Calculate scrum team's year's predictability

First, you need to define what "predictability" means. Until there is a clear definition, you can't know what to measure to determine how predictable the team is.

Within the Scrum framework, I would define predictability as the team's ability to meet their Sprint Goals. At every Sprint Planning event, the team sets a singular goal for the upcoming Sprint that captures why the Sprint is valuable. Assuming that each Sprint Goal is truly valuable to one or more stakeholders, knowing how often the team can commit to a goal and then achieve it is helpful. However, it relies on the Sprint Goal being valuable and the team not being too aggressive in descoping with the Sprint Goal.

Another measure of predictability could be deployment frequency and stating how often the Scrum Team makes a new Product Increment available for stakeholder use. Making Increments available is the first step in a feedback loop. However, just making the Increment available doesn't say how quickly the team responds to the product feedback or the quality of each Increment.

Building on that idea, lead time and cycle time for changes could measure predictability. If the team can respond and deploy changes to get feedback and then quickly react to that feedback, they demonstrate that they are responding to changes in the stakeholder environment.

Unfortunately, in my experience, many organizations consider the amount of work done as a measure of predictability. Rather than commitments to goals, there is an expectation of a commitment to a body of work without respect for the value of that body of work. This often fails to account for the team's ability to respond to changes as the body of work changes. I wouldn't consider this a good, reliable indicator of the predictability of a team, even though it's what many stakeholders expect.

Perhaps there isn't just one measure of predictability, and you need to consider one or more options.

Situation is : top management needs to know the predictability of a team for a year, anyone has ever done it? please let me know.

You look for the hard evidence of Done increments being produced every Sprint and immediately put to use.

If the ability to do so escapes the team for some organizational reason, ask management what they are going to do to resolve it.

I agree with the comments made my @Thomas, it is important to define what predictability means. If it is just how many features the teams are able to deliver then I don't think it gives the entire picture (what about value delivery , customer outcomes etc.). As @Ian mentions, the hard evidence available is the done increments (have these helped customer outcomes).