Investigating the Viscosity Structure of Venus

The interior structure and geodynamic style of Venus are largely unknown. One of the most informative ways available to investigate the planet’s interior is through the joint analysis of gravity and topography datasets, lastly obtained by NASA’s Magellan mission, which orbited the planet between 1990 and 1994 (Konopliv et al., 1999; Rappaport et al., 1999). Several studies have shown that the long-wavelength (over 1000 km) gravity and topography signature of Venus is dominated by convective flows in the mantle (e.g., Sjogren et al., 1980; Banerdt, 1986; Kiefer et al., 1986), commonly referred to as dynamic support.

The analysis of long-wavelength gravity and topography has been performed in several investigations to better understand the properties of Venus's mantle. Many of these utilized the dynamic loading model developed by Hager and Clayton (1989), in which stresses driven by density anomalies linked to mantle flows cause surface deformations that depend on viscosity variations. This model has been applied to estimate the spatial distribution of mantle mass anomalies (e.g. James et al., 2013) and to investigate the planet's mantle viscosity structure (Pauer et al., 2006, Maia et al., 2023). However, these studies applied a simplistic viscosity profile, in which the mantle is divided into 4–5 layers of constant viscosity, which can freely vary in magnitude. This approach has the drawback of considering a large number of free parameters, which linearly increase with the number of viscosity layers added, and allow for physically unrealistic viscosity structures.

We present a new investigation that adopts a more robust parametrization approach, using an Arrhenius law, often employed in geodynamic models (Breuer and Moore, 2015), in which the mantle viscosity profile is pressure- and temperature-dependent. This approach not only allows for more physically robust viscosity structure but also directly provides insights into key rheological parameters, such as the activation volume, mantle potential temperature and thermal lithosphere thickness, which are essential for constraining geodynamic models. In this approach, we also allow for the presence of a viscosity jump, with its order of magnitude and depth set as free parameters.

During this investigation, we also explore two different parametrizations of the mantle density anomalies: a single-mass sheet and a depth-constant density anomaly. In the single mass-sheet scenario, density anomalies are assumed to be concentrated within a thin surface mass-sheet at a depth treated as a free parameter, yielding density anomalies with units of kg/m². This approximation is motivated by the expectation that anomalies are largely confined to a relatively thin, laterally extensive layer associated with the plume head beneath the lithosphere (Herrick and Phillips, 1992). In the depth-constant (δρ-constant) scenario, density anomalies are instead assumed to be uniform with depth, extending from the base of the lithosphere down to a free-parameter depth, providing a volume to these anomalies.

In a first phase of the investigation, we conduct a quasi-global study using a multitaper spatio-spectral localization approach (Wieczorek and Simons, 2007) to suppress signals from highland regions predominantly supported by thickened crust - namely Ishtar Terra and Western Aphrodite Terra - from the gravity and topography observations. We explore a range of creep regimes, from dislocation or diffusion creep dominated mantles to mixed rheology scenarios, and consider the mass-sheet and δρ-constant scenarios for the density anomalies. Inversions are performed using a Bayesian framework, similar to Maia et al., 2023.

A second phase focuses on localized inversions using a single-window localization approach to isolate regions of interest that are thought to be linked to mantle dynamics, namely Atla, Beta, Phoebe, Themis and Imdr Regiones, Lada Terra, and Atalanta and Niobe Planitiae. This allows us to probe potential lateral viscosity variations in Venus's mantle. Here, dislocation creep dominated and mixed rheology cases are examined for the mass-sheet scenario, while only dislocation creep dominated rheology is considered for the δρ-constant case. 

Both phases consistently yield two best-fit viscosity structures: a one-to-two order-of-magnitude increase at relatively shallow depths, consistent with Maia et al. (2023), or a decrease of comparable magnitude deeper in the mantle. The former produces a low-viscosity zone potentially indicative of partial melting. The latter may reflect compositional stratification, such as the transition from a depleted upper mantle to an enriched lower mantle, or even local reservoirs, formed due to mixing in the mantle.

These interpretations would carry significant implications for our current understanding of Venus. A low-viscosity zone caused by partial melt would support recent observations of a volcanically active planet, while a chemically stratified mantle, separated into depleted and enriched layers, or local reservoirs associated with mantle mixing, have major implications on Venus’s planetary evolution. The localized investigation also reveals similarities between geographically adjacent regions, namely regarding the predicted viscosity jump and lithospheric thickness, suggesting that Venus's rheological parameters and viscosity structure exhibit a significant degree of regionality.

 

References:  Banerdt (1986), Journal of Geophysical Research: Solid Earth; Breuer & Moore (2015), Treatise on Geophysics; Hager & Clayton (1989), Gordon and Breach Science Publishers; Herrick & Phillips (1992), Journal of Geophysical Research: Planets; James et al. (2013), Journal of Geophysical Research: Planets; Kiefer et al. (1986), Geophysical Research Letters; Konopliv et al. (1999), Icarus; Maia et al. (2023), Geophysical Research Letters; Pauer et al. (2006), Journal of Geophysical Research: Planets; Rappaport et al. (1999), Icarus; Sjogren et al. (1980), Journal of Geophysical Research: Space