I think if you’ve looked at numerical models then you have enough knowledge to at least judge the underlying assumptions: boundary conditions, numerical stability, discretisation methods, validation methods and results.
In the same way that I can verify a SAT solution, I don’t need to know how to code a sat solver.
Unfortunately there are fundamental disagreements about most of the critical parts of climate science, so going by the opinion of the field isn’t foolproof (who to choose? How to choose?). Many fields have had false consensus beliefs before, and most of their problems weren’t 1000th the difficulty of climate modelling.
In the same way that I can verify a SAT solution, I don’t need to know how to code a sat solver.
Unfortunately there are fundamental disagreements about most of the critical parts of climate science, so going by the opinion of the field isn’t foolproof (who to choose? How to choose?). Many fields have had false consensus beliefs before, and most of their problems weren’t 1000th the difficulty of climate modelling.