Date: Mon, 20 Nov 1995 15:34:39 -0600 (CST) From: raymond thomas pierrehumbert Subject: Moore critique =================== An essay on Thomas Gale Moore's "GLOBAL WARMING: A Boon to Humans and Other Animals" (Hoover Institution Essays In Public Policy #61). R. T. Pierrehumbert Dept. of Geophysical Sciences University of Chicago 1. Introduction "Global change is inevitable; warmer is better, richer is healthier." With these words at the conclusion of his essay, Thomas Gale Moore of the Hoover Institution expects us to dismiss all concern over the climatic effects of increasing atmospheric CO2. My objective in the following is to see whether there is any substance to Moore's argument. There seems to be a presumption running through Moore's and similar essays that environmentalists have a vested interest in believing that global warming will occur and be an unmitigated disaster. Some fringes of the movement may indeed see this as an opportunity to provoke lifestyle changes that they see as philosophically desirable. Others may just be genuinely alarmed at the evidence that we are now living in a world that is more like a fishbowl than an infinite universe -- we can see the impact of our existence in the global atmosphere. This should indeed give us pause, and it certainly heralds a necessity of a new level of awareness of what we may be doing to the environment. If we trash the planet, there will no longer be anyplace to run. Be that as it may, environmentalists have just as much reason as captains of industry for wanting to know the truth about global warming, since some of the "cures" for carbon dioxide emissions -- e.g. nuclear energy, expanded hydropower, and biomass fuels -- are in themselves quite environmentally destructive. For another perspective on this, let's suppose for the sake of argument that the Moore and his friends are right, both in asserting that global warming will be an insignificant problem, and that costs of CO2 mitigation would be high -- suppose on the order of $500 billion per year worldwide. If we knew that global warming weren't going to be a problem, then we could split the difference with our detractors, and spend the remaining $250 billion dollars on other environmental needs, such as acquiring land for wilderness and wildlife habitat. Only around $200 million would suffice to buy the land for the whole Maine Woods national park, for example. Regrettably, the trade-offs are hardly ever presented in this light. More typically, global warming skepticism comes packaged ideologically with a broad-based assault on environmental protection of all types. Perhaps, in the face of uncertainty about climate change, it is more cost effective to set aside large, connected habitats to allow for latitudinal migration. I have never heard any large voice in favor of anything like this from the Young/Pombo/Chenoweth "environment-be-damned" wing of the Republican Party. Those who predict certain disaster from global warming are overstating the confidence of the science; as a rhetorical device for alerting the public to what might go wrong, and why there is cause for concern, they may not be so far off the mark. Nonetheless, some balance in discussion of the impact of global warming would be welcome. Moore's article does not provide balance, however. Just what is his point? If it is that predictions of certain disaster are unjustified, we already know this and Moore's sloppy reasoning adds nothing to the assessment of the uncertainty. However, Moore implies much more than uncertainty Ñ the title and the narrative clearly imply that global warming will most likely be good for people, and for the natural world in general. As will be seen shortly, in this Moore is wholly off the mark, and his arguments need be given no more credence than we would accord any other propaganda set-piece. There are three parts to Moore's argument. The first attempts to argue for the benefits of warming, and against predicted dangers, on the basis of fundamental physical principles. The second argues that certain periods in the past provide a valid analogue for what a doubled-CO2 world would be like. The third Ñ and by far the bulkiest part Ñ is a historical discussion of good things that happened during relatively warm periods and bad things that happened during cold periods. Moore never delivers on the promise of the title to argue that warming would be a boon "... for other animals." This issue is not even addressed at the shoddy and inept level accorded to impacts on humans. 2. Moore's basic misconceptions regarding the physics of climate Of the many distortions, errors of fact, and tendentious over-simplifications in Moore's essay, one is so egregious as to merit special attention. "Studies of carbon (CO and CO2) in the atmosphere show that at times in the last eight thousand years the level has been substantially higher than it is today and greater than it is likely to reach any time soon." This is disinformation of the purest sort. The reference to CO is a red-herring, as carbon monoxide is a very minor constituent and in any event not a significant greenhouse gas. The truth is that CO2 is already at a higher level than it has ever attained anytime in the past 160,000 years (Crowley and North, hereafter CN, page 123), and we are contemplating doubling the concentration within a generation. How did Moore make such a blunder? He cites CN page 70 in evidence. The graph on this page indeed shows an increase as we go back in time -- but it is a graph of the fraction of carbon-14 in the atmosphere, not a graph of atmospheric carbon! Evidently, Moore is willing to take any graph which goes in the direction he wants as evidence for his thesis, without bothering with such niceties as figure captions or axis labels. There are also a number of other errors of fact and uses of obsolete references which, though not so fatal to Moore's argument, nonetheless show the Moore's tendency to root out and uncritically accept whatever seems to support his preconception. For example, he cites Newell's argument that evaporation keeps the tropical sea surface temperature from ever much exceeding 29 degrees C (Newell and Hsiung 1984). This argument had been discredited even at the time of the cited reference, and has been thoroughly rejected on unassailable physical grounds (see Pierrehumbert 1995 for a recent discussion of the issues). Similarly Moore's thinking about clouds is based on Lamb's 1972 analysis, which showed a net cloud cooling effect in the tropics and a warming effect in midlatitudes. The more recent ERBE data (ERBE team, 1989) yields a completely different picture, with cloud greenhouse (warming) and shading (cooling) effects almost exactly cancelling in the tropics, and summing to a net cooling in the midlatitudes -- in the present climate. I will have more to say about clouds shortly. In general, Moore reduces the theory of climate to the same kind of broad-brush homilies that gave us such disasters of Republican economics as