Uncertainty in climate modelling: Why Should we trust climate model predictions for future climate change?

During the role-plays, I got the opportunity to examine the scientific justifications critically. Climate model critics frequently claim that climate models are unreliable and inadequate at reflecting future projections, so why should we believe them? In this blog article, I tried to delve deeper into this criticism.

Climate models represent the climate system as mathematical computer codes which run on supercomputers, their fundamentals are based on physical laws, such as conservation of mass, energy, and momentum, along with a variety of observations. Critics claim that climate models are meaningless because long-term projections cannot be accurately evaluated and little can be learned from a prediction without verification. In fact, there is no clear, well-established comparison to learn from for the climate change problem, which is peculiar in the past. However, estimates from climate models nonetheless offer useful information, despite the claims mentioned above. Climate models’ reliability is supported, for large scales and specific variables, by their ability to replicate the current climate, the most recent trends as well as those from the distant past, the fact that they are based on physical principles, and the fact that we can comprehend and interpret many of the results from well-known processes.

The boundary and initial condition uncertainty, our insufficient theoretical grasp of the system, parameter uncertainty, and the flaws in our models all contribute to the uncertainty in model projections. The spatial and temporal scales and variables also affect the model’s uncertainty. Although many parts of the climate system are chaotic, it is believed that they are predictable on decadal or longer time durations and at broad regional dimensions. The issue is that a model’s lifespan is substantially shorter than the amount of time necessary for observations to evaluate a prediction. Therefore, we need to assume the equations, parametrizations, and assumptions built into the model that can be extrapolated beyond the range in which they are evaluated. The subject of interest determines whether a model is “excellent” or “poor” because climate models make projections in a wide range of distinct quantities. Since forecasts can be checked daily, skill is relatively readily defined for numerical weather prediction, but it is more challenging to define a specific overall figure of merit, metric, or skill score for long-term climate projections.

Let’s do a thought experiment: Would the world be any different if we could predict the future with absolute confidence using a perfect model? Would we be able to combat the climate change issue more successfully? For local adaptation, accurate information on anticipated trends is essential, and it is true that climate model projection uncertainties are a problem. However, it is unlikely that they will be the stumbling block keeping us from taking action on the climate change issue rather than just talking about it. It should be underlined that each model has constraints, presumptions, and inaccuracies; for instance, models are frequently used in the design of buildings and aircraft; since we are willing to accept the possibility that these models may be inaccurate, why not climate models?

Let us further understand this scenario by a simple example, suppose you need to take a train to travel tomorrow but you don’t have the internet to check the exact timings and you are in a foreign land where you don’t speak the same language. In that situation, you would strive to arrive at the station as early as possible to lessen the likelihood that you would miss the train. You could wind up having to wait at the station for a long, but you have nonetheless decreased the likelihood that you would miss the train.

This situation is similar to climate science. Think of the worldwide agreement among the nations to keep global warming to less than 2 degrees Celsius as the train you need to catch. What steps then should we take to minimize the risks? Uncertainty about the departure time of the train translates into more investments (in time) to reduce the chance of missing it. Faster actions are necessary to increase the likelihood of reaching the climate target, such as by lowering carbon dioxide emissions. Therefore, while there are uncertainties, this in no way justifies inaction. We could wait for better information, but much like in the training example, if we wait until we hear the train approaching the tracks, it might be too late.

In conclusion, it can be said that the physical foundation and capability of the climate model to represent past and present climate change give the model confidence. As of now, these capabilities have proven to be incredibly crucial tools in understanding the climate, and there is a great deal of confidence that these capabilities can provide credible quantitative estimates of future climate change, particularly at larger scales. Nevertheless, they have continuously shown a strong and clear picture of a major climate warming in response to rising greenhouse gas levels throughout the course of several decades of model development.

References:

1. 1. Knutti, Reto. “Should We Believe Model Predictions of Future Climate Change?” Philosophical Transactions: Mathematical, Physical and Engineering Sciences 366, no. 1885 (2008): 4647–64. http://www.jstor.org/stable/25197431.
   2. Bony, S. et al 2006 How well do we understand and evaluate climate change feedback processes? J. Clim. 19, 3445-3482. (DOI:10.1175/JCLI3819.1)
   3. Forest, C. E., Stone, P. H., Sokolov, A. P., Allen, M. R. k Webster, M. D. 2002 Quantifying uncertainties in climate system properties with the use of recent climate observations. Science 295, 113-117. (DOI: 10.1126/science. 1064419).
   4. Claussen, M. et al 2002 Earth system models of intermediate complexity: closing the gap in the spectrum of climate system models. Clim. Dyn. 18, 579-586. (doi:10.1007/s00382-001-0200-l
   5. Why uncertainty is not an excuse for inaction on climate change. https://drawdown.psu.edu/research/blog/why-uncertainty-not-excuse-inactionclimate-change.
   6. IPCC AR4, FAQ 8.1: How Reliable Are the Models Used to Make Projections of Future Climate Change? https://www.ipcc.ch/publications_and_data/ar4/wg1/en/faq-8-1.html