I do not have that luxury, so I have to go on what i read and see.
You could (ab)use the surface finish and try different settings, and then compare it to test results. But this is sort of a black box with only discrete settings.
Another option may be what they call the "failure strength reduction factor Kf". That way you have more control over the actual reduction numbers, and you can tune your model more precisely to the test results.
In any case it will be difficult to get good results. But you will have the same trouble with other software or hand calculations. The influence of environment on fatigue is just difficult to predict.