How do you validate a traffic simulation model without field data?

I am working on a microsimulation study for an urban corridor, but we have very limited field data to calibrate against.

Reviewers often ask for validation that the available data cannot support.

How do you handle this in your own work, and what do you tell reviewers?