Plows, Plagues, and Petroleum: How Humans Took Control of Climate

Book Review: Plows, Plagues, and Petroleum: How Humans Took Control of Climate – by William F. Ruddiman (Princeton University Press, 2005/2007)

This is a fascinating book by a renowned retired climate scientist that puts forth some interesting possibilities about how humans have been influencing climate since the advent of agriculture. Ruddiman considers the most important question – By how much are we influencing climate? He does not downplay the current anthropogenic influences on climate which are considerably more than those of the past. He simply shows that there were small but significant increases (and some relative decreases) in man-made greenhouse gases since about 8000 years ago and that they may have significantly affected to climate system. The agricultural effects of increased CO2 and methane involved cutting and burning forests, clearing land for planting, constructing and irrigating rice farms, and tending livestock in large numbers. Burning wood, peat, and coal for heat also had effects. 

Ruddiman goes through the basic sciences of evolution, dating, and human history to preface his study. He notes that both paleoclimatology and paleohistory have similar limitations, being studies of the past. He compares them to crime-solving. His first clue that something might be amiss was that atmospheric methane began to rise about 5000 years ago but the trends suggested that it should have gone down. Then it was a similarly anomalous atmospheric rise in CO2 that began about 8000 years ago. He also goes through the history of life on earth as we know it and paleoclimatology as we know it. He introduces the major climate change triggers from astronomy – the Milankovitch cycles which affect solar climate forcing, triggering icehouse and greenhouse conditions. He goes through everything from plate tectonics to ocean circulation to develop the conditions necessary for the most recent (geologically speaking) series of ice ages and the warmer periods between them (such as now). While the astronomical cycles are the triggers, changing concentrations of atmospheric greenhouse gases, mostly CO2 are the likely mechanisms. In the past, volcanoes have been a major source of CO2 and as in the present chemical weathering of rock is a major sink for CO2. Plate tectonics is powered by the energy of radioactive decay in the earth and as more parent product decays into daughter product the process gradually slows down through time, which means less volcanic CO2 overall through geologic history. He continues co-chronicling climate history on the vast geologic time scale to human history of the last few million years. Though it made for interesting reading it seemed a departure from his main thesis. He does give a temperature chart of the last 5 million years which shows a steadily cooling temperature with many fluctuations that coincides with our hominid history. 

Next he goes in the Milankovitch cycles that trigger climate change. He covers the history of the development of the idea that astronomical and orbital cycles affect climate and trigger climate events like glaciation. In the mid-1800’s Swiss geologist Louis Agassiz was a key figure in determining that there were multiple ice ages. Through the next century many of the features and extents of the various ice ages evident on the earth’s surface would be described and mapped. When radiocarbon dating came about in the 1940’s and 50’s the details were refined much further and paleoclimatology could better be studied. The study of pollen grains could determine the extents of different types of forests in different time periods. The study of tree rings could determine local and regional climatic conditions back through recent time. It was found that not only the sun but the planets and even the moon have gravitational effects on the earth. The larger planets like Jupiter and Saturn (mainly Jupiter) have larger gravitational effects on the earth at certain points in their elliptical orbits. This affects the axial tilt of the earth. The axial tilt of the earth changes slightly in a regular cycle from about 22.2 deg to about 24.5 deg which directly affects how much sun is received at the poles, altering solar irradiance at high latitudes where the sun is always low in the sky. That full cycle lasts about 41,000 years. Small changes in the eccentricity (amount of circularity or elliptical-ness) of the earth’s orbit around the sun cause the earth to be about 3 million miles closer to the sun in one part of the orbital cycle than in the other extreme. This cycle lasts about 100,000 years. This cycle is more irregular than the change in axial tilt which is more predictable and regular. There is also the precession, which is a wobbling effect, which takes about 22,000 years. So that is 22,000 annual revolutions around the sun and over 8 million rotations of the earth (days) that make up the precession effect. Both precession and eccentricity work together affecting solar radiation: changes in eccentricity act as a multiplier on the precession cycle. These combined effects can be summarized graphically and they correlate quite well to known climate cycles. Many scientists were skeptical early on about the effects of the orbital cycles because they thought that the effects in one season would be cancelled out in the opposite season when the earth was tilted oppositely or at an opposite place in its orbit but they did not take into account the intricacies of the earth’s climate system. For instance, the Southern Hemisphere has much more ocean that heats up slower than the predominant land of the Northern Hemisphere. Other differences are between flat land and mountains, deserts and forests, snow cover and non-snow cover. In 1875 James Croll of Scotland made preliminary calculations of changes of solar radiation received by the earth, followed by Ludwig Pilgrim of Germany in 1904. But it would be Milutin Milankovitch who would refine these tedious calculations of the effects of tilt, eccentricity, and precession and predict their effects on climate beginning in 1911.

