A lot more of this has to do with the physical nature of the cable than the signal protocol, in that you get the best cable you can and then adjust your signalling method to move as much data as physics will allow.
The classic solution for 1km distances is RS422 though various alternatives exist, including all those intended for two wire telephone use that include things like adaptive speed changes based on noise. You can also consider optic fibre as they may come in lighter than copper and avoid having a 1km long antenna.
The bigger questions are about how the cable will be used. Will it be run out by a human in a space suit or from a fixed length drum by robot. Will it be pre terminated on earth for a specific length and purpose or in situ by someone in a suit (only two wires, PLEASE!). Does it need to survive multiple day/night cycles or just long enough to do a single test/experiment. Will it be deployed by being dragged out across the rocks or unrolled from a moving spool. Will it need to be re-used/retermineted in place. Does it need a tight bending radius to fit in the stowage space (most optic fibres do not like being tightly coiled).
Once you know some of the answers above and how much weight you have you can work out what sort of protocol will work on the cable you got. Shielding is great but will add a lot of weight, more cores will give you more data but add complication in any connections (two wires=one right one wrong way, 8 gives almost 100 wrong ways) . Light plastic insulation will get more capability in mass allowance but may fall apart under UV and cold. If length matters above all else you can go single conductor and earth return for maximum length for a given weight of copper.
If doing this as a basic project a good starting point is 8 core ethernet cable because it a is cheap, well understood with lots of off the shelf hardware. Probably running RS422 with some cores left over for power.
If looking deeper (and have deeper pockets) the optic fibre tethers used for under water devices would be the starting point since they are dealing with similar challenges and have some off the shelf hardware/data/research papers/failures to learn from.