If many different methods agree

.. then they must probably be right. Here's the famous Hockeystick curve (published by Mann et al in '99). The curve is based on historical projections of temperature based on several different methods (tree rings, etc..).