How do we provide the information that society needs to make decisions on climate mitigation and adaptation, with only a limited number of scientists and experts?
A great thing about working in climate science, is that people care what you find out. The benefits that climate science might offer to society come to mind often when I’m working, which is a help when I have to use IDL1
One important trend in climate science is the expansion in science useful to inform adaptation. For many years, the fundamental research in climate science has been used to inform mitigation policy2. In effect, this has involved answering a single, big question, for a really small set of decision makers. The decision makers were the governments of the world, the question was “do we really need to stop emitting so much CO2?”, and the answer was, broadly, “yes”3.
During this research, it became apparent that the Earth has been merrily storing energy in the oceans for quite a while now, and that even dramatic cuts in CO2 emissions aren’t enough to stop the climate changing for a good number of decades. On top of that, it turns out that human society isn’t particularly well adapted to the natural variability of the climate anyway. Human systems will be obliged to adapt to a changing climate over the next few decades, whatever the greenhouse gas emissions turn out to be.
The adaptation problem is different from the mitigation problem. The timescales are much shorter; mostly months and years, perhaps a small number of decades for big infrastructure problems. There are a huge number of decision makers; from individuals, through companies, industry, local government, all the way to national governments, and supranational bodies. At the same time, there are an enormous number of ‘domains’ that are impacted by the climate; agriculture, water, urban planning, insurance, shipping, … you’re bored already, and that’s just some of the human systems. Every part of the Earth-human system is affected by climate change, and the people who work and live in that system might benefit from information on how it will change. The need for climate information is growing massively then, along with the diversification of the types of information needed for all those different decision makers and domains.
This complexity and diversity leads to some really interesting challenges. Each climate impacts and adaptation problem is essentially unique, and relies on deep expert knowledge, both in the problem domain, and the climate domain, to avoid poor decision making. When you meet with people who work in industry, or health, whatever, you realise just how deep their knowledge of their particular domain is. They often have lots of data (although less often, the tools to analyse it effectively), lots of experience in keeping the lights on, and a good gut feeling for how they can improve things. What they often don’t have is a real feel for how the climate impacts them, and how that might change in the near future. This is understandable, as often, there are many more pressing issues than climate and climate change on their mind. This is mirrored in the climate scientist’s relative lack of knowledge about the domain, and what the domain expert knows is important. The climate information problem has gone from being one-to-one, to many-to-many, as information flows in both directions.
Climate impacts problems can drive really interesting science, beneficial to both the climate scientist (how much do we know about important climate change?), and the domain expert (how do I integrate climate information into my decision making?). The danger is that a lack of interaction between the climate scientist and domain expert might lead to poor analysis and decision making. What happens when there simply aren’t enough climate scientists to go round?
All of this was brought to mind during a lecture that I heard recently from Darius Campbell, head of climate change at DEFRA. It was heartening to hear that the department’s policy on climate adaptation is to make sure that the right science is done, and then to make sure that it is available to those that need it. They don’t want to regulate, they want to inform, and let individual decision makers plan their own strategy4.
So, is the scientific community delivering the right climate science? This is probably an essay in itself, so I’ll duck the question for the moment. There is a lot of climate information that is well known within climate science, that hasn’t got to the people that might find it useful; I’ll focus on that.
A major technical challenge is to make sure that climate information gets to those people and institutions that need it, in order to “own” their personal climate risk5. They need to understand the uncertainties involved, and recognise when it is appropriate to use the climate information. They need the information presented in a form that makes sense to them, via a medium that they can use, and that is cost effective. As with the information itself, the communication channel tends to be different for each individual user.
There can be no single optimal strategy for presenting climate information to interested users, but what can we do? I’ll outline three approaches that span a range of interactivity and cost to scientists (I’m thinking from a scientist’s perspective here).
At one end, is simple data provision. Why don’t we just pump out every climate observation and model run to the web, and let the interested parties ingest them and use them as they like? While I think we should be doing this anyway (people are smart), this misses real opportunities to ‘add value’ (sorry) to the data. I’m of the firm belief that much of the really useful information about climate lies within the minds of those that know it best, rather than within any particular climate model. Providing only raw data misses opportunities to access that information, as well as ignoring scientifically interesting domain problems. As mentioned before, the risks of misunderstanding and inappropriate use of information here are huge.
In the middle, is consultancy. For each interesting domain problem, you pair a small group of climate scientists, with a small group of domain experts, share data, methods, and understanding, and work through the problem. This works really well (we’ve tried it), you get good science and understanding out at the end, and generally, the customers are happy. The problem is that it isn’t really scalable. At the moment, there simply aren’t enough climate scientists to go round – and you don’t really want to divert too many resources away from doing fundamental research, and consultancy skills are difficult to recruit in the first place. Many potential users of climate information are in countries or organisations that simply cannot bear the expense of hiring a bunch of climate consultants in the first place.
And so, the most expensive, time consuming, and ultimately, rewarding strategy is capacity building. The solution to a world with a vast number of varied domain problems and information needs, is to help people generate the information they need themselves. This will involve lots of training of scientists, but also the free sharing of code, data, and even computing power. We need to spread the knowledge and tools that we have, to those that might find them useful.
The time when a small scientific community locked itself away, and came back with ‘the answers’ to the climate problem is long gone. The future looks much more interactive.
1) Having said that, people might be surprised how unengaged many climate scientists are when it comes to climate policy – It just doesn’t come up that often in conversations, except perhaps in the abstract. The majority of climate scientists I know are driven by an interest in finding out how the Earth system works, more than any notion of doing good for society. I’m convinced that the notion of the ‘activist climate scientist’ is little more than a myth, or at least, very rare.
2) I wouldn’t say that the science has all been completely directed towards informing mitigation policy, more that the timescales that climate modellers have been interested in are multidecadal or centennial in nature.
3) For some value of “yes”. This statement is a hideous generalisation of course, but maybe we’ll have that argument another day.
4) For this to work well, you need to make sure that any scientific uncertainty is explicity included in the climate information. This was made crystal clear by Darius during the question and answer session after the lecture: if there is genuine uncertainty about what the climate might do, they need to hear about it.
5) I believe that was the phrase
Thanks to Richard Betts for commenting on an earlier version of this post.