The Rightful Place of Science

by Roger Pielke

  The campaign to have me ousted as a writer at FiveThirtyEight was nonetheless successful. After subsequently publishing a few of my pieces on sports, Silver eventually refused to publish anything further and we parted ways soon after. I harbor no hard feelings, as Silver was put under an impressive amount of pressure in a setting where being popular seems to be more important than being right. I doubt that Silver will go near the climate and disasters issue again. If you can’t take the heat, it is best to get out of the kitchen.

  Recalling Harry Truman’s sage advice, I am interpreting the furious attacks against me as one reason to prepare this primer on disasters and climate change. Here I’m also throwing in my lot with John Kay, a columnist for the Financial Times, who explained in another context about campaigners who try to create their own reality: “Whatever initial misconceptions spin doctors may promote, reality will out.”[35]

  2

  The Scientific Question Addressed Here

  This short volume addresses a very specific scientific question—a question that can be addressed empirically, that is, with data:

  Have disasters become more costly because of human-caused climate change?

  There are many other questions which could be asked about climate change, many of which are arguably far more important, but are not addressed here in any significant depth. I have addressed some of these other questions in other venues.[36] However, the central question addressed here stands on its own. It is not a stalking horse for anything else.

  Over the years, I have addressed in my work many other questions related to climate change science and policy. Here in a capsule format are some of the other important questions, and my views on them.

  Is climate change real?

  Yes.

  Does climate change have human causes, notably from the emission of greenhouse gases?

  Yes.

  Does human-caused climate change pose risks, perhaps significant ones, for life on Earth?

  Yes.

  Should policy makers around the world take action to reduce emissions toward eventual stabilization of greenhouse gas concentrations?

  Yes.

  Does a price on carbon make sense?

  Yes.

  Do scientific projections suggest that some extreme events may become more common or intense?

  Yes.

  Does current science suggest that episodes of extreme heat and intense rainfall may be increasing in some areas as a consequence of increasing concentrations of greenhouse gases in the atmosphere?

  Yes.

  Does any of the work summarized in this short volume counter my answers to any of the above questions?

  No.

  For those who may be interested in my more fully developed perspectives on these questions, I recommend my book, The Climate Fix (Basic Books, 2011), which discusses them in detail. I now return to the main question that is the focus of this volume.

  Looking for Signals in Complex Data

  Studying the relationship between climate change and disasters is challenging because there are many moving parts that contribute to the outcomes that we care about, like property damage or casualties. For instance, the frequency and intensity of extreme events can change and vary over time, but so too does the exposure and vulnerability of human settlements which are subject to experiencing disasters. Much of the work that I have been involved in over the past few decades involves efforts to separate out the role of physical factors (the “natural” in “natural disasters”) from societal factors in long-term trends in disaster losses.

  One way that researchers make this separation is by adjusting historical loss data to account for relevant societal changes. We seek to standardize losses to a common base year—a process which we have called “normalization.”[37] If this procedure is done properly, then trends in the resulting normalized time series should match up well with climatological trends in the relevant extreme events.[38]

  For example, according to data kept by the National Hurricane Center, in 2005 Hurricane Katrina caused $80 billion in damage (in what’s known as current, or 2005, as opposed to inflation-adjusted, dollars) and in 2012 Superstorm Sandy caused $50 billion (in 2012 dollars). The scale of these numbers makes some sense to us.

  But consider the Great Miami Hurricane of 1926. It caused $76 million (in 1926 dollars) of damage almost nine decades ago, or about 1% of the damage caused by Katrina. That number makes little sense in today’s context, because that same storm hitting downtown Miami would surely cause much more damage than either Katrina or Sandy. But exactly how much damage would it cause today?

  The answer to this question has great relevance to insurers and reinsurers (those companies that provide insurance to insurance companies), policy makers, and residents of hurricane-prone regions. An entire area of financial analysis called “catastrophe modeling” has developed in the past several decades to address questions like these.

  Normalizing disaster losses leads to an estimate that the Great Miami Hurricane of 1926 would cause almost $200 billion in damage were it to hit in 2014. This would make it the most costly hurricane since 1900, if we were to rank all past storms based on what damage each would cause if they hit with today’s level of population and development.

  The logic here is simple, and can be illustrated with an example. Imagine a house on the beach in 2005. A hurricane comes through and badly damages the house, causing $100,000 in damage. Now imagine that same stretch of the beach ten years later, in 2015. Now, due to coastal development there are two identical houses on the beach. A hurricane of the exact same strength as the earlier storm blows through and damages both houses. The most recent storm causes $200,000 in damage.[39]

  In this hypothetical example, storm damage has doubled over a decade. However, the increased damage was not because of stronger or more frequent storms. The increase in damage was entirely due to the doubling of the amount of exposed property.

