It's been a great challenge to project tropical cyclone (TC) activity, particularly in terms of TC frequency, in the coming warm climate. Different climate models do not agree with each other, even on the sign of projection. Part of the reasons is that these climate models are not well constrained by a short instrumental record of TCs that dates back to the middle of the 19th century. It is therefore crucial to extend TC records back over a longer period, e.g., the last millennium, using paleoclimate reconstruction methods.