Next he describes the process of “ablation” whereby snow melts at different speeds based on latitude and heat. Summer melting has a powerful effect on all the snow except at the furthest latitudes, where snow can accumulate, year over year. Thus, Milankovitch concluded that the solar radiation received in summer was the critical factor in determining the growth and melting of ice sheets. Cores containing pollens and cores of ocean basins (where melting ice deposited debris) seem to match Milankovitch’s predictions of a time lagged effect from increased solar radiance to ice melting. Radiocarbon dating could only go back about 30,000 years and pollen grains where only reliable in certain areas and also recent times but ocean sediments with greater deposition rates reached back further and showed correlations with Milankovitch’s cycles. Oxygen isotope ratios in ocean basin cores showed how extensive the ice sheets were as well as aging them. The same ratios in coral reefs work as a “dipstick” of paleo sea levels. In 1976 marine geologists Hays, Imbries, and Shackleton examined oxygen isotope ratio variations of the last 300,000 years and found cycles that matched eccentricity, tilt, and precession in the 41,000 year, 22,000 year, and 100,000 year cycles. They also found that ice volume changes roughly matched Milankovitch’s predicted lag times. New ocean floor drilling in the 1980’s showed that the current cycle of ice ages began about 2.75 million years ago and there have been 40-50 of them since then with the 41,000 year tilt cycle being most correlatable until about 1 million years ago when the 100,000 year eccentricity cycle began to have the most powerful effect on climate. This shift is explained by the gradual overall cooling of the climate so that the 41,000 year and the 22,000 year cycles could no longer melt the snow at the poles. Milankovitch died in 1958 and so never got to see the verification of his theories by climate science.

Dating of lake sediments in North Africa were done to try and determine why the Sahara desert was greened much more, earlier than 5000 years ago when there were lakes and grasslands there. It was found that tropical monsoons were triggered by changes in summer solar radiation at the 22,000 year orbital cycle. The rains saturate tropical wetlands where plants decompose in stagnant oxygen-poor water which also accelerates methane emissions from the reducing environments. Thus orbital changes control monsoon cycles and can also affect greenhouse gas emissions on a small regional scale. Ruddiman sees meteorologist John Kutzbach’s theory of orbital control of monsoons as being as important to climate change as Milankovitch’s discoveries since half of the earth’s land surface is between 30 deg N and 30 deg S latitudes and thus affected by monsoon changes. Air bubbles in the Arctic Vostok ice core gave a history of atmospheric methane changes that matched Kutzbach’s hypothesis quite well. Thus it was concluded that atmospheric methane changes were driven by changes in tropical monsoons which were in turn driven by changes in summer solar radiation as a result of the 22,000 year orbital cycle. The same cycle is also thought to control temporary summer release of methane from natural bogs in Arctic Siberia and Canada. The tropical monsoon cycles possibly also affected human evolution in Africa in terms of available water and causing early hominids and humans to develop the ability to adapt to changing climates, suggest some scientists. Ruddiman thinks, however, that they would have just moved as the climate changed.

Early hominid ingenuity: stone tools, plant ropes, fishing nets and snares, awls to sew clothing, hunting weapons and spear points, and other tools, allowed our ancestors to extend their environmental ranges. By 50,000 years ago humans made it to Australia, likely by boats across the then quite shallow seas and proceeded to hunt species to extinction. Even better evidence ties humans to mass extinctions of large mammals (megafauna) in the Americas around 12,500 years ago. Although some have argued that climate change was responsible for much of the extinctions, the evidence does strongly suggest humans. He cites the evidence of geoscientist Paul Martin in the 1960’s as “absolutely convincing” that it was NOT climate cycles that wiped out the animals, since those conditions occurred many times previously. More recently, there is more than ample evidence from around the world that ancient humans hunted many species to extinction. Recent analysis through population ecology shows that this is very likely. This suggests that notions that ancient humans lived in harmony with nature was not always the case. By around 12,000 years ago human ingenuity reached a new turning point: the discovery and adoption of agriculture, mainly in the “fertile crescent” of the eastern Mediterranean but also in the Yellow River region of northern China.