the "Laffer curve" and the ballooning deficits of the Reagan years. In the following, I discuss Moore's principal conceptual failures which lie in this category. For the sake of economy of expression, I shall refer to the mythical world in which Moore's assertions apply as "MooreWorld." The Hydrological Cycle -- In MooreWorld warmer simply means wetter (by virtue of increased evaporation),and the only question is who gets the benefit of the attendant increase in precipitation. Infact, it is not even true that warming necessarily increases evaporation. Evaporation depends on many factors besides air temperature; increases in cloudiness or decreases in wind speed can reduce evaporation even in the face of warming of the atmosphere (cf Pierrehumbert 1995). Further, it is rather glib to dismiss concerns about drought by waving hands toward the distance and noting that increased precipitation must fall "somewhere." It does little good for agriculture if increased precipitation falls mainly over the ocean. Moore's lack of understanding of tropical climatology is astounding for one who claims to be addressing the impacts of global warming. Apart from monsoonal systems and the locally recycled rain over the Amazon, tropical precipitation is largely concentrated in a narrow band called the "Intertropical Convergence Zone" (ITCZ). The upward motion in this zone is compensated by the subsiding branch of the Hadley circulation. The subsidence -- which covers most of the tropics, is hot, clear and dry; it is the principle reason for the existence of the Sahara desert. If one warms the tropics and increases precipitation in the ITCZ one strengthens the subsidence and makes these areas even drier. The fact that the ocean is getting a better dump of rain will be of little solace to people trying to grow crops or graze cattle in sub-Saharan Africa. Ironically, this same drying in the subsiding regions was invoked by Lindzen (Sun and Lindzen 1993) as mechanism stabilizing tropical temperatures against CO2 increases, but the possible effects on drought were never emphasized; it could well be that the main effect of increased CO2 in the tropics may show up as moderate warming plus increased subtropical desertification, rather than a more severe warming. It should also be noted that while increased evaporation over the ocean yields at least the possibility of increased precipitation over agricultural areas, increased evaporation over land is on the contrary inimical. The increases in land evaporation in a hot summer can offset increased precipitation in the rest of the year, leading to decreases in soil moisture -- which is what plants really are responding to. Moore conveniently forgets this consideration, at least until he needs to invoke it to dry out Medieval swamps. Moore's attitude to this, by the way, shows an interesting bias against the value of wetlands and this is far from the only instance in which Moore's biases reveal a striking adherence to the Republican Party line. Drying out wet areas can be disastrous for many forms of wildlife. An interesting example, noted by McKibben, is that the open canopy of some managed Eastern industrial forests lets more sunlight reach the forest floor, eliminating pools of water which frogs and other amphibians require for breeding (McKibbin 1995). Storms and Winds-- In MooreWorld, a warmer climate means lower pole-to-equator temperature contrasts, which lead to less energy source for the jet stream winds and storms, implying weaker storms and weaker winds. In jumping to this conclusion, Moore is fatally ignorant of the role of latent heat release (due to condensing moisture) in energizing the atmospheric circulation. Let's take the matter of jets first. While the midlatitude jet stream is at least partly driven by meridional temperature contrast, the subtropical jet (which is fuzzed together with the midlatitude jet in the present climate) is caused by displacement of air from the equator toward the pole, which creates a jet due to angular momentum conservation. The jet is a lot weaker than it would be for a purely conserving flow, owing to large scale turbulent dissipation in the wave-breaking zone in the subtropics. The circulation driving the subtropical jet is the Hadley circulation, which is energized by latent heat release in the deep tropics. Warming, and more vigorous convection, can lead to a stronger subtropical jet. There are some model indications that this may have happened in warm climates such as the Cretaceous (Barron and Washington, 1982). Thus, far from warming simply weakening winds everywhere, we might expect a shift in the circulation pattern to favor the subtropical jet over a midlatitude jet. While there isn't currently any indication that the midlatitude jet stream would get stronger in a 2xCO2 world, for the record I note that midlatitude jets do not really require strong meridional temperature gradients for their existence, since eddy fluxes can spin up strong jets with little vertical shear. The eddies "kick" the jet as if it were a flywheel, and allow it to build up speed. This process can be energized by latent heat instead of the pole/equator temperature gradient. The gas giant planets (like Jupiter) provide a good example of this; in fact the planets with the weakest temperature gradients (going out to Uranus and Neptune) appear to have the strongest winds. Although Moore's argument does not tell us much one way or another about the behavior of jets, there is in fact modelling evidence that at least the midlatitude jet would becomes weaker in a warm climate (e.g. Barron and Washington 1982). This is not at all a good thing. It is the winds of the jet that carry the moderating effect of the oceans over into the climate of the interior of continents. When the jet becomes weaker, this effect is less pronounced (e.g. Sloan and Barron 1990). The consequences are particularly worrisome for summer; weak winds can allow stagnant hot air to pool over the continents unchecked by the moderating influence of maritime air. This amplifies summer warming over land, exacerbating killer heat waves such as experienced by the U.S. Midwest in 1995. Deleterious effects on agriculture are also to be expected from such episodes of extreme heat. Now on to storms. Moore is confusing the two main types of storms: baroclinic eddies (which are primarily driven by potential energy in the meridional temperature gradient) and tropical storms (which are primarily driven by latent heat release. In this Moore is in good company, as the confusion in the lay mind dates back at least to Aristotle's Meteorologica, as I mention in my review article (Pierrehumbert and Swanson, 1994). Tropical storms, like hurricanes, live off latent heat -- that's why they primarily affect coastal areas. They are most prevalent in warm-water regimes, and do not need any meridional temperature gradient for their existence. It is this class of storms that might become more prevalent or more intense in a warmer world. It