  We would be able to recognize the reason for the increased losses by looking at the data, and asking how much damage the 2005 storm would have caused in 2015. In this simple example, we would simply multiply the 2005 losses by two to arrive at our answer, as there is twice as much exposed property. The 2005 storm, had it occurred in 2015, would have caused $200,000 in damage. Because we have assumed identical storms in this simple example, the only variable that changes is the exposure to damage.

  In the real world, however, things are not so simple. Most obviously, houses are built using different practices. So let’s consider an alternative scenario. In this second scenario the second home is built with greater attention to damage potential, perhaps with reinforcements or a change in style. When the second storm passes through in 2015 there is only $50,000 in damage, for a total of $150,000 (that is $100,000 damage to the older home, plus $50,000 damage to the newer, stronger home).[40]

  Under this second scenario, how much damage would the 2005 storm have caused if it occurred in 2015? If we were to simply multiply the 2005 damage times two—reflecting that in 2015 there are now two houses—we would get $200,000, which is much higher than the $150,000 that is actually observed. We would thus have a bias in our results because we failed to account for the stronger house of 2015. It would be erroneous to claim from these data that a weaker hurricane occurred in 2015 because of the lesser damage.

  We could identify a bias in our results by comparing the normalized losses to the physical characteristics of the storms. Because the storms in this thought experiment are assumed to be identical, we might initially expect damages in 2015 to be twice those of 2005, as we did under the first scenario. The fact that the data do not match up (100% increase in storms but only 50% increase in damage) indicates that there is something left out of our normalization methodology. By comparing trends in damage to trends in storm frequency and intensity, we can check for evidence of a bias in our adjustments.

  Now consider a third variation on the thought experiment. In this version, imagine that there are
two houses on the beach in 2015, both identical to the single house present in 2005. In this case, a stronger storm makes landfall in 2015, causing $250,000 in damage to the two houses. In this case, after adjusting the 2005 storm to 2015 values, we would see that the normalized damage from the earlier storm had increased from $200,000 to $250,000, and that this increase in damage would be attributable to an increase in storm strength. We would know this not because of the adjusted loss data, but because of the data on storm strength.

  Of course our coasts have trillions of dollars in property across many millions of structures. Hundreds of storms over many decades create a complex record of damage and costs. Performing a normalization properly requires paying careful attention to the many societal factors which influence losses, but also to trends in the frequency and intensity of storms.

  It seems obvious, but is often overlooked, that in order for climate change, human-caused or otherwise, to contribute to increasing disaster losses, extreme events must become more frequent, more intense, or both. With more frequent or intense events, we would expect to see similar increases in normalized losses. Because data on the frequency and intensity of extreme weather events are usually collected separately from dollar losses, these two independent types of data provide a very important consistency check for any normalization process. Trends in each dataset should match up. If they don’t then there remains something to be explained in the data.

  When we conduct research to adjust past loss events to present day values, we are in effect asking how much damage would occur today if events of the past occurred with today’s level of population and economic development.

  So the first step in evaluating any normalization of disaster losses is to see if the trends in the adjusted losses correspond with trends in the frequency or intensity of the relevant events. If they don’t then there is a remaining bias in the procedure which needs to be addressed. You will see some numbers on this type of check in the sections which follow, but the logic here is simple: Trends in normalized losses should match up with trends in the relevant weather events.

  It is important to underscore that we do not look at normalized loss data in order to identify changes in extreme events or to discover a “signal” of climate change. The best place to look for evidence of changes in the frequency or intensity of extreme events is, not surprisingly, in data which directly reflect those extreme events.[41] Scholars are well aware of this issue, even if it does not make it into popular discussions of disaster losses, where growing damage by itself is sometimes simply assumed to reflect climate trends. At the conclusion of the next chapter, I summarize in a flow chart the role of normalization research in the search for a signal of human-caused climate change in the growing toll of disasters.

  As you will see in the chapters which follow, disaster losses have increased dramatically over recent decades. However, once past losses are adjusted for societal changes (more houses, more possessions, and so on), there is no remaining increase, leaving essentially no residual trend to be explained. In the several dozen normalization studies that have been published for phenomena around the world, you will find many different approaches to normalization. More frequent or intense extreme events, whatever the cause, are not necessary to explain the dramatic increase in disaster losses. Not only are they not necessary, but the evidence on extreme events is perfectly consistent with the normalization results.

  But I am getting ahead of myself. We will get to the data shortly, but first it is necessary to address a common complaint.

  Can Science Prove a Negative?

  Over the many years that I have worked on disasters and climate change, a common response to my work and that of my colleagues has been that it does not conclusively prove that human-cause climate change is not influencing extreme events.

  Not only are such responses absolutely true, but they are in fact truisms. Science cannot prove a negative.

  All that we can say is that the record of disaster losses is entirely explainable by changes in society. There is at present no evidence that human-caused climate change is responsible for any part of the global increase in disaster costs. We cannot say that there is no such influence.

