Note: This Premium Climate Web Roadmap expands significantly upon the “Lite” version of the Roadmap, available here, and is provided as a free example of a Premium Roadmap. To access our full collection of Roadmaps, we have a Roadmaps Climate Site. If you’re new here, we’d suggest right clicking on the “New to the Climate Web?” link at right to get oriented, given that it’s the closest thing to a collective climate intelligence.
You might think that everyone, including business decision-makers, has been talking about climate risks non-stop for the last couple of years. That may be, but if climate risks are being significantly under-estimated despite all of that talk, the steps being with undertaken with the goal of managing those risks may not only be too little, too late, but may be fundamentally misguided.
There is a lot of evidence to suggest that from the Intergovernmental Panel on Climate Change on down, many if not most climate risk assessments today are downplaying or entirely missing potentially important sources of climate risk. That’s really not surprising given the characteristics of climate change, in particular how fast relevant variables are shifting, as well as human nature when it comes to perceiving and responding such risks. Some of the key variables likely to result in under-estimating climate risks include:
- Climate Change Probability Distributions Are Long-Tailed/Fat-Tailed
- Climate Change Probability Distributions Are Actively Shifting
- Climate Change Itself May be Accelerating
- Tipping Points Are Largely Absent from Climate Risk Assessments
- Systemic Climate Risks Are Under-Valued
- Climate Models Are Commonly Risk-Neutral as Opposed to Risk-Adverse
- We Mistakenly Pay More Attention to What We Can Measure Than What We Can’t
Based on these variables, it is quite possible that in many cases societal and business climate risk assessments are missing a substantial fraction of, if not the majority of, climate risk. In making this statement we are not suggesting we are better able to predict the future than anyone else, and we’re not saying that all of the outcomes discussed here will come to pass in conventionally business-relevant time frames, or should be considered material in business decision-making. But we are suggesting that if these and other variables are being downplayed in or are missing altogether from risk assessments, A LOT of climate risk might be unaccounted for. And that’s a problem.
Our Two Introductory Videos
A short (6.5 min) topical overview video
click on image above or available here.
Interested in a Briefing on Under-Estimated Climate Risk? Contact Us!
A short (8 min) video introducing you to some of the Climate Web’s most relevant topical resources for digging into climate risks.
Click on image above or available here.
The Roadmap
For purposes of this Topical Roadmap we’re going to use the following definition of risk:
Risk = Probability x Consequence
Risk is a fascinating topic, and we’ve assembled a GIF of quotes from risk experts that can help set the stage for thinking about under-estimating climate risks. Note the focus on the less probable over the more probable in assessing risk.
The next GIF starts off with former Secretary of Defense Donald Rumsfeld’ss famous quote, and then gets into a series of quotes having to do with the fact that, for better or worse, we face real psychological challenges when it comes to grappling with known unknowns, much less unknown unknowns!
Long-Tail and Fat-Tail Probability Distributions
The relevant definition of risk for our purposes is “Risk = Consequence x Probability.” We’ve visualized a climate consequences probability distribution below. Don’t get too hung up on the precise shape of the probability distribution; what’s important is that the distribution can be split into two parts: expected risks (higher probability, lower consequence), and fat-tail or long-tail risks (lower probability, higher consequence) outcomes. Note that this is NOT a climate risk probability distribution, since risks (probability x consequence) will tend to rise rapidly as you move farther out on the consequences axis toward “black swan” events.
The problem in under-estimating climate risks often comes from focusing primarily on the “expected risk” part of the probability distribution, and much less so on the “grey rhino” and “black swan” parts of the distribution. As you can see from the slide shown here, there are many grey-rhino variables that could be missed by focusing on “expected risks.” And that’s without even trying to identify some of the black swan risks.
There are a number of reports in particular that to a good job of exploring long-tailed risk issues. Through the links below you can access the reports themselves, as well as a small number of key extracted ideas and graphics. These are by no means the only particularly relevant materials in the Climate Web, and we welcome suggestions for others we should highlight via this roadmap.
