Thermal Support of the Devana Chasma Rift, Venus K
Total Page:16
File Type:pdf, Size:1020Kb
Lunar and Planetary Science XLVIII (2017) 1400.pdf THERMAL SUPPORT OF THE DEVANA CHASMA RIFT, VENUS K. L. Bell1, W. S. Kiefer2, M. B. Weller2,3 1Imperial College London, South Kensington, London SW6 2BP, [email protected], 2Lunar and Planetary Institute, 3600 Bay Area Blvd, Houston, TX 77058, kief- [email protected] 3Institute for Geophysics, University of Texas, Austin, TX 78758, [email protected] Introduction: vana due the mantle plumes beneath Beta and Phoebe The Devana Chasma rift lies between Beta Regio Regio. and Phoebe Regio on Venus. Both Beta and Phoebe A five dimensional grid search algorithm has been are likely sites of upwelling mantle plumes [1-5]. It is developed to explore combinations of the flexural pa- believed that Devana Chasma, which reaches depths of rameters and Te for a given profile; 25 profiles were 3.3 km, is directly linked to the formation of Beta and tested in total. The fit between the profile and the mod- Phoebe Regio, with extension as the result of uplift [2, el is assessed using the Root Mean Square (RMS) mis- 6]. Devana Chasma is split into two distinct arms. The- fit; the best-fit values for di and α are those correspond- se segments may be a result of two different adjoining ing to the minimum misfit. An example of a best-fit rift systems, with the northern and southern sections model is illustrated in Figure 1.The eastern and west- originating from Beta and Phoebe Regio respectively ern sides of the profiles are tested independently; this [6]. method avoids the central depression of the rift, and Previous studies have used flexural modelling of 1 offers an additional constraint on the model results. or 2 topography profiles [4, 8] and models of the rela- The relationships between parameters in Eq. 1 contain tionship between gravity and topography [4, 9, 10] to information regarding the structure of the lithosphere, determine the thickness of the elastic lithospheric (Te) e.g. whether the lithosphere is broken or continuous. If at rift systems, volcanic rises, and coronae on Venus. d1=d2 the implication is that of a continuous lithosphere Te can be related to both the lithosphere’s thermal gra- and in the condition where d1≠d2 and d2=0 this indi- dient and heat flux [11]. cates that the lithosphere is broken [12]. The Magellan This work used Magellan altimetry data for Devana altimetry data has an estimated uncertainty of about 20 Chasma and a lithospheric flexure model to determine m for this latitude range on Venus [13, 14]. The uncer- how Te varies along the rift. Our objective was to as- tainty of the best-fit parameters is estimated by looking sess how heat flow varied with distance from the Beta at the range of values that fit the data with RMS misfits Regio and Phoebe Regio convective upwellings and to within 20 m of the minimum misfit. use this data to improve our knowledge of mantle In some cases, it was also found that the flexure plume dynamics on Venus. Our study looked at pro- model had difficulty in fitting large areas of noise or files spaced every ~100 km along the rift to obtain a topographic features. Along the south eastern edge of higher resolution and better understanding of the litho- the rift there is tesserae, a region of thickened crust. spheric structure. This feature creates an increase in crustal thickness at Method: the end of most of the eastern profiles. In order to mit- The flexure profiles were modelled based on the igate its affect, selected profiles have been clipped to equation for elastic flexure of a line load [12]: remove the anomalous region. Further, many of the profiles in the south eastern side were not tested be- x x x x cause the extent of deformation of the rift axis was so w(x) = d exp(− )cos + d exp(− )sin + d x + d (1) 1 α α 2 α α 3 4 severe, the model would not be able to fit the profile. Results: T is a function of the thermal structure of the litho- where w is the flexural topography induced by the lith- e sphere, therefore it should vary smoothly with distance ospheric load d and d . For the Devana rift, the load 1 2 and have similar values on the east and west sides of comes from normal faulting, which creates a central the rift. In our model T values range from 4 km to 89 rift valley by crustal thinning, which in turn leads to e km along the rift. These values