Farming, notes Ruddiman, is not “natural.” It is the largest alteration of the earth’s surface that humans have ever achieved. The methane cycle, based on orbital changes, is more predictable than the more complex CO2 cycle, he notes. Agriculture is the first instance of human-caused increase of these two greenhouse gases. The warming effect caused by these anthropogenic greenhouse gases counteracted a long-term cooling trend and may have prevented a small-scale glaciation, he suggests. Farming appeared independently in several areas at different time periods. The combo of wild large-seeded cereal grasses (einkorn and emmer wheat, barley, and rye) and legumes (peas and lentils) made foraging in the Fertile Crescent nutritious. The region was also home to the largest concentration of domesticate-able wild animals, mainly goats, pigs, sheep, and cattle. By 6000 years ago farming was practiced all over Europe and most of Asia. Over 8000 years ago selection of wild rice strains began with full farming beginning to occur over the next two centuries in Southeast Asia. In the Americas squash was grown 10,000 years ago and corn, beans, and avocados were cultivated between 9000 and 7500 years ago. Farming spurred civilization and population growth which increased food needs. Farming and population growth went hand in hand. Animal domestication also led to animals pulling plows so that farming could be expanded. Irrigation was yet another innovation, begun around 8200 years ago in Anatolia. The invention of the wheel and the ability to plant cuttings, select fruits, and maintain orchards also aided population growth. Ruddiman gives a good overview and timelines of all these developments and much more. Fairly widespread land clearing occurred in all of the agricultural areas. 

Ruddiman noticed that atmospheric methane concentrations followed John Kutzbach’s theory of orbital control of summer monsoons until about 5000 years ago when there was an anomaly – methane increased when it should have decreased. It continued increasing all the way up through the Industrial Revolution. Methane had followed the orbital controlled monsoon solar radiation predictions for 100,000 years then departed 5000 years ago. There was nothing different noted in the two major world sources of methane: the tropical wetlands and the boreal wetlands. Methane increase from the boreal wetlands was first thought to be the cause but was ruled out when comparing methane in the Greenland and Antarctic ice core bubbles. What could be the new source, he wondered? The most plausible explanation, he notes, is the diversion of rivers to create wetlands all over China to grow rice. At the same time irrigation was occurring in other agricultural areas, also increasing artificial wetlands. While other human activities generated methane such as livestock manure, enteric fermentation in the stomachs of livestock ruminants, and the burning of biomass, Ruddiman points out that the effects of those activities were more population dependent and so would have increased with population increase directly and gradually. Human waste was another source of methane, but relatively small and also tied to gradual population growth. He thinks irrigation was the main cause since when it became well established in China/Southeast Asia the population there was in the tens of millions and large areas were irrigated for rice farming. Concurrently with the development of cities in Sumeria and the Indus Valley areas there were large irrigation projects. He thinks the change in atmospheric methane concentration from 5000 years ago to 500 years ago (from 10 ppb to 100 ppb) is consistent with the growth in human population and corresponding irrigation needs. He goes through some mathematical formulas to show that this is plausible and notes that the full anomaly should be 250 ppb since natural methane was decreasing according to the orbital influenced solar radiation changes. He also suggests that early rice farming was less efficient than the modern form, with far more areas flooded per equivalent yield and also more weeds left to decompose into methane. He notes that from 1950-1990 there were efficiency changes in rice growing that yielded less methane potential per eq. yield. He notes that this idea still should be considered a hypothesis and requires more quantitative data. However, it is quite a reasonable and plausible hypothesis.

The CO2 story, he suggests, is more complicated. Changes in CO2 concentrations are affected by all three orbital cycles. There are other natural CO2 cycles that are not yet well understood. Over the last 400,000 years atmospheric CO2 concentrations have fluctuated between 300 ppm in interglacials and 180 ppm at glacial maximums. The current 