Each of the GIFs below pull together selected extracts from a key report, linked to below the GIF. All the extracts for each report (and in some cases quite a few more) are available through the link just below the GIF if you want to explore them more systematically. Obviously the literature relevant to under-estimating climate risk is far larger than reflected in these GIFs, and we’ll continue to add materials to this page.
2018 Spratt_What Lies Beneath - the Understatement of Existential Climate Risk
2019 Spratt_Existential climate-related security risk A scenario approach
Other very relevant reports and papers include:
- 2020 Spratt_Climate Reality Check 2020 PPT
- 2021 Ehrlich_Ghastly Future A Survival Revolution in Response.docx
- 2021 Bradshaw_Underestimating the Challenges of Avoiding a Ghastly Future
- 2015 King_Climate Change a Risk Assessment
In this Topical Roadmap now we’ll dig deeper into relevant Climate Web resources through the story and links organized below.
There are a bunch of psychological and other reasons we tend to focus on “expected risk,” and some key parts of the relevant literature are listed below (with links into the Climate Web), with some extracted thoughts below that.
While we most often think of “normal” probability distributions, climate change isn’t an example of that. The potential for significantly under-estimating future temperatures, for example, is much greater than the potential for over-estimating future temperatures. This is partially a function of uncertainty about the “sensitivity” of the climate to rising levels of greenhouse gas (GHG) emissions. 2011 Probability distributions of climate sensitivity
Insight Page. As a result of how we model climate change outcomes it is very easy to fall into the trap of focusing much more on “expected outcomes” that we are most confident in as opposed to the long tail of the probability distribution The Forecasting Challenge
Insight Page. Expected outcomes are a poor proxy for climate risk. Expected Change vs. Risk.
- Why climate change is fat-tailed The fat-tail of climate risk
- Why climate change is fat-tailed 2015/9 The 'Fat Tail' of Climate Change Risk | Michael E. Mann
- With the return periods for extreme events changing, tail risk rises substantially. 2010 Extreme event "tail risks" are substantial
If you’re not familiar with “black swans” or “grey rhinos” in the context of risks and risk management, take a look at their Index Entries
Black swans, commonly characterized as “unknown unknowns,” can be hard to get your head around. But you can see how the National Intelligence Council (which has done some of the most interesting scenarios work over time) thought about potential black swans (not necessarily climate related) in its 2012 report looking out to 2030. Potential Black Swans
The reality is that the long tail of the distribution of climate change outcomes includes average global temperature change of 3, 4, or even more than 5o C, as well as many other kinds of impacts with potentially very high and negative consequences.
There is not a huge amount of discussion in the climate change literature regarding long-tailed and fat-tailed climate risk, but there is some. It’s organized in the Climate Web under these headings:
Index: I:FatandLongTailClimateRisk
The most on-point topical headings:
That said, there are quite a few other topical headings that will allow you to dig deeper into important concepts going into fat-tail and long-tail risk.
- S - Worst Case Climate Change which you might think of as getting into black swan territory. It’s not clear how useful “worst case” thinking really is, since climate change becomes “unacceptable” long before we get to anything like worst case, but it is a commonly used term. You can also explore news and opinion for the topic here N - Worst Case Climate Change and materials we’ve extracted from the literature here E - Worst Case Climate Change
- S - Probabilistic Decision-Making which pulls together the “how and why” literature for thinking more probabilistically about climate change (and risks generally). In a similar vein, S - Historically Low Probability Outcomes as Business Risk weaves the need to think about low-probability events into business decision-making.
- S - Climate Uncertainties Unknowns explores the literature focusing on the fact that while we know a great deal about climate change, there’s a great deal we don’t know (Donald Rumsfeld’s known unknowns). And we’re probably only scratching the surface of what the unknown unknowns might be!
