I work in Construction, so variability risks for me would be performance of folks in the field, or weather conditions. These are things that we can estimate based on the past, or have a Monte Carlo done to simulate the impact.
Ambiguity risks would be things that are much more difficult to estimate or model, for instance if you're using a new technology, or even as we're seeing now during covid the volatility of market conditions. You can also simulate these, but without good assumptions the impact can vary greatly. I would typically rely on subject matter experts, and benchmarking based on what others have experienced.