Thomas,

Although I'm not completely sure, we can note in the red (All Forcing) curve from SPEAR a general downward trend over the whole time period shown.  This is to first order a response to greenhouse gas warming, which in the SPEAR model, produces a downward trend TC frequency over the Atlantic basin.  The only exception to the downward trend is a temporary period of rising trend, from about 1980-2017, which coincides with a reduction in aerosol cooling in the model.  So I think that what happens is the TC frequency rises for this brief period in response to the temporary reduction of aerosol-induced cooling.  Once the reduction is basically "finished" (because you can only reduce it so far) then the greenhouse warming-induced reduction really takes over in the model.  The behavior you mention around 2017 is at the transition point where the aerosol induced change (increase in TC frequency, due to reduced aerosol forcing) is replaced by the greenhouse gas-dominated reduction signal.  Presumably in the pre-1980 period there was a gradual reduction in TC frequency over time due to the (weaker) greenhouse warming-induced reduction along with the gradual effect (also a TC frequency reduction) from the gradual build-up of aerosol loading and cooling in the pre-1970s period.