Gravity Observations of Structure in Valles Marineris, Mars

Introduction Valles Marineris is one of the most prominent tectonic structures on Mars. It is approximately 4000 km long [1] and up to 11 km deep [2]. Valles Marineris is oriented radial to the Tharsis volcanic province and formed primarily by normal faulting in an extensional rifting environment [3]. Detailed geologic mapping indicates that it formed primarily during the Early Hesperian [4], corresponding to deformation stages 2 and 3 in the deformation stratigraphy of Anderson et al. [5]. Gravity and topography observations provide clues that can help to constrain the mechanisms which produced Valles Marineris. Previous studies by McGovern et al. [6, 7] and by McKenzie et al. [8] used admittance modeling to constrain the compensation state of the region. McGovern et al. favored generally large elastic lithosphere thickness (TE = 60-120 km) and either a very low surface density (around 2000 kg m) or the presence of significant buried loads. McKenzie et al. favored a somewhat smaller elastic thickness, around 50 km, and a surface density of 2350 kg m. The nearby Tharsis volcanic province has a basaltic composition corresponding to a higher crustal density (> 2800 kg m), and layering in the Valles Marineris walls has been interpreted in terms of flood volcanism [9]. The very low surface densities inferred in some of these models pose significant challenges to our understanding of Valles Marineris, as they require either a very different crustal composition from Tharsis or a high porosity maintained throughout up to 11 km of exposed stratigraphy. Method The admittance spectrum method used in past studies of Valles Marineris [6-8] is an excellent approach for defining the compensation state of a region. However, purely spectral models can not be used to define the amplitude and spatial distribution of possible buried loads or to assess the geologic causes of such loads. For this reason, my modeling approach [10, 11] uses a combination of both spectral domain calculations and spatial domain interpretation. I calculate the residual gravity anomaly after removing both the effects of the topography and of the best-fitting flexural compensation model, RLM = GLM FLHLM. Here GLM and HLM are spherical harmonic expansions of the free-air gravity anomaly (model JGM95I, [12]) and topography [13] and FL is a flexural response function calculated using thin-shell elastic flexure theory [14]. The residual anomaly, RLM, is the part of the observed gravity anomaly that can not be explained by the topography and its compensating root. Thus, the residual anomaly represents either subsurface mass anomalies or regions where the surface density or compensation mechanism differs significantly from the assumed regional values. By mapping RLM into the spatial domain, I can compare the residual anomaly with the known surface geology and make plausible inferences about the mass anomalies that produce the gravity signature. Figure 1 shows an example of this approach, calculated assuming an elastic thickness of 60 km, a mantle density of 3400 kg m, a crust density of 2900 kg m, a mean crustal thickness of 50 km, Young’s modulus of 10 Pa, and Poisson’s ratio of 0.25. Current gravity models for Mars have statistically useful signal up to about harmonic degree 72 [13], but terms above degree 60 are damped by constraints imposed in the inversion process [12]. Thus, Figure 1 conservatively includes only gravity and topography up to spherical harmonic degree 60, corresponding to a horizontal half-wavelength resolution of 180 km. Figure 1 shows that significant residual anomalies exist in several parts of Valles Marineris after removing the contribution of topography and its compensation. In this initial survey, I focus on three aspects of Valles Marineris.