is well established physics that the maximum intensity of hurricanes is determined by the sea surface temperature (c.f. Emanuel 1987,88). I also note that there are a wide variety of latent-heat-driven mesoscale systems over land that lead to dangerous weather over land (e.g. severe thunderstorms and squall lines. I also wouldn't be so quick to dismiss major changes in the nature of the midlatitude synoptic storm tracks. While the primary energy source for midlatitude storms is potential energy, they do draw some energy from condensation of water, and as the climate warms, they can draw more and more energy from this source. Further, condensation reduces the effective vertical stratification, and since the characteristic scale of storms is the "deformation radius," which is proportional to the stratification, one would expect a shift toward smaller vortices. In general warming begins to blur the distinction between synoptic eddies and tropical storm systems. At present it is unclear whether warming would lead to more frequent but weak storms, or infrequent but very intense small eddies. It is clear, though, that monkeying with the climate will change the nature of the midlatitude storm tracks, and not necessarily for the better. Clouds -- In MooreWorld there is little work for cloud physicists, because clouds behave in such a stunningly simple way. A quote from Moore's appendix is symptomatic: "Any warming of the earth will increase evaporation and cloudiness, which will moderate increases in temperatures from the equator to at least the 35degree north and south meridians..." One would have to work hard to cram more misconceptions about clouds into a single sentence. As noted previously, warming need not increase evaporation, but even if it does, increased cloudiness does not follow as an inevitable or even particularly likely consequence. The appearance of clouds is limited not by the amount of water in the air, but by the amount of saturated air the climate system produces; there is more than enough water in the air already to create a cloud-swathed planet like Venus, if only it were distributed properly. The amount of uncondensed water vapor in a column of tropical air would make a puddle 2-3cm deep if rained out, while the condensed cloud water in one of the all-important high cirrus clouds would make a scant tenth of a millimeter. Clearly, water supply is not the limiting factor. Hyper-arid air at 5% relative humidity with a temperature of 27C would become saturated and make a cloud if lifted about 8 km. One could make a plausible (though simplistic) case that warming would reduce cloudiness, as warming exponentially increases the amount of water air can hold before condensation occurs. To be sure, if evaporation increases, then precipitation must also increase. However, most clouds are not precipitating, and increased precipitation does not require an increase in cloud cover, as it can be accomodated in limited area but intense systems. This is not the end of the problems with Moore's view of clouds and climate. As Moore correctly notes,clouds have a dual effect -- they act as a greenhouse gas, which traps heat and warms the planet, but they also reflect solar energy back to space, which cools the planet. The net effect of clouds is the difference of two large terms. In the present-day Tropics, these terms are individually up to 80 Watts per square meter (compared to average incident solar radiation of around 400 Watts per square meter) but sum up to nearly zero. A warmer world need not preserve the proportion between cloud greenhouse and shading effects that happens to prevail today. If the cloud greenhouse effect became greater as compared to the cloud shading effect, then increased cloudiness would substantially amplify the warming due to the direct effects of CO2. Given the large values of the individual cloud terms, it would not take a very big shift in the typical cloud height to produce a significant destabilizing feedback. Atmospheric models incorporate such effects, and indeed difference in cloud treatments account for much of the differences in predicted warming amongst the models. There is much more to be learned about cloud modeling, but for better or for worse, these models represent the best guidance we presently have about the range of possible behaviors of the atmosphere. Even if it should turn out that clouds stabilize the surface temperature, this would nonetheless represent climate change. The increased cloudiness would reduce the surface solar radiation, weaken the hydrological cycle, change the vertical distribution of atmospheric heating, and probably increase the stratification of the upper troposphere. I don't think it can be said yet that we understand what the resulting world would be like. Moore's cloud fallacy make a number of appearances. In his discussion of the effect of warming on agriculture for example, he states, "As a result of more evaporation from the oceans, a warmer climate should intensify cloudiness." Moore also shows his willingness to jump to any appealing (to him) conclusion, when he asserts that clouds account for the observed reduction in the day/night temperature range. These things can be quantified, and scientists are aware of the issue, but it takes more than an assertion to make a theory. In fact, the cloud effects have been studied (Hansen et al 1995; see esp. the survey in Harvey 1995.), and it was found that they could not account for the reduction in temperature range without also leading to a substantial cooling (contrary to observations). It appears that aerosols from sulfate pollution are playing a temporary role in the observed temperature range decrease, and that as CO2 continues to increase, the daytime temperatures will ultimately resume their rise in step with night-time ones. I note further, in this connection, that while Moore is happy to posit beneficial effects of clouds in moderating the day-night temperature range, he doesn't give any consideration to the fact that increased cloudiness would also reduce the amount of solar radiation reaching the ground -- reducing the energy source upon which photosynthesis -- hence agriculture-- depends. Glaciology-- The increased precipitation in MooreWorld is quite simply expected to lead to growth of the ice caps. This greatly simplifies the work of MooreWorld glaciologists. In RealWorld, glaciologists have to contend with the fact that ice is a fluid, and that the growth or decay of ice caps depends on a delicate balance between accumulation and flow to the margins, where ice breaks off and melts. Warming exacerbates the latter, and it is when this effect wins out that the ice cap shrinks. In support of his thesis, Moore invokes a Letter to the Editor of the Wall Street journal, purporting to show that the Greenland icecap is growing rather than shrinking. It would have been nice if Moore had cited the more balanced discussion of the Greenland ice cap which appears in the IPCC report, but that