“But Predictions”

Various ways to attack the predictibility of climate science. As always, watch out for deflection after asking for specifics.

See also #But12Years, #ButModulz, #ButFalsification.

Examples

Exxon scientists were actually wrong with their predictions.

(Luthertarian)

Not ONE single climate doomsday prediction over the past 60 years has come true. Not one.

(NealB)

Objections and Replies

Always. Climate scientists predictions always fail—
☞ False: Zeke assessed models from 1970 to 2007. Their predictions fared quite well. If we look at 2019, it’s right in line with the old CMIP5 models. There are many other comparisons, as climate scientists tend to test their models a lot. See my collection.

Chaos. Climate is too chaotic to be sure of—
☞ Seasons still follow one another. Think of climate like a horse in an hospital: one expects things to be OK in the end, but we don’t know what’s going to happen next.

Complexity. Climate is too complex to be predicted—
☞ Without the human factor it’d be easier. Climate itself is still deterministic.

Ehrlich. Climate predictions bring to mind unenlightened futurists as Ehrlich—
☞ Not sure how your analogy works exactly. The possibility to be wrong is the hallmark of the empirical sciences.

Exxon. Exxon was wrong—
☞ Look at Andrew’s graphs. See also this article.

Handwaving. Some prediction P failed, most extreme predictions failed—
☞ That’s easy to say. Which one do you have in mind?

Impossible. Climate science cannot make prediction—
☞ Either it fails to predict, or it cannot. Make up your mind.

IPCC. The IPCC itself says that climate is chaotic and impossible to predict—
☞ The next sentence reads: Rather the focus must be upon the prediction of the probability distribution of the system’s future possible states by the generation of ensembles of model solutions.

Non-linear. An ECS of 2.1 doesn’t pose the same mortal threat as 3C—
☞ We still need a carbon-neutral economy ASAP, as the risks are only shifted by a few decades. Non-linearity cuts both ways: the field that studies non-linearities in dynamical systems is called Catastrophe theory. Like uncertainty, non-linearity is nobody’s friend.

No shit. Climate science can’t predict a shit—
To be fair, that’s not really what they are designed for.

Soon. If cycles continue as in the past, the current warm cycle should end—
☞ Don made that prediction in 2001. It did not pan out very well for him.

Useless. If a theory cannot make predictions, it’s useless—
☞ AGW is not of prediction problem, but of control problem: we need to decide on how much and how fast we’ll dump CO2 into the atmosphere.

When. When, where, and how severe—
☞ Statistical inference only provides tendencies {6}. RTFR for specific events.

Notes

{} Dutch Book. One can’t lose by saying both that predictions are necessary and impossible.

{} Luckwarm Inertia. Beauty sleeping through a few decades to realize we’re on a Titanic that is even harder to move back on track may not be the best strategy.

{} Predictability. Predictability is less clear than we usually presume.

{} Projection and Prediction. A prediction can be distinguished from a projection, but for our purpose that does not matter much.

{} Robustness. Whining about climate modulz does not cohere with appeals to economic modulz. Yet many contrarians will do both, like Adam.

{6} Dice. Two dice suffice to refute most contrarians appeal to “but predictions.”

* * *

Resources

2018-06; Hansen’s 1988 predictions – 30 year anniversary; Nick haz an online tool.

2017-12; Checkmate: how do climate science deniers’ predictions stack up?

Further Readings

2020-11; A 50-Year-Old Global Warming Forecast That Still Holds Up; the first prediction was made by Mikhail Budyko.

2019-01; Advances in weather prediction; https://doi.org/10.1126/science.aav7274

2018-08; Hothouse Earth: here’s what the science actually does – and doesn’t – say.

Climateball Episodes