchronous orbit, the seismic measurements will be made at a moving point on the surface. This presents additional challenges, but may provide us with interesting methods for determining surface wave velocities from a single measurement. To test this, we adapted the mode summation code to synthesize seismograms at a moving observation point. Seismic displacement from an event that occurs near the trajectory of the orbiter recorded on an observation point moving away from the source location produces a seismogram at sufficient time after the event with a resonant frequency (Figure 3). This resonance is caused by a wavepacket of a given frequency having a group velocity which closely matches the velocity of the observation point. Because the frequency at which the group velocity will match a given orbital velocity depends on the ice shell model, an observation of this resonance phenomenon may be diagnostic.

Figure 35.3: 3000 seconds of seismic displacement (in m) for an observation point that starts 15$^o$ (410km) east of the source at the event origin time, and moves north with an apparent surface velocity of 1.4 km/s calculated in the low attenuation 5 km thick ice shell model. The arrivals at about 120 s and 350 s are the P and S waves respectively.

Conclusions

Long-period seismic observations on Europa have potential to greatly expand our knowledge of the satellite. Long-period displacement measurements with millimeter accuracy may be able to determine the current tectonic activity of Europa's surface, the presence of a liquid ocean, and the thickness of the ice shell. These observations hold considerable promise relative to shorter period acceleration and velocity measurements, as the peak amplitudes occur at frequencies that minimize the complications from unknown 3D heterogeneity and finite source dimension and duration. Such displacement measurements may be possible from orbit, but many instrument design and data processing details need to be carefully considered. Determination of deeper structure with seismic measurements is much more difficult in the presence of a global liquid subsurface ocean, which acts to decouple deeper seismic energy from the surface.

References

Cammarano, F., Lekic, V., Manga, M., Panning, M., Romanowicz, B. (2006) Long-period seismology on Europa: I. Physically consistent models, J. Geophys. Res. - Planets, in press.

Nimmo, F., Schenk, P. (2006), Normal faulting on Europa: implications for ice shell properties, J. Struct. Geol., in press.

Panning, M., Lekic, V., Manga, M., Cammarano, F., Romanowicz, B. (2006) Long-period seismology on Europa: II. Predicted Seismic Response J. Geophys. Res. - Planets, in press.

Woodhouse, J.H., (1988) The calculation of eigenfrequencies and eigenfunctions of the free oscillations of the Earth and the Sun, Seismological Algorithms, 321-370, ed. Doornbos, D.J.