Thread by @RogerPielkeJr

OK, another short, but data-rich thread on the new Kossin et al. PNAS paper on tropical cyclones, which has been widely misreported, misunderstood and which misrepresents its own findings ... (benefiting from collaboration with @RyanMaue)

Let me start by saying that the paper's conclusions, that in some ocean basins a greater proportion of TCs have become intense since 1980, is neither new nor controversial
Kossin et al. did not perform an attribution analysis but is widely claimed to have done so (which is odd)

1-KKOV20 says of new satellite dataset:
"key point here is that the ADT-HURSAT record is homogenous in time and by region, whereas the best-track data are not"

OK let's compare the datasets (those diffs👀)
Any inhomogeneity in best-track is hard to see if HURSAT is homogeneous

2-KKOV20 says of the two datasets:
"it is not always clear which of the two data records is the more accurate for any particular estimate"
OK then let's then use std dev of differences as an "error" on actual intensity estimates
Here is what their Fig.1 looks like w/ error bars

Bottom line
KKOV20 makes a great case study from start to finish
Science promotion
Hurricanes & climate (yay!)
Media hype
Gross misrepresentation
Questionable science
But that is why we publish ... to spark debate & discussion
And KKOV20 have certainly achieved that goal
/END