by Roy Spencer, January 17, 2019 in GlobalWarming

I wanted to expand upon something that was mentioned in yesterday’s blog post about the recent Cheng et al. paper which was widely reported with headlines suggesting a newer estimate of the rate of ocean warming is 40% higher than old estimates from the IPCC AR5 report in 2013. I demonstrated that the new dataset was only only 11% warmer when compared to the AR5 best estimate of ocean warming during 1971-2010.

The point I want to reemphasize today is the huge range in ocean warming between the 33 models included in that study. Here’s a plot based upon data from Cheng’s website which, for the period in question (1971-2010) shows a factor of 8 range between the model with the least ocean warming and the model with the most warming, based upon linear trends fitted to the model curves:

Yearly ocean heat content (OHC) changes since 1971 in 33 models versus the recent Cheng reanalysis of XBT and Argo ocean temperature data for the surface to 2,000m layer. The vertical scale is in both ZettaJoules (10^21 Joules) and in deg. C (assuming an ocean area of 3.6 x 10^14 m^2). The Cheng et al. confidence interval has been inflated by 1.43 to account for the difference between the surface area of the Earth (Cheng et al. usage) and the actual ocean surface area.

…