I expect climate change to be consistently worse than the mainstream predictions.
There’s a huge political incentive to err on the side of downplaying than to exaggerate, and those models should have been bad enough to make the point.
They first created a simplified model. That model was still a fairly good match with the fully featured models. They then varied the parameters around reasonable possible values. If scientists were using most likely assumptions, you would expect a 50/50 split of better/worse predictions.
In practice it was 93% worse. Scientists were (unconsciously) using a lot of best case figures, not expected case.
When I read about that, and the complete lack of follow-up coverage, I knew we were fucked.
Which I don’t blame the original scientists or studies for. Their predictions were bad enough that we should have done something then. It’s been clear for at least 30 years.
They did what they thought would have the most traction and not get them dismissed as crackpots. And, to be clear, they were still within a margin of error. They weren’t generally wrong, considering their margins. It’s just that they’re not going to be perfect and they made the better choice.
Even now it would be easier to dismiss climate change if the reality ended up better than predicted, even if it was still bad.
I came to the same conclusion after reading papers and articles for years and every year it got a little bit worse and a little bit sooner. Just following the trajectory of the changes in their expectations told me that we would hit ~+6°C by 2100.
One simplified model is not the reason for this. They’ve been expected to use a number of different climate scenarios in their work for a long time. Stop making shit up.
The point was to test for human bias in the data. The initial complaint was that scientists were cherry picking the data to make it look worse than it was. By looking at the spread, it showed they were minorly cherry picking to understate the problem. A fact that is now more and more obvious as it plays out.
It wasn’t deliberate. It was more likely scientists being overly cautious about overstating their case. The end results pushed the normal agreed values closer to best case predictions rather than middle.
This was back when the main models needed weeks of super computer time, so only a few were run, and everyone else analysed the shared results.
I expect climate change to be consistently worse than the mainstream predictions.
There’s a huge political incentive to err on the side of downplaying than to exaggerate, and those models should have been bad enough to make the point.
About a decade ago a group did a study on it.
They first created a simplified model. That model was still a fairly good match with the fully featured models. They then varied the parameters around reasonable possible values. If scientists were using most likely assumptions, you would expect a 50/50 split of better/worse predictions.
In practice it was 93% worse. Scientists were (unconsciously) using a lot of best case figures, not expected case.
When I read about that, and the complete lack of follow-up coverage, I knew we were fucked.
Which I don’t blame the original scientists or studies for. Their predictions were bad enough that we should have done something then. It’s been clear for at least 30 years.
They did what they thought would have the most traction and not get them dismissed as crackpots. And, to be clear, they were still within a margin of error. They weren’t generally wrong, considering their margins. It’s just that they’re not going to be perfect and they made the better choice.
Even now it would be easier to dismiss climate change if the reality ended up better than predicted, even if it was still bad.
I came to the same conclusion after reading papers and articles for years and every year it got a little bit worse and a little bit sooner. Just following the trajectory of the changes in their expectations told me that we would hit ~+6°C by 2100.
One simplified model is not the reason for this. They’ve been expected to use a number of different climate scenarios in their work for a long time. Stop making shit up.
The point was to test for human bias in the data. The initial complaint was that scientists were cherry picking the data to make it look worse than it was. By looking at the spread, it showed they were minorly cherry picking to understate the problem. A fact that is now more and more obvious as it plays out.
Funny you say that, my wife does climate change mitigation research. She definitely isn’t cherry picking data. Climate change is just very uncertain.
It wasn’t deliberate. It was more likely scientists being overly cautious about overstating their case. The end results pushed the normal agreed values closer to best case predictions rather than middle.
This was back when the main models needed weeks of super computer time, so only a few were run, and everyone else analysed the shared results.