- S - Climate Sensitivity pulls together the literature around this absolutely critical variable. We have a pretty good idea of what the sensitivity of the atmosphere to greenhouse gas (GHG) concentrations has been in the past, but how relevant is that history given that the atmosphere today is being “forced” by far more rapidly changing GHG concentrations than has ever occurred in the past. N - Climate Sensitivity does the same for news and opinion relating to climate sensitivity (of which there isn’t that much), and E - Climate Sensitivity organizes extracted materials. Some particularly interesting examples include (note that extracted graphics are dated, extracted ideas are not. Also note that in almost all cases the original source of extracted materials is present can accessed through a “parent thought” linked above the flagged thought.)
- 2008 How much sensitivity matters
- 2010 Linking temperature to cumulative emissions
- 2010 Modeled sensitivity per 1000 GTs emissions
- 2011 Probability distributions of climate sensitivity
- 2011 Uncertainty in climate sensitivity
- 2014 Implications of climate sensitivity
- 2014 Where Climate Sensitivity Estimates Come From
- A wide range of climate sensitivity estimates
- Climate feedbacks like methane releases could be considered part of climate sensitivity
- Climate sensitivity may be higher than we think
- Exactly how sensitive is the climate to atmospheric change?
- For a 1.5o C target you have to assume a lower carbon sensitivity
- Taylor: We just don't know a lot of key things about climate change. It all depends on the planetary sensitivity
Changing Probability Distributions
The Roadmap so far on the fact that we have been neglecting a lot of if not most of the overall climate risk by focusing on the “expected” portion of the climate change probability distribution. Compounding that problem, however, is the reality that the climate change probability distribution is also shifting due to climate change.
Insight Page. First a general introduction to probability distributions. Probability Distributions
Insight Page. Obviously a shifting probability distribution when it comes to climate outcomes will have a lot of implications for climate risk. Conceptually it’s easy to see that a shift to the right will tend to result in more heat extremes, and fewer cold extremes. Risk and Changing Probabilities
And there is a lot of data to show that this is already happening, including dozens of climate change fingerprint variables. It’s important to note that global conclusions don’t necessarily apply all over the world at the same time.
- The ratio of high to low temperature extremes 2012 Ratio of high to low records in the U.S. 1950's to 2000's
- 2011 Fraction of surface area that's cold or hot shifting significantly
- 2011 Anomalies are clearly trending towards hot and very hot temperatures
- As an example of geographical differences, 2011 The U.S. anomalies are not clear
We are also seeing individual events that should be so unlikely as to never be observed (if the climate were stable). Here’s an example of a specific event The 2003 European heat wave against the normal distribution that arguably proves that probability distributions are changing. Another recent examples estimates how unusual a recent heat wave in Switzerland really was Suggesting that Switzerland's climate has to have changed
Another way to think about changing probability distributions is through return-periods for extreme events. You can explore numerous examples of changing return periods from many sources in the Climate Web through this Extracted Materials head, E - Changing Return Periods
It may not be intuitively clear how significantly the economics of extreme events can change with seemingly small changes in probabilities, so we’ve addressed this topic via:
Accelerating Climate Change and Tipping Points
Some of these variables aren’t even factored into most climate models upon with climate risk assessments are based. Climate change tipping points, for example, are generally not factored into climate models because scientists don’t know enough about their “when” and “how.” But the likelihood of triggering such tipping points is increasing as climate change progresses, and will increase further if climate change is in fact accelerating.
Further compounding the problem of under-estimated climate risks is the apparent acceleration of climate change itself.
There is a lot of news coverage of this topic N - Accelerating/Worsening Climate Change but there are also recent studies suggesting that climate forcing could accelerate quickly in the coming decade, in part due to the combination of human activities and natural forcings.
One aspect of accelerating climate change that deserves particular attention is the relevance of accelerating climate change to potential climate change tipping points. S - Climate Change Tipping Points
2020 Cascading potential climate tipping points Tipping points are clearly linked to global temperatures, and even successful achievement of the Paris Accords might not avoid some of the key tipping points.
