Long-wavelength topography on Mercury is not from folding of the lithosphere

Previous work suggested that the lithosphere of Mercury could undergo folding in response to global contraction, and indeed observations from the MESSENGER mission revealed several regions where long-wavelength topography is present. Here, we test, via finite-element simulations that use a more realistic rheological model than that earlier work, lithospheric folding as a formation mechanism for long-wavelength topography on Mercury from interior secular cooling over the last 3.8 Gyr. This radial contraction has been estimated from geological observations to be less than 10 km, which translates into small amounts of horizontal shortening of < 0.3%. Under expected surface temperatures of ∼440 K, the development of even modest fold amplification in such low strain environments is untenable. The scenarios under which there is this positive fold amplification begin with a fully compensated crust, but amplifications are small (factors < 1.1). Under other, non-compensated scenarios (e.g., a constant thickness crust), the collapse to isostasy overwhelms any folding instability. In order to produce lithospheric fold amplitudes that match those observed on Mercury, unrealistically large amounts of horizontal shortening (in excess of 10%, corresponding to hundreds of kilometers of radius change) are required. Therefore, we find that lithospheric folding cannot produce the observed long-wavelength topography on Mercury, and conclude that this topography must be buoyantly supported.