Update: our most powerful supercomputer, currently Titan Cray XK7, capable of 27 Petaflops, is also surrendered to run climate models now! Not surprising that we are trailing China (Tianhe-2). On the topic of the computing power: its increase cannot help climate models. Solutions for Navier–Stokes equations are unstable for all but very simple special cases.
Originally published on 10/21/2015: inspired by this press-release from Potsdam Institute for Climate Impact Research, boasting a new super-computer capable of 212 Teraflops, I’ve decided that a Teraflop is an excellent unit to measure the otherwise countless failures of the “climate scientists.” (1 Teraflop = 1 trillion flops.)
Let us define 1 flop as a failure of a certain magnitude, comparable to a fourth grade student making a mistake in multiplication tables in a classroom. Not a big deal. If a fashion model tells her friend that solar activity does not influence climate, that is between 10 and 50 flops (depending on how many cocktails she has had that night). If she does the same on TV, in front of millions of viewers, it counts as 50 MFlops (Megaflops). If somebody calling himself a scientist does the same, that failure can be estimated at 5-10 GFlops (Gigaflops). The failure value is proportional to the audience size and presumed qualifications of the author(s), and the degree of claimed certainty. Failures should not be confused with mistakes. Thus scientific failures can be measured, using reliable scientific criteria.
So, how big are the following scientific failures?
- Al Gore’s movie An Inconvenient Truth
- Each report by the IPCC, including Summaries for Policymakers (SPAM) and messaging by environmentalist NGOs – IPCC observers
- Obama’s statements on climate and weather
I have not done the math, but hundreds of Teraflops is a plausible estimate for each of them.