change rapidly with flexural uplift of the rift flanks. d is the regional slope, 3 distance and are often significantly different on the and d is the topographic regional mean. The flexural 4 eastern and western flanks of the rift. These observa- parameter, α, is a function of T , and x is the position e tions suggest that the elastic flexure model (Equation along the profile. This specific solution to the flexure 1) only explains part of the observed topography and equation models linear loading when the load is ap- thus cannot be used to derive constraints on the litho- plied exclusively at the origin. Non-zero values of d 3 spheric thermal gradient and heat flux. This observa- and d indicate that part of the topography has a non- 4 tion was only possible because our study modelled a flexural origin. This may be a strong influence in De- Lunar and Planetary Science XLVIII (2017) 1400.pdf of Devana, they apparently do not fully penetrate the lithosphere in most cases. Alternatively, if d2=0, this indicates a broken plate, in which faulting completely penetrates the lithosphere. Our results show that this occurs in some localized regions of Devana Chasma. Figure 1: Profile 16° N, eastern flank. This graph Figure 2: Variations in d4 with latitude along the rift. shows the fit between the model parameters and the When d4 is large, it means that a substantial part of altimetry profile. The misfit between data and model the topography has a non-flexural origin. The non- is 73 m. This model clearly fits well with the profile, flexural topography decreases southward with in- and the noise within the data is small. creasing distance from Beta Regio and also decreas- es northward with distance from Phoebe Regio. This is consistent with thermal support for part of the rift large number of profiles along the rift. This further topography, which would be expected to be largest suggests that studies examining a single transect and beneath central Beta and central Phoebe, where the extrapolating those results to the rift systems as a upwelling plumes are centered. Error bars have not whole may be misleading [e.g., 4]. been included on the graph as the errors on these Figure 2 shows the relationship between d4 and po- values are much smaller than the incremental values sition along rift. Both the eastern and western flanks of the data. (red and blue lines, respectively in Fig. 2) show a high degree of correlation. The graph indicates a decreasing References: trend as the rift moves away from Beta Regio to a min- [1] Basilevsky and Head (2007), Icarus 192, 167-186. imum at ~ 10°N, corresponding to the region of rift [2] Stofan et al. (1989), Geol. Soc. America Bull. 101, arm interaction, after this point the value of d4 begins 143-156.[3] Kiefer and Hager (1991), J. Geophys.Res. to increase again. A plausible interpretation is that for 96, 20,947-20,966. [4] Barnett et al (2002), J. Ge- Devana Chasma, d4 is measuring topography that is ophys.Res.107, No E2.[5] Kiefer and Peterson, Ge- thermally supported by a hot lithosphere. The trend of ophys. Res. Lett. 30, 2002GL015762, 2003.[6] Kiefer the d4 graph indicates that the rift system lies between and Swafford (2006), J. Structural Geol. 28, 2144- two mantle plumes. Interestingly, values of d4 in the 2155. [7] Nimmo and McKenzie (1996), Earth Plan- southern region, near Phoebe Regio, are similar to et.Sci. Lett. 145, 109-123. [8] Johnson and Sandwell those near Beta Regio, in the North, despite Phoebe (1994), Geophys. J. Int. 119, 627-647. [9] Simons et al being thought to be an older plume [5]. Instinctively it (1997), Geophys. J. Int. 131, 24-44. [10] Hoogenboom might be expected that there would be less dynamic et al (2005), J. Geophys.Res. 110, E09003.[11] McNutt support as a result of the greater age. Furthermore the (1984), J.Geophys.Res. 89, 11,180-11,194.[12] Tur- lack of correlation between the values of Te on either cotte and Schubert (2002), Geodynamics, 2nd ed., side of the rift could imply that the dynamic support Cambridge University Press, New York.[13]Rappaport beneath the rift is asymmetric. et al., (1999), Icarus139, 19-31.[14] Ford and Petten- In the analytic flexure model for a line load [12], gill, J. Geophys.Res. 97, 103-114. d1=d2 occurs when the elastic lithosphere is continuous beneath the rift, which is usually the case in this study. Although many faults are visible in Magellan images .