- 2017 Temperature vs. the likelihood of tipping points
- 2017 Some of these tipping points could occur even with a successful Paris Agreement
We’ll be adding more information to this Roadmap, bringing in relevant videos and other materials, e.g. to address changing extent and thickness of Arctic ice.
Systemic Climate Risk
Another commonly under-valued risk variables is “systemic climate risk.” Systemic risk used to be a term of art in the financial sector, but it is increasingly being recognized that climate change could trigger systemic risk events in its own right, from global food shortages to pandemics and even global conflict. Systemic climate risks are key because they can’t be managed through conventional business risk management tools, e.g. diversification and resilience, and are arguably becoming the risk elephant in the room.
Systemic risk is relatively new to the climate change conversation, although not to the financial sector generally. Why the idea of systemic risk should be applied to climate change To the extent that systemic risks are under-weighted or ignored in societal and business decision-making, it’s a significant source of under-estimated risk. And because the systemic risk conversation is so new, it’s safe to say that it is not well factored into risk assessments and decision-making.
Insights Page. Accounting for Systemic Risk
Arguably Lloyd’s of London’s Global Food Shock report of 2015 kicked off the systemic risk conversation, but it has advanced substantially since then. One of the things that make systemic risks so significant is that they can’t be conventionally insured or diversified against. Systemic shocks are by definition non-insurable and non-diversifiable , and it could be that responding to systemic risk requires a fundamental change in how business thinks about climate risk A Business Turning Point? In part because of the dilemma of Incremental vs. Transformational Change (Insights Page)
You can easily access a core library of Systemic Risk Sources in the Climate Web, as well as the extensive systemic risk conversation relating the financial sector. Systemic Risk News
Risk Adversity
The variables introduces previously in this Topical Roadmap focus on the actual consequences associated with the long- or fat-tail of the risk distribution. This variable focuses on the psychology of how we perceive climate risks, and the fact that we tend to underweight risky futures in our decision-making.
First, it’s important to recognize just how much uncertainty exists with respect to future climate change outcomes, and that most of that uncertainty is associated with the “more than expected” side of the probability distribution.
You can explore the topics of known unknowns and unknowns in the Climate Web, including through this collection of sources S - Climate Uncertainties Unknowns as well as a couple Insights Pages that pull together relevant thinking. Climate Known Unknowns Climate Unknown Unknowns
The consequences of climate change are potentially so severe Why a "risk averse response?" that the absence of a substantial risk response on the part of societal decision-makers could be seen as surprising. But a lot of work has gone into understanding the reasons Why AREN'T we responding in a risk averse manner?
It doesn’t help that we We are over-confident in predicting the future, which is one of the things that is contributing to our under-estimation of climate risks. Nor does it help that we tend to Discount the Future so severely.
One of the best examples of our failure to think risk-adversely involves sea level rise. How should you interpret prevailing seal level rise forecasts? Interpreting Sea Level Rise Risk
What would a risk-adverse response to climate change look like? That’s an enormous topic in its own right, but we can explore one slice of the question of risk-neutral vs. risk-adverse framing via the Social Cost of Carbon. A Risk Averse SCC
To Rethink Climate Risks
The Climate Web is structured to support climate risk assessment and re-assessment in a number of ways:
Climate Assumptions Audits help you identify and challenge assumptions that may characterize organizational thinking around climate topics.
Business Premortems are a story-telling technique that can help open up decision-makers’ minds to futures that they may not have thought about.
Scenario Planning for climate change is critical, and should often encompass a significantly wider range of outcomes that a lot of scenario planning does today. The Climate Web is specifically structured to support sector and company-appropriate scenario planning.
You can utilize the Climate Web on your own, or the Climatographers can help you utilize it Web via customized briefings and Roadmaps, and even Ebooks based on Climate Web content. We can help you take instant advantage of thousands of hours of analysis and knowledge curation across not only the topics covered in this Roadmap, but hundreds of others.