is somewhat beside the point. Greenland does not represent the main repository of the world's ice, and the behavior of Greenland is not necessarily indicative of the behavior of other ice caps; it has been shown for example, that snow accumulation on Greenland is sensitive to minor shifts in the storm tracks (Kapsner, et al 1995), and not primarily to temperature. Nor are short term trends indicative of the long term fate of the ice caps; the recently discovered importance of glacial surges shows that glaciers can accumulate for a long time before building up enough pressure to melt the basal ice, wherupon they purge themselves in a catastrophic surge (Macayeal 1993ab). The behavior of glaciers could easily have a number of nasty surprises in store. Climate modeling -- Although MooreWorld is a congenial place for cloud modellers -- because MooreWorld clouds behave according to a few simple rules which have remarkably escaped the notice of RealWorld cloud modelers -- woe betide any General Circulation Modelers who try to actually use physical principles to predict changes in climate! While no atmospheric scientist would argue that climate models are perfect, Moore consistently exaggerates the inadequacies of the model. He says, "Should the world warm --and there is little evidence or theory to support such a prognostication..." No theory? Does he want to throw out all of quantum theory and radiative transfer developed in this century, that tells us how much CO2 and water vapor absorb infra-red radiation? All the thermodynamics that says why much of the planet becomes moister in a warmer world? All the satellite evidence that has been confirming the nature of the water vapor feedback? All the evidence that warm past climates like the Cretaceous (the last age of the Dinosaurs) cannot be accounted for unless the sensitivity to CO2 increase is something like what the models say it is? Moore complains that the models "...cannot even replicate the current climate." This is a really vacuous statement. Just what does he mean by "replicate?" Do we need to get it right to within a millionth of a degree to satisfy Moore? Or is it OK if the models just get fundamental things (like the pole to equator temperature difference) right, which they certainly do? Actually, the accomplishments of the models have been quite remarkable. They have been tested against the observed seasonal cycle and El Nino cycle (see survey in IPCC 1995, Section 4.4), and against the cooling effect of the Mt. Pinatubo eruption. The comparison extends even to the verification of the important water vapor feedback against satellite data (Soden 1995). The models have been checked against climate variations over the past glacial cycle (COHMAP 1993). Britain's Hadley Center modelling group has been able to match the observed temperature record over the past century by taking the temporary cooling effect of sulfate polution into account. The models even are able to reproduce detailed statistics of short term variability with considerable accuracy (Lau and Nath 1987). Moore remarks that the National Academy of Sciences is "skeptical" of the treatment of clouds in models, and on this basis would like to have us discount the model results altogether. In this, Moore shows his ignorance of the operation of science. Scientists have to be "skeptical" about their models and theories all the time, otherwise there would never be a drive for improvement. This does not mean that nothing of value is known, and it is a gross distortion of the Academy's position to infer that the models have no predictive value. There is a big difference between not knowing everything, and knowing nothing whatsoever. In the case of clouds, the effects are indeed incorporated in all models, and different assumptions have been made in the different models. This means that the range of predictions amongst the different models already to some extent reflects the uncertainty in the range of possible cloud responses; if the cloud feedbacks were really so horribly in error, it is unlikely the models would do as well as they do in reproducing the observed seasonal cycle. Moore can certainly claim that the case for pronounced warming is not airtight, but it is ridiculous to say there is "little evidence or theory" to support warming. The models make a more than adequate case that there is a realistic risk of warming. Insofar as models are imperfect, it is erroneous to assume that all model errors will work in our favor. Future revisions are as likely to yield larger warmings as smaller. This remark applies especially strongly to the uncertainties in cloud effects. Moore disparages atmospheric models even though they are based on firm physical principles and even though they have been subjected to stringent and quantitative verification studies. Strangely, he does not hold economic models-- upon which many global warming impact studies rely-- to the same impossibly high standards. He is happy to quote the study of Mendelsohn et al. (1994) on the impact of warming on agriculture, but the manifest shortcomings of their model do not seem trouble Moore in the least. Their model applies a uniform 8% increase in precipitation across the United States during all seasons, even though many atmospheric models indicate a decrease in rainfall over much of the U.S., and all indicate considerable seasonal and regional variations. The model assumes a uniform 2 degree C (5 Fahrenheit) warming over all latitudes and seasons, whereas agriculture is very sensitive to extremes in temperature and the local temperature extremes in the US are expected to extend far beyond the 2C globally averaged warming. The regression model used by Mendelsohn et al. does not adequately account for the steep rise in evaporation with summer temperature, and so cannot handle the competition between increased precipitation and increased evaporation; without this, the effects of reducing soil moisture (which is what agricultural plants really care about) cannot be reliably estimated. The model in many cases implicitly assumes that farmers will react by shifting to higher-value crops (like oranges or avocados) that require more irrigation, but fails to take into account whether there will be a market for the increased production, or the water supply or irrigation cost issues, or the distorting effects of water subsidies on prices. Finally, the regression-based method employed by Mendelsohn et al. seeks to infer effects of future climate change from present regional variations in climate; however, it cannot adequately distinguish amongst the effects of different variables that are highly correlated in the present climate. For example, southern areas of the US are warmer, but also receive more light averaged over the year. It is nearly impossible to disentangle the two effects, but we do know that warming the planet will not change the latitude of Madison, Wisconsin! The Mendelsohn et al (1994) study is really quite interesting, and the authors are very scrupulous in noting its shortcomings. I have no quarrel with them. I do find fault with Moore's willingness to uncritically accept results that support his thesis, while fabricating reasons to doubt those he would like to ignore. 