Do the models back up your approach?
Computer models to project future scenarios are constantly being refined in order to improve their skills and accuracy. As such, they can provide useful guidance in how we approach and understand challenges, and seek solutions. However, models should not be relied upon wholly to provide accurate assessments. Models are only as good as the data they are supplied with and should always be regarded as best guess rather than absolute fact. Seaweed Generation is determined to supply more information to inform models and will use them to guide further research.
Now there’s a question. The short answer is yes.
Everyone loves a model, and we are no different at Seaweed Generation. Over the years, modelling has proven to be incredibly powerful and useful to scientists and has allowed us to identify and subsequently start responding to the climate crisis. Without the models, there would be no one active in the CDR sector now, because no one would know there was a problem.
But. There is a well-known saying that all modellers worth their salt agree with: All models are wrong, but some are useful1. Which ones though? That’s a real problem to work out. Modelling on a global scale is difficult. That’s why there are so many modellers in the world, and so many (wrong) models.
CDR is a rapidly evolving field, market and opportunity. We need modellers and their models, and we need them to be as accurate as possible. However, what we don’t need is retrospective forcing of generalist inaccurate models on to specific solutions and processes without undue and necessary attention and thought. It’s not easy to model global systems, we all understand that all too well. Those of us with CDR solutions need accurate models to inform and provide insight on the impacts of our work. We need the CDR models to be useful, not wrong.
It is very easy for existing models to be adapted under the banner of CDR technology appraisal and return a rather spurious outlook23. Poorly understood components of models are given often subjective weightings, sometimes as simple as a 0 or 1 (not significant or significant). Modellers are often forced to guesstimate a number somewhere in between, and usually without sufficient data and with a broad resolution. It’s not their fault. They need the data, which they usually simply don’t have. It will take time to fill this void in data and understanding.
Unfortunately, time is something we don’t have. We have been sleepwalking into a climate crisis, but are slowly awakening and responding to the situation as it unfolds in front of us. There’s a danger that complex, inaccurate, misapplied models could hinder and slow our response. They might divert us down the wrong track or divert us off the right track. Right now, no one knows. Everyone in the newly developing CDR field is scrambling to work it out.
Our approach at Seaweed Generation is data heavy and is part of our evidence based process, so we’ll be getting the data the modellers need to improve their models as we go along. We’ll be getting our data to them as quickly as we can. Not everyone has this luxury though. We have positioned ourselves so we are exploiting an entirely natural process that has been going on for hundreds of millions of years, so we are confident that our approach has already been proven to work, and will have no unintended environmental impacts. Some would argue (not us!), that we don’t even need a model.
So, let’s not kid ourselves: models can and should be used to help inform on all CDR approaches. But right now, in the early assessment of CDR strategies, we have to be careful that the tail doesn’t wag the dog. The climate emergency is real, the alarm bells are ringing and ringing loud. We’ve already got it very, very wrong; aside from a nuclear war, there is little humanity could do to diminish, destroy and degrade our natural environment any further.
Let’s not stand rooted to the spot in indecision because of a lack of understanding of what might happen if we get it wrong in our response.
As we return to the saying “All models are wrong, but some are useful”, we need to be mindful of the broader picture. The complexity of the response required is beyond any current model, regardless of how wrong or useful it is. Favouring one solution over another today because their model is currently less wrong/more useful might very well hinder tomorrow’s solutions from being developed.
We need to use the models to guide us, they shouldn’t be directing us. It’s a subtle difference, but one that is often lost in the noise of scientific debate.