  But as I have explained on many occasions, from a practical standpoint a signal that may exist, but which cannot be detected, is indistinguishable from a signal that does not exist.

  Yet the desire to prove a negative persists.[42] Some borrow a phrase most commonly associated with arguments over the existence of God and aliens: “absence of evidence is not evidence of absence.”[43] This catchy phrase is of course a tautology, hardly helpful and of little relevance to the central question of this volume. Science is concerned with evidence, not with supporting pre-existing beliefs.

  The philosopher Bertrand Russell provided a useful analogy for such arguments:

  Many orthodox people speak as though it were the business of skeptics to disprove received dogmas rather than of dogmatists to prove them. This is, of course, a mistake. If I were to suggest that between the Earth and Mars there is a china teapot revolving about the sun in an elliptical orbit, nobody would be able to disprove my assertion provided I were careful to add that the teapot is too small to be revealed even by our most powerful telescopes. But if I were to go on to say that, since my assertion cannot be disproved, it is an intolerable presumption on the part of human reason to doubt it, I should rightly be thought to be talking nonsense.[44]

  Russell continues, explaining that if the belief in the celestial teapot were widely affirmed and instilled, then “hesitation to believe in its existence would become a mark of eccentricity and entitle the doubter to the attentions of the psychiatrist in an enlightened age or of the Inquisitor in an earlier time.” While Russell was referring to arguments over the existence of God, he might as well have been talking about the climate debate circa 2014.

  The state of climate science today suggests that we should no more expect to see a signal of human-caused climate change in the increasing disaster losses of the past several decades than we should expect to find a teapot orbiting the sun.[45] More precisely, the IPCC’s climate model projections of changes in extreme events do not show identifiable increases in disaster losses for many decades, and often much longer.[46]

  Richard Dawkins, a long-time partisan in debates over religion and science, argues that a focus on the significance of an “absence of evidence” encourages “sloppy thinking”:

  Agnostic conciliation, which is the decent liberal bending over backward to concede as much as possible to anybody who shouts loud enough, reaches ludicrous lengths in the following common piece of sloppy thinking. It goes roughly like this: You can't prove a negative (so far so good). Science has no way to disprove the existence of a supreme being (this is strictly true). Therefore, belief or disbelief in a supreme being is a matter of pure, individual inclination, and both are therefore equally deserving of respectful attention! When you say it like that, the fallacy is almost self-evident; we hardly need spell out the reductio ad absurdum.[47]

  It is of course true that a role for climate change in the growing toll of disaster losses has not been excluded in any of the studies or assessments that are discussed here. In fact, such exclusion is a logical impossibility. Science cannot by its nature prove a negative. It would be equally true to state that science does not exclude a role for solar influences, cosmic rays, or, for that matter, evil leprechauns in explaining trends in disaster losses.

  The good news about the subject of disasters and climate change is that there is lots of evidence to look at; we don’t need to rely on clever logical constructions. The following sections of this volume assess some of the evidence with respect to the central question:

  Have disasters become more costly because of human-caused climate change?

  There are two possible answers to this question:

  1. Yes, the evidence indicates that disasters have become more costly because of climate change.

  2. No, ther
e is not sufficient evidence to indicate that disasters have become more costly because of climate change.

  There is a third possible position of course—agnosticism. However, given the strength and depth of the research in this area, the only legitimate reason for agnosticism is a lack of awareness of the relevant science and data. If you keep reading this volume, you won’t be able to take that position!

  A good analogy here would be evidence in support of claims that the Earth has warmed, on average, over the past century or so. The strength and depth of research on this topic indicates that it has. But it does not rule out the possibility that it has not warmed, of course (remember, science cannot prove a negative). The IPCC reports its findings with degrees of certainty, which never reach 100%. The degree of certainty expressed by the IPCC with respect to warming is similar to its degree of certainty about disasters and climate change.

  As a tonic against lack-of-awareness-induced agnosticism, the next sections present some of the data that help us to address the central question which is the focus of this volume.

  Have a look, and come to your own conclusions about what the evidence says.

  3

  The IPCC Framework for Detection and Attribution

  This section describes the framework of detection and attribution which underlies the work of the IPCC, and how it has been applied in studies that focus on the central question of this book.

  I have noticed in my years writing and speaking on climate change that what “climate change” actually means is not widely understood in a common way. For instance, the IPCC and the UN Framework Convention on Climate Change (and its most widely known policy agreement, the Kyoto Protocol) use different definitions of “climate change.”[48]

  Here I explain in some detail what it means to detect and attribute a signal of human-caused climate change in the disaster record, so that there is no confusion by what I mean when I discuss this topic. I rely on the IPCC definitions here, not because they are necessarily the best or final words on these subjects, but rather because they are widely accepted and used in the climate science community.