Abstract
Titan’s surface icy shell is likely composed of water ice and methane
clathrate [1, 2]. Methane clathrate may play a role in Titan’s
methane cycle [3–5] affect Titan’s thermal profile [6] , and
may affect the habitability of Titan’s ocean. Although the bulk
properties of clathrates are similar to those of pure water ice, the
thermal conductivity of methane clathrate is about 20% the value for
pure water ice [7, 8]. The lower thermal conductivity acts to
insulate Titan’s icy shell, changing the thermal profile of Titan. As
seismic wave speeds [9, 10] and attenuation [11] are dependent
on temperature, any changes to the thermal profile will result in
changes to seismic waveforms recorded by seismic instrumentation. Here,
we compare the seismic waveforms of model with a 100 km thick pure water
ice shell, versus a model with a 10 km clathrate lid over 90 km of pure
water ice. Our results have implications for the upcoming Dragonfly
mission, which will carry seismic instrumentation as part of its payload
[12]. Methods: We use PlanetProfile [13] to create interior
structures models of a pure water ice shell and a model with a pure
water ice shell with a 10 km clathrate lid. The interior structure
models are used as inputs with AxiSEM [14] and Instaseis ([15]
to generate seismic waveforms. We interpret the results to quantify the
differences in seismic velocities, arrival times of seismic phases, and
amplitudes of seismic waveforms at the surface of Titan. Results: The
interior structure models show a clathrate lid will reduce the
conductive lid thickness by ~ 2/3 compared to the pure
water ice shell model. As a result, the clathrate lid model reaches
higher temperatures at shallower depths (Figure 1a). The temperature
profile affects the seismic velocity (Figure 1b), and the seismic
quality factor (Q, Figure 1c) profiles. A clathrate lid creates a
steeper negative gradient in seismic velocities and Q. The greatest
difference in seismic velocities occurs at the base of the clathrate lid
(Figure 2). Because of the change in seismic velocities, the arrival
times and observable distances of seismic phases will be different
between the two models. Using TauP [16], we calculate the
differences for several seismic phases. We find that the change in
seismic velocity profile results in a difference of a few seconds at
most in arrival times. The range of observable distances will also vary
by a few degrees. The small changes might be noticeable on waveforms,
but would require high signal to noise ratios, and precise
determinations of location and depth of the event. The changes in
seismic velocities and Q will also impact the observed ground motion.
Using AxiSEM and InstaSEIS, we create a database of seismic waveforms
spaced 1 degree in epicentral distance. We compare the same event
magnitude and distance between source and seismometer for the two
models. For each waveform we calculate the root mean square (RMS) using
ground acceleration.