6. References [**CN = Crowley and North. Paleoclimatology. Oxford University Press] Soden B 1995: [**New J. Climate paper on water vapor intercomparison.] IPCC 1995: Climate Change : The IPCC Scientific Assessment . Cambridge University Press. COHMAP group 1993: Global Climates since the Last Glacial Maximum . University of Minnesota Press, Minneapolis. 568 pp. Manabe R and Stauffer 19** [**Coupled ocean run; use as citation for regional warming in doubled CO2 vs. Medieval Optimum] Hansen J, Sato M, and Ruedy R 1995: Atmos. Res Ê37, 175-209. Harvey LDD 1995: Warm days,hot nights. Nature 377, 15-16. Pierrehumbert, R.T. and K.L Swanson 1994: Baroclinic Instability Ann. Rev. Fluid Mech. 27, 419-467. Pierrehumbert, R. T. 1995: Thermostats, Radiator Fins, and the Local Runaway Greenhouse. J. Atmos. Sci. 52, 1784-1806. Mendelsohn R., Nordhaus WD, and Shaw D 1994: The impact of global warming on agriculture: A Ricardian Analysis. The American Economic Review 84, 753-771. Sloan, L. C. and E.J. Barron 1990: "Equable" climates during Earth history? Geology 18, 489-492. Newell, R. E. and J. Hsiung, 1984: "Sea Surface Temperature, Atmospheric CO2 and the Global Energy Budget: Some Comparisons between the Past and Present," in Morner & KarlÚn, eds. [1984]: 533-562. McKibbin, B. 1995: An Explosion of Green. Atlantic Monthly 61-83. Barron, E.J. and W. M. Washington 1982: Atmospheric circulation during warm geologic periods: Is the equator-to-pole surface-temperature gradient the controlling factor? Geology 10, 633-636. ERBE science team, 1989: Cloud-radiative forcing and the climate: Results from the Earth Radiation Budget Experiment. Science 243,57-63. Sun, D-Z, and Lindzen, R. S. 1993: Water vapor feedback and the ice age snowline record. Ann. Geophysicae 11, 204-215. Emanuel, K. A. 1987: The dependence of hurricane intensity on climate. Nature 326, 483-485. Emanuel, K. A. 1988: The maximum intensity of hurricanes. J. Atmos. Sci. 45, 1143-1155. Macayeal, D.R. 1993a: Paleoceanography 8, 767-773. Macayeal, D.R. 1993b: Paleoceanography 8, 775-784. Lau, N-C and Nath, M. 1987: Frequency dependence of the structure and temporal development of wintertime tropospheric fluctuations--Comparison of a GCM simulation with observations Mon. Wea. Rev., 115, 251-271. W.R.Kapsner, R.B.Alley, C.A.Shuman, S.Anandakrishnan & P.M.Grootes 1995: Dominant influence of atmospheric circulation on snow accumulation in Greenland over the past 18,000 years. Nature 373 , 52-54 An essay on Thomas Gale Moore's "GLOBAL WARMING: A Boon to Humans and Other Animals" (Hoover Institution Essays In Public Policy #61). R. T. Pierrehumbert Dept. of Geophysical Sciences University of Chicago ===========CONTINUATION; SECOND AND FINAL HALF=================== 3. Global Warming deja vu -- Have we been there before? Moore claims that the human species and human society have experienced sustained warm climates in the past, which are comparable in magnitude to what is expected from doubling CO2, and have if anything flourished in such climates. Most of his evidence is drawn from the Medieval Warm period, with a few anecdotal pieces drawn also from the Holocene Optimum (about 6000-7000 years ago). These can be clearly shown to be invalid analogues for a doubled CO2 world, both with regard to the magnitude and the character of the climate change. There are also a few oblique references to the Eem interglacial (about 120,000 years ago) and warm climates of the distant past, such as the ice-free Eocene (50 million years ago) or Cretaceous. I will concentrate my discussion on periods that are warmer than the present. Clearly, the Ice Age (little or big) was a tough time, but the issue isn't whether climates much colder than today would be a problem. The issue is how people and wildlife would cope with a substantially warmer climate. Before beginning my detailed discussion of the periods in question, I would like to make one remark on a tacit assumption underpinning all of Moore's methodology: namely, that modern technological society can deal with climate change better than primitive past societies, so that examining past impacts represents a worst case. It is not even clear if this is true for impacts on human societies, narrowly construed. We have a far larger population than in the distant past, so there is less opportunity for peoples to adapt by migration. We have more people living in vulnerable areas, or areas of marginal productivity. We have complex systems that may be very expensive to adapt to climate change: think of the effect of shifting precipitation patterns on existing hydropower facilities, or on the complex web of water and irrigation rights. However, I am willing to accept that at least the richer societies will find some way to survive without great harm, though perhaps at greater cost and inconvenience than if CO2 emissions had been sensibly limited. Moore's thesis is clearly untrue for wildlife, though. In the past, wildlife has responded to climatic change largely through migration, with many extinctions being associated with barriers to migration. Our ecosystems today have been fragmented as never before, and the area of habitat has already been woefully reduced by economic development. Opportunities for wildife to survive by migrating are far more limited than in the past. Global warming will bring changes in forest type (deciduous vs. conifer), a reversion of much tundra to boreal forest, changes in the ocean's nutrient distribution, and changes in the seasonal pattern to which many creatures time their life cycles. I fear that global warming will administer the coup de grace to many species already stressed and brought to the brink of extinction. We will never get them back. To those who imagine some technical fix to help species adapt, I need only ask how well we have done with the relatively simple problem of helping the Idaho Snake River salmon population survive the environmental change due to dams and logging. The answer is, "not well." In fact -- not at all. The Medieval Warm Period -- The term "Medieval Warm Period" refers to a period of a century or more centered roughly on 1200AD, which was conventionally believed to be warmer than the present. Almost all of Moore's historical anecdotes about the effect of warm climates on society come from this period. The failure of Moore's argument with regard to this period is especially simple: contrary to Moore's supposition, the documented local temperature increases (relative to the present) over land during this period are far less than the temperature increases expected in connection with a doubling of CO2. Part of Moore's problem comes from a use of outdated references, and most notably an excessive reliance on the documentary evidence given by Lamb. There has been