Vigorous convection as the explanation for Pluto’s polygonal terrain

Abstract

Pluto's surface is surprisingly young and geologically active. One of its youngest terrains is the near-equatorial region informally named Sputnik Planum, which is a topographic basin filled by nitrogen (N2) ice mixed with minor amounts of CH4 and CO ices. Nearly the entire surface of the region is divided into irregular polygons about 20-30 kilometres in diameter, whose centres rise tens of metres above their sides. The edges of this region exhibit bulk flow features without polygons. Both thermal contraction and convection have been proposed to explain this terrain, but polygons formed from thermal contraction (analogous to ice-wedges or mud-crack networks) of N2 are inconsistent with the observations on Pluto of non-brittle deformation within the N2-ice sheet. Here we report a parameterized convection model to compute the Rayleigh number of the N2 ice and show that it is vigorously convecting, making Rayleigh-Bénard convection the most likely explanation for these polygons. The diameter of Sputnik Planum's polygons and the dimensions of the 'floating mountains' (the hills of of water ice along the edges of the polygons) suggest that its N2 ice is about ten kilometres thick. The estimated convection velocity of 1.5 centimetres a year indicates a surface age of only around a million years.

Formation of the Sputnik Planum basin and the thickness of Pluto's subsurface ocean

“…For a R 0 = 300 km, typical of our models, reducing the depth of SP by 0.5 km can increase the central mass anomaly by 1.6 × 10 17 -1.7 × 10 17 kg for ρ ocean from 1100 to 1200 kg/m 3 . For R 0 = 300 km, typical of our models, increasing the thickness of the N 2 layer by 1 km would reduce the central mass anomaly by 2 × 10 16 -4.8 × 10 16 kg for ρ ocean from 1100 to 1200 kg/m 3 indicating that our results ( Figure 4) are relatively insensitive to the assumed thickness of the N 2 ice layer within the reported range of 3-10 km thick [McKinnon et al, 2016;Trowbridge et al, 2016]. Thus, increasing the thickness of nitrogen ice, d N2 , reduces the average thickness of the ocean layer, t ocean , by the same thickness (see equation (2)).…”

“…The convective cells in SP, however, suggest that depth of the basin floor underlying the N 2 ice deposits gradually decreases with distance from the basin center [McKinnon et al, 2016;Trowbridge et al, 2016]. Without modeling the postimpact evolution of basin structure (e.g., loading by N 2 and flexural response), we cannot use iSALE models to constrain d topo and d N2 .…”

Thermal convection as a possible mechanism for the origin of polygonal structures on Pluto's surface

“…Yamashita et al [] studied the rheology of nitrogen ice for a temperature ranging from 5 K to 77 K and for a strain rate varying between 10 −4 s −1 and 10 −2 s −1 . Using these experimental results, Trowbridge et al [] inferred that the viscosity variations caused by temperature variations across Sputnik Planitia would be about 1 order of magnitude, which can be considered as isoviscous. Alternatively, McKinnon et al [] suggested that the viscosity contrast across Sputnik Planitia is between ∼150 and 2 × 10 5 .…”

“…Alternatively, McKinnon et al [] suggested that the viscosity contrast across Sputnik Planitia is between ∼150 and 2 × 10 5 . The large difference between these two studies is caused first by the different Δ T considered, 11 K for Trowbridge et al [] and 27 K for McKinnon et al [], and second by the fact that McKinnon et al [] also considered the effect of the grain size and the activation energy for volume diffusion [ Estève and Sullivan , ; Eluszkiewicz , ]. Overall, it is important to keep in mind that the uncertainties on the determination of the viscosity and its corresponding temperature variations are very large and should be taken with great care.…”

“…The centers of the polygonal cells are typically raised by ∼50 m compared to their edges. These structures have been identified as a surface expression of thermal convection within the glacier [ Moore et al , ; McKinnon et al , ; Trowbridge et al , ]. Following this hypothesis, the rift separating the polygons would correspond to dynamic topography induced by downwellings.…”