By “on the surface” I mean exactly that.

My Sabkha-style model envisages damp ground surfaces over thousands of square km (not all at once, or always damp, but perhaps seasonally damp as a surface ice layer thaws and sulphate-rich geothermal waters reach the surface to evaporate away).

The layers form as annual or decadal layers of sulphate crust upon sulphate crust. This would build-up as a smooth mantle over the said thousands of square km, gleaming pale in the weak sunlight. Any topography would be smoothed over and filled-in. Occasional meteor craters might make a local hollow which could become a temporary pond or lake, but would soon fill in due to focusing of the groundwater and evaporation in the hollow.