a great deal of progress in the study of historical climate change in the past decade, and much of Lamb's simplistic picture has been thrown into doubt. In a recent survey of the subject, Hughes and Diaz (1994) report, "The documentary evidence presented by authors such as Lamb for a widespread Medieval Warm Period has been weakened to a significant degree by recent work ... based on the rigorous application of historical criticism to sources." They note that 45% to 50% of the historical sources used by Lamb have been found to be invalid or dubious, and that Lamb's "winter severity index" for England's climate is 80% in error. In many instances, even Lamb's assessment of whether a period was warmer or cooler than present is dubious; Lamb's quantitative estimates of the magnitude of the supposed warming is even more in doubt. Reviewing a broad range of physical and historical indicators of past climate, Hughes and Diaz conclude that "the simplified representations of the course of global temperature variation over the last thousand years reproduced in various technical and popular publications (....) should be disregarded, since they are based on inadequate data that have, in many cases, been superseded." Such warming as has been confirmed is limited in magnitude, applies to only certain geographical regions, and the warmings in the different regions are not strongly synchronous. Rigorously calibrated tree-ring data provide some of the best quantitative indications of the magnitude of warming over the past millenium, and show that when short term variations are smoothed out, the maximum warming over the land sites that apply to Moore's analysis is generally less than .5 degrees C (.9 degrees F) -- in most cases much less. Some examples: For the far north of Fenno-Scandia, Briffa et al. (1992) show a maximum summer warming of no more than .54 degrees C for any period in the past millenium. Temperature reconstructions for Northern Italy (Serre-Bachet,1994) show barely one tenth of a degree C warming. The Sierra Nevada shows maximum warmings of only a quarter degree, when averaged over fifty years. How do these compare with the magnitudes of warming expected from doubling CO2? First we have to clear away some further wool Moore has attempted to interpose between the reader and the truth. Predictions of global warming from doubling CO2 are generally summarized by a "global, annual average" warming, meaning an average over the whole globe and over the whole year. This measure is currently expected to lie in the range 2-4 degrees C (3.6 to 7.2 degrees F). Some regions and seasons will warm much more than the average, and others much less. Moore confuses -- let us be charitable and say inadvertently confuses -- the comparison in two ways. First, he sticks to the 2C value for global warming, though this value represents the middle of the range of the predicted warming and not necessarily the most probable value. The higher end, 4C, should have been considered as a plausible worst case, if we are really trying to see how bad things could get. Second, and more critically, he compares the figure for global annual mean warming in the models, to regional (and sometimes seasonal) warming in the climate record. In doing so, Moore greatly underestimates the warming predicted for the specific midlatitude regions for which his arguments are mainly made. Actually, given the small estimate of the Medieval warming understood from modern analyses, even the predicted global mean warming is far in excess of any sustained warming seen in the pre-industrial millenium. When regional variations of the predicted warming are taken into account, Moore's claim that the Medieval era is any kind of analogue for a doubled CO2 world becomes even more ridiculous. For example, the summer warming predicted by the GFDL model over the central US is fully 5C (9F)by the year 2050 -- which is TEN TIMES the medieval warming seen in the tree-ring data. To put that into perspective, that means that the 104 degree heat wave experienced in the Midwest in the summer of 1995 will seem like nothing unusual in 2050; heat waves of comparable rarity will be 113 degree heat waves. That is not the end of the bad news, because the predicted winter warming, and high latitude warming, can be even greater. Some might think that winter warming would be no bad thing, but one must consider that control of many agricultural pests and forest infestations (e.g. the oporinia moth) relies on hard freezes to kill the dormant eggs. The high latitude warming is expected to lead to massive reductions in tundra ecosystems, with unfavorable impacts on wildlife and on migratory animals who time their migrations to expected appearances of their dominant food sources. Warmer winters increase the maximum possible severity of snowstorms, because the current very cold air is too cold to hold much water; it is not yet possible to say how frequent such events would be, but the risk is there. It is sobering to note that the figures quoted for 2050 are not even equilibrium figures. Even if the CO2 increase were to be held constant after reaching twice its pre-industrial value -- and there is no reason to expect this to be the case barring mitigating policies -- the world would continue to warm for at least another 50 years. So, this is what I have learned from Moore. If all the changes in Medieval society which Moore attributes to climate truly are so caused, then we are in truly big trouble if we double CO2 -- because human society is really terribly sensitive to changes of temperatures of a half degree C or less! The Holocene Climatic Optimum-- The "Holocene Optimum" refers to the period about 6000 years ago, which is the warmest period over land in the Northern Hemisphere experienced since the last great ice sheets retreated. While there is little doubt about widespread warming of some sort, quantitative estimates of temperature change over land are hard to come by and of uncertain reliability. The yearly average temperature in the Western US increased by no more than .7C, though local summer warmings of as much as 2.3C are estimated for the warmest subregion (COHMAP 1992, p. 497). The pattern for the Eastern half of North America is quite complex, with January coolings of 4-8C in the Southern regions, compensated by warmings of up to 4C in the Canadian regions (COHMAP 1992, p. 451). Summer warming is more wide spread, indicating a climate 1-2C warmer than the present summer (COHMAP 1992, p. 452). Europe may have warmed by about 2C in summer (CN p. 68). The largest of the summer warmings encountered locally do begin to become comparable to what is expected from doubling CO2. There is little evidence that the oceans were significantly warmer than at present, on the other hand. Despite the significant warming, the Holocene Optimum is an unreliable analogue, because it was caused by climate forcing mechanisms that operate very differently from doubling CO2. 6000 years ago (roughly speaking), the Earth was closest to the sun during the Northern Hemisphere summer, i.e. when the North Pole was tilted toward the sun. Conversely, the Earth was farthest from the Sun during Southern Hemisphere summer. This enhances the heat of Northern summer and decreases the heat of Northern winter, while decreasing the heat of Southern Hemisphere summer and moderating the cold of Southern winter. Unlike CO2 doubling, which warms the whole planet everywhere, year 'round, the orbital effect cancels out to nearly zero at any point on the globe when averaged over the year, and cancels out to zero at any time of year when averaged over the globe. This is why the slow-responding oceans show less warming than land -- in effect they average out the sunshine changes over the year, whereas land responds essentially instantaneosly to the stronger Northern Summer sunshine. It is also why the most pronounced warmings in the Holocene Optimum occur in Northern Hemisphere summer. It should be noted that changes in the atmospheric circulation pattern redistribute warm and cold air, and can cause local climate changes which differ substantially from the simple picture that might be expected on the basis of changes in amount of absorbed sunshine. Moore is fond of quoting the noted paleoclimatologists Crowley and North, but only when it suits him, or can be made to appear so by taking a conclusion out of its context. Here is an apt quote he has missed though. Referring to the Holocene Optimum, they say: "In fact, although seasonal warming was locally impressive, there may have been at most a few tenths of a degree warming in mean annual global temperatures during these times. Even minimum greenhouse scenarios project warming larger than this. The analogy between future and past warmings in the Quaternary therefore does not appear to be appropriate." ( CN, p 89). Moore would argue that land warming is all he needs in order to look at impacts on human societies (and ecosystems, I suppose, if he ever got around to it). This ignores the fact that there is more to climate than temperature changes, and that many important features of the climate are strongly affected by temperature patterns at remote distances over the ocean. Not all warm climates will be the same with regard to precipitation, storms, winds and the seasonal cycle. All these features are of great importance to the way in which climate affects human societies and wildlife. The strength of the jet streams and the storms that grow on them, and the growth of hurricanes, are determined largely over the oceans. The atmospheric behavior over the oceans affects the large scale bending of the jets that can lead to extreme fluctuations like episodic droughts. The precipitation vs. evaporation over the ocean is critical to driving the oceanic circulation, which in turn greatly affects the degree of climate variability (cf. Broecker, Scientific American 1995). For that matter, the ocean circulation is controlled globally, since events in the Antarctic couple to and feed back on the circulations in the North Atlantic which give present-day Europe its mild climate. It is also obvious that truly global warming with strong high latitude impacts -- such as expected from doubling CO2-- will have a very different effect on the Antarctic ice cap than a warming that increases Southern Summer temperatures little if at all. Then, too, a warming which occurs primarily in the summer will have a very different impact on insect pests, disease vectors, and forest vegetation than one (like doubling CO2) which yields significantly milder winters. The caveat that not all climate forcing mechanisms yield the same kind of response applies especially to Moore's favorite eye-catching example: the blooming of the Sahara during the Holocene Optimum. It is quite well established (COHMAP 1993) that the increased rainfall in this region during the Holocene Optimum is due to an enhanced Monsoon circulation. Monsoons are driven by the difference between ocean and land temperatures, which arise because land can heat up more rapidly in response to the intense summer sunshine than can the well-mixed ocean. This driving was stronger during the early Holocene, because of the orbital effects mentioned earlier. Increased carbon dioxide affects the climate in a very different way from the orbital forcing. The CO2 effect is not seasonally dependant, and increases the energy input to land and ocean equally. While the behavior of monsoons in a doubled CO2 world is an interesting and important topic, there is no reason to expect the monsoons in a greenhouse world to behave like the monsoons in the early Holocene. Again, Moore's supposed analogue is not apt. Other warm periods -- We are currently in a period known as an "interglacial," between major ice sheet advances. These happy circumstances occur roughly once every 100,000 years. During the previous ("Eemian") interglacial era (120000 years ago), Western Europe may have been 1-3C warmer than it is in the current interglacial. There is scattered evidence of warming from other land sites around the globe, but it is not clear that the warm periods were synchronous. However, it is quite clear that the ocean temperatures were not much different from the present interglacial (see CN p 116ff), and so the Eemian is still not much of an analogue for a doubled CO2 world. There is also some disturbing, though preliminary, evidence from the Greenland ice cores that the warmer Eemian also experienced much greater climate variability than we do presently. In any event, if Moore wishes to argue that higher land temperatures are generally good for people, he will have to account for why civilization didn't arise during the Eem interglacial. Our Eemian ancestors were hardly genetically different from anyone you would meet on the street today, and so were as equipped as anybody to take advantage of the warm climate they were "enjoying." This leaves us with the very warm, ice-free climates predating the current spate of ice ages: most recently the Eocene (50 million years ago) and Cretaceous (65 million years ago, the last age of the Dinosaurs). These ages are in fact the best argument that the sensitivity of climate to CO2 is something like what our models says it is --- because it would be hard to account for the existence of such climate states otherwise. These certainly are warmer than what is expected from doubling CO2, so if Moore wants to use these past climates to argue that the earth won't boil off the seas in a runaway greenhouse, and that life won't come to an end, that's fine with me. However, these really were massively different climates than the present, and he can hardly expect to turn the clock back 50 million years in the course of a century, without causing some massive disruption to the Earth's societies and wildlife. For that matter, it may be a mistake to assume the hothouse world of the Cretaceous was all that pleasant. In discussing the "lush, rich world" of the Cretaceous, Moore may have been overly influenced by the fanciful swampy murals hanging in the Field Museum of Chicago, and many other natural history museums. It is known now that Apataosaurus and many other favorite dinos actually lived in a climate that was dry as a desert for much of the year. Much of the tropics evidently received much less rainfall than the present tropics, and the interiors of tropical continents may have been barely habitable. Global warming in the perspective of earth's climate history-- For the past 10000 years, the Earth has enjoyed a remarkably stable climate (cf Broecker, Scientific American 1995 for an accessible review). During this time, human civilization arose and flourished. In doubling CO2, we are breaking out of the top of that regime, and there is a realistic danger that ocean circulation feedbacks may kick in to restore the greater chaotic variations in climate prevailing in earlier times. Over the past million years, climate has fluctuated between warm states similar to the present, and glacial states 4-5C colder than the present. This encompasses the whole of the experience of the human species, and of many of the species with which we share the planet. CO2 is already somewhat higher than it has ever been during any of this time, and if we don't take action, it is expected to double at least. In doing so, we are taking the system past the top of the range it has been in for the past million years. Crowley and North point out that the combination of a warm, high CO2 atmosphere with glaciated poles may be unique in the billions of years of earth history. As a species, we are leaving our climatic home, and we are departing at (by geologic standards) breakneck speed. And, given current uncertainties in climate models, we are going there with only the guidance of a road map crudely drawn by children just getting used to their crayons. Perhaps it would be prudent to slow down the trip. 4. Playing fast and loose with history =======SECTION TO BE COMPLETED SOME OTHER TIME============== 5. Conclusion Moore is right in pointing out that climate fluctuates naturally, but he is wrong in his assertions that the human species has lived through climate changes of the type expected from doubling CO2. Further, to the extent that there is natural climate variability, it is erroneous to assume that this will be suspended just because people are also causing climate to change. The human-induced changes can add to, and interact with, the natural variability to yield extremes of fluctuation that are not easy to comprehend at this point. Moore wonders if it "...might just be hubris." on our part to think that our own activities can change the climate. I find this a curious attitude. It is hardly a novel, or even controversial, idea that the presence of life has changed the planet. The advent of life accelerated carbon burial, and the drawdown of carbon dioxide in the atmosphere. The advent of photosynthesis gave us an oxygen atmosphere, lead to the ozone layer, and changed the temperature structure of the stratosphere. The advent of tiny animals in the upper ocean, owing to their fecal pellets, allowed carbon to be taken out of the upper ocean faster than its decomposition could suck up oxygen, leading to still higher oxygen levels. The advent of burrowing worms on the sea floor, once the bottom was oxygenated, stirred the sediments and increased the layer involved in ocean chemistry from a millimeter to a couple centimeters. Now, Homo Sapiens is the ultimate burrowing worm, reaching kilometers down into the crust of the Earth and rooting out carbon sources that had remained safely buried for a hundred million years or more. Are we to believe that humans, accounting for a significant and expanding proportion of the Earth's biomass, have less ability to change the climate than the lowly worm or fecal pellet? I doubt it, but let us hope that at least we have more foresight about what we are doing. 6. References Barron, E.J. and W. M. Washington 1982: Atmospheric circulation during warm geologic periods: Is the equator-to-pole surface-temperature gradient the controlling factor? Geology 10, 633-636. Briffa KR, Jones PD, Bartholin TS, Eckstein D, Schweingruber FG, Karlen W, Zetterberg P and Eronen M 1992: Fennoscandian summers from AD 500: temperature changes on short and long timescales. Climate Dynamics 7, 111-119. COHMAP group 1993: Global Climates since the Last Glacial Maximum . University of Minnesota Press, Minneapolis. 568 pp. Davis OK 1994: Climatic Change 26 , **-**. Emanuel, K. A. 1987: The dependence of hurricane intensity on climate. Nature 326, 483-485. Emanuel, K. A. 1988: The maximum intensity of hurricanes. J. Atmos. Sci. 45, 1143-1155. ERBE science team, 1989: Cloud-radiative forcing and the climate: Results from the Earth Radiation Budget Experiment. Science 243,57-63. Hansen J, Sato M, and Ruedy R 1995: Atmos. Res Ê37, 175-209. Harvey LDD 1995: Warm days,hot nights. Nature 377, 15-16. Hughes MK and Diaz HF 1994: Was there a `medieval warm period', and if so, where and when ? Climatic Change 26 , 109-142. (NB. This is part of a special issue devoted to climate change over the past millenium.) IPCC 1995: Climate Change : The IPCC Scientific Assessment . Cambridge University Press. Lau, N-C and Nath, M. 1987: Frequency dependence of the structure and temporal development of wintertime tropospheric fluctuations--Comparison of a GCM simulation with observations Mon. Wea. Rev., 115, 251-271. Macayeal, D.R. 1993a: Paleoceanography 8, 767-773. Macayeal, D.R. 1993b: Paleoceanography 8, 775-784. Manabe R and Stauffer 19** [**Coupled ocean run; use as citation for regional warming in doubled CO2 vs. Medieval Optimum] McKibbin, B. 1995: An Explosion of Green. Atlantic Monthly 61-83. Mendelsohn R., Nordhaus WD, and Shaw D 1994: The impact of global warming on agriculture: A Ricardian Analysis. The American Economic Review 84, 753-771. Newell, R. E. and J. Hsiung, 1984: "Sea Surface Temperature, Atmospheric CO2 and the Global Energy Budget: Some Comparisons between the Past and Present," in Morner & KarlÚn, eds. [1984]: 533-562. Pierrehumbert, R. T. 1995: Thermostats, Radiator Fins, and the Local Runaway Greenhouse. J. Atmos. Sci. 52, 1784-1806. Pierrehumbert, R.T. and K.L Swanson 1994: Baroclinic Instability Ann. Rev. Fluid Mech. 27, 419-467. Serre-Bachet F 1994: Middle Ages temperature reconstructions; a focus on Northeastern Italy. Climatic Change 26 , **-**. Sloan, L. C. and E.J. Barron 1990: "Equable" climates during Earth history? Geology 18, 489-492. Soden B 1995: [**New J. Climate paper on water vapor intercomparison.] Sun, D-Z, and Lindzen, R. S. 1993: Water vapor feedback and the ice age snowline record. Ann. Geophysicae 11, 204-215. W.R.Kapsner, R.B.Alley, C.A.Shuman, S.Anandakrishnan & P.M.Grootes 1995: Dominant influence of atmospheric circulation on snow accumulation in Greenland over the past 18,000 years. Nature 373 , 52-54 [**CN = Crowley and North. Paleoclimatology. Oxford University Press]