This is the eighth and final post in a series to do with scenario analysis and the proposed SEC Climate-disclosure rule. The background to the series is explained in the first post: SEC Climate Rule: Scenario Analysis - Part 1.
We started the series of posts with an evaluation of the relevant sections of the proposed rule. We quote the wording of the rule using the page numbers in the SEC’s document (The Enhancement and Standardization of Climate-Related Disclosures for Investors). The first three posts reviewed some of the material on pages 83 to 86 of the proposed rule.
Scenario Analysis
The need for creating “imaginary gardens” or for determining the “unknown unknowns” can be seen in the following definition for scenario analysis.
Scenarios are coherent, credible stories about alternative futures. They are created around a synthesis of multiple, wide-ranging perspectives on a particular problem, rather than detailed development of a single viewpoint. Scenario planning does not forecast, predict or express preferences for the future; rather the story-telling paints internally-consistent pictures of alternative worlds, which might emerge given certain assumptions, that are credible in the light of both known and lesser known factors.
(Spratt & Dunlop, 2021)
Alternative Futures
Most climate analyses follow the first of the roads discussed above. Such analyses project a future that is a linear projection of the past and present. We see reasoning on the following lines:
Human emissions have caused the concentration of CO2 in the atmosphere to increase by 40 ppm every 10 years ever since the early 1950s.
This has had the effect of increasing atmospheric temperatures by 0.3°C every ten years.
We anticipate continued economic growth. Therefore, the CO2 concentration is expected to increase by at least 50 ppm over the course of the next ten years.
Therefore, temperatures will increase by about 0.4°C during that timespan.
The assumption of linearity built into the above sequence of events can be challenged. Potential non-linearities include the following.
Currently, about half of the heat generated by the greenhouse effect is absorbed by the oceans. There are indications that the oceans cannot continue to absorb heat at the same rate. Therefore, the atmosphere will probably warm more quickly than anticipated.
Increased temperatures may fire the “clathrate gun”.
The world may be entering a severe recession (partly caused by climate change), in which case economic expansion may go into reverse. This may be bad for human prosperity, but it would be good for the climate.
Each of the above scenarios creates an alternative future. Some of them — the ocean’s heat absorption, for example — would cause temperatures to increase more quickly than predicted. Other scenarios, such as an economic slowdown — would reduce the rate of heating.
The above scenarios are plausible — indeed, they have a high probability of taking place. Some scenarios, however, are of low probability. For example, increased volcanic activity could create a natural form of geoengineering (page 152) and cause atmospheric temperatures to fall.
Stories
The SEC approach to climate reporting and the management of risk is rightly based on the use and presentation of data. The catch is that humans do not communicate with charts, graphs and data tables. Humans communicate by telling stories to one another. As one of the commenters (staggering_god) at the reddit Collapse site said,
More and more, people are going to have stories. True stories about things that have happened to them, their families, or their friends. That's what will ultimately persuade people--not models, projections, scientific experts, or charts. Like a lot of farmers today, it will start like, "I don't say it's climate change, but the growing season is out of whack”.
Only when a large number of people are telling stories to one another about climate change are we likely to see much action.
People have to hear stories from people that they know and trust, and who have a similar cultural background. With respect to the above quotation, farmers will believe stories told by other farmers living in their community.
Mathematical Models
Climate models work best when there are linear relations between the variables in those models. Major discontinuities resulting from step changes or tipping points in one of the key variables are very difficult to model. Even if the model can handle a discontinuity, the timing of the event is never known. For these reasons, models from authoritative organizations such as the IPCC (page 37) have limited utility when it comes to predicting what may happen more than a few years from now.
Mathematical models have even more difficulty incorporating variables from other domains. For example, climate trends can be affected by variables such as the following.
Systems
As has been stressed in the previous chapters, climate change is not a stand-alone activity. It is part of a system that includes resource usage, economic factors, and human behavior. These factors are very difficult to understand on their own. When included in a climate model, they become almost impossible to comprehend. For example, as the supply of oil declines it would seem likely that emissions will go down as society switches to cleaner sources of energy. However, companies may also switch to coal and other energy sources that have higher greenhouse gas emissions for the same amount of energy delivered. Techniques that can help with understanding systems are What If Analysis and Why Tree Analysis. But these are narrative techniques. They do not contribute toward quantitative analysis.Economic trends
It is always difficult to model the world’s economy. Yet economic trends are likely to have a major impact on greenhouse gas emissions.Human behavior
It is virtually impossible to predict how individuals will react to a given situation. The behavior of large groups may be slightly easier to predict, but not by much. For example, people generally are not taking much action in response to climate change. But some sort of psychological or perceived tipping point may occur that will quite suddenly lead to large numbers of people taking action. Such behavior is impossible to model.
High Consequence Events
The discussion to do with risk analysis in Chapter 1 drew a distinction between the likelihood of an event taking place, and its consequences. Generally, there is an inverse relationship between the two. Most scenario analyses focus on high consequence /low likelihood events, i.e., those events that could radically change the parameters of the analysis, and form which recovery is difficult, i.e., tipping points.