,
19 tweets,
4 min read
Read on Twitter
If your Ph.D. is from Twitter Science U, then you’ve probably tweeted something like “A single observation can falsify a theory." In some cases, this is true — if you see a black swan, then you’ve falsified the theory that “all swans are white”. 1/
Other times, contrary observations don’t do much to confidence in a theory. A good example was the observation of neutrinos going faster than light. en.wikipedia.org/wiki/Faster-th… 2/
No physicist looked at that observation and said, “special relativity must be wrong.” Rather, everyone just assumed that the scientists making the observation had made a mistake. Indeed, that was correct: there was a loose cable on their apparatus that threw off their results. 3/
But what if the investigators had not been able to find something wrong with their apparatus. Would everyone then have thrown out special relativity? No. Special relativity is a very successful theory, so it would take more than one observation to generate doubt. 4/
Let’s assume the measurement was verified by other, independent experiments. Would that “disprove” relativity? Probably not. After all, it’s a successful theory that explains a lot about the world. 5/
The first thing physicists would try to do is to create a carve out for this observations — explain why it’s generally true that nothing exceeds the speed of light, except in some limited cases. 6/
If the carve out is successful, the theory will survive. If other problems appear, perhaps the entire paradigm will shift to a new theory. But that one observation of faster-than-light neutrinos will not destroy the theory. 7/
How is this relevant to climate change? One “discrepancy” you see pointed out is that atmospheric temperature observations disagree with climate models simulations. Some people refer to this as the “missing hotspot” in the tropical upper troposphere. 8/
What this means is that the atmosphere is not warming as fast as we would expect. This, Twitter scientists say in a very serious voice, “falsifies CO2 as causing climate change". I agree that there is a disagreement here. What follows is my interpretation of it. 9/
I think the disagreement is interesting, but it’s also hard to believe it’s real — just as it was hard for physicists to believe neutrinos were going faster than light. 10/
First, the history of the atmospheric temperature measurement from satellites does not provide confidence: 11/
Second, the connection between surface and atmospheric temperatures is governed by simple thermodynamics (google Moist Adiabatic Lapse Rate). There is nothing specific to greenhouse gas warming here. 12/
If the Sun were causing the warming, or clouds, or ocean variability, we would expect the “hotspot” to appear. Note that the “fingerprint” of greenhouse warming is a cooling stratosphere, not the “hotspot”. 13/
If the observations turn out to be correct, then it requires new physics to be operating — but only on long time scales. On shorter time scales (i.e., El Nino cycles), we do indeed see the atmosphere warm as expected. No one has any ideas for what that physics could be. 14/
If such physics were discovered, would that cause us to reassess climate science? It's possible, of course, but I think it’s unlikely — climate science is a very robust science that explains almost everything we see in the world. 15/
The most likely outcome is that it would be incorporated into mainstream theory, but the main results would survive. Perhaps it would cause us to re-evaluate climate sensitivity or other parameters. 16/
But, before that happens, we must 1) validate that the atmosphere is waring less than expected on long time scales or 2) come up with a theory that explains how it could happen. Without either of those, it's just like the observation of faster-than-light neutrinos. 17/
One last point: I saw Mike Brown (@plutokiller) give a talk at recently where he said that he didn't believe observations that couldn't be reconciled with theory (I'm paraphrasing). That's exactly the point I'm making here. 18/
Another great addition to this string:
