Reference Steel Design Guide Series 1 Column Base Plates, AISC 1990.
If you only want a weld and a baseplate, you will have to check in yield strength at least the compression side and the tension side. This can be made by yield lines but will require some trials and be knowledgeable on such.
An alternative -of complicated execution is to weld circumferentially to one plate preset in the concrete; anchors and weld would be at same radius.
If acceptable to your intent, you could proceed this way.
1) would weld star arranged rectangular plate foot corbels to the column
2) at the external ends of the star arranged plates is where the bolts to anchor the column will be. You may hide the bolts within 2 angles making a box welded at such external ends of the plates. Such 2 angle boxes would be of the same height than the star arranged rectangular plate corbels. Nuts would be aparent atop the boxes.
3) preferably (if possible) make the plates' dimension in plan as short as to be possible to assume the forces be passed in shear; in any case make von mises checks. If necessary, reinforce the thickness of the cylinder if you see
4) with the above approach the thickness of the baseplate becomes almost non-consequential, since you can put as much star arranged plates as required to meet the only problem, compression of a plate at least simply supported in 3 sides, which you can judge in yield line approach.
5) you of course anchor bolts to required extent
6) the stresses in the concrete must be weak enough to not have any problem. These you can ascertain by analyzing the projected plate for flexocompression.
A more rational approach dependent on availability is simply make "solid brick" models of the thing and analyze behaviour on thicknesses till you deem it satisfactory.
I have been faced with this problem numerous times, both with base plates and with splicing two prestressed concrete members together. I developed several Excel tools that helped me and then found a ready-made, quite modestly priced, spreadsheet at the following site:
This program was originally formulated to be a welded flange connection a steel pipe stack, but I found that it could be adapted quite well to a concrete column.
You can even download a trial version to see how it works.
