He may be right, but such curve-fitting reasoning is an example of confirmation bias.

WSJ: Matt Ridley on Confirmation Bias and Global Warming | Mind & Matter

In fact, fitting any simple curve to the data does no better than the straight line.

FORBES: Global Warming Flatliners

Roeder's work revises predictions of lightning deaths by fitting a statistical curve to the historic data instead of averaging over time.

MSN: News flash: Lightning deaths continue to decline in US