Which case is missing?
The Worst Case.
This blog post looks at Climate Science and worst case scenarios. Why are worst case scenarios not routinely presented. Why do climate scientists avoid them?
The discussion around worst case scenarios has been highlighted by the recent debate of the use of the RCP 8.5 pathway. (An explainer here)
Many scientists have been critical of it’s use and presentation in literature, causing some scientists to write counter articles to dissuade others from its consideration.
The key phrase used in this counter article is
“more-realistic baselines make for better policy“
Really?
So what is the point of a worst case scenario.
I’m an engineer (I don’t design bridges – but it’s going to be my analogy of choice)
In my job I deal with worst case scenarios all the time. They are the fundamental aspect of my job. So putting them to one side is curious to me.
Imagine you are designing a road bridge. What do you design that bridge for?
The design case for the bridge might be something like, a queue of articulated lorries stacked both ways, with 6 inches of snow, on a windy day.
Not only this, but you would also be likely to consider an earthquake, even in benign regions of tectonic activity.
Maybe your earthquake occurs when there is a queue of lorries, on a snowy day…you get the idea.
But not only this, you want the bridge to last 100 years. So what will everyone be driving then? How about individuals drive around in their own personal tank. (Oh wait people do that now) How about you forecast that the payloads of those lorries are now 50% heavier, maybe you envisage double decker trailers.
You get the idea.
Importantly, as the bridge designer, I don’t care whether this scenario ever occurs. I am not interested in how likely it is. The goal is to design a safe bridge.
What about an ever so slightly topical subject like vaccine development.
When you run trials and licence a vaccine, you do so on the basis of the worst cases. if 1 in 1000 people drop dead after taking it, then that’s probably not going to be good enough.
Let’s imagine that we set our bridge building policy and our vaccine development, on “more realistic scenarios”.
Okay we carefully study bridge traffic over a period of months and determine that the mean amount of traffic on the bridge at any one time is six cars and a couple of pedestrians walking their dog. So we design the bridge for this loading.
It collapses during the first rush hour.
Now we carefully study all the side affects of the vaccine. It seems like the majority of people have no problems, so we licence away and witness a wave of death that makes the disease we are trying to protect against look benign by comparison.
So what is the rationale for dismissing worst case scenarios from climate science and subsequent policy? We want our flood defences to be high enough, we want our emissions cuts to be fast enough, don’t we?
There is some legitimacy in using mean, or more likely outcomes to plan adaptation. Partly because if you want to adapt to a +5C world, good luck!
However when it comes to emissions reduction there is no reason. The only purpose that average, more likely scenarios serve is to provide false comfort and increase risk.
They increase risk because they reduce the room for unkowns. They reduce the margin of error.
In my bridge scenario, I have room for my bridge to survive a foot of snow. This is because I assumed a queue of lorries that is unlikely. It covers for my underestimation of the snow.
In my “most likely” climate scenario I have limited room, a small margin of error, for unexpected feedback loops or unknowns not modeled. Like this one, or this one.
It seems much more like these realistic scenarios are there to pacify investors and policy makers. To avoid defeatism.
All I can say if we have leaders that give up due to defeatism, then we have the wrong ones.
Luckily we are nearing the point where climate science no longer matters and the predictions are moot, because we need to do everything we can, as fast as we can.
How did we get here? Maybe by considering “most likely, average scenarios”.