Subdifferential calculus extends the concept of derivatives to non-differentiable functions by using subgradients, allowing for optimization in non-smooth contexts. It is particularly useful in convex analysis and optimization, where it provides tools to handle functions that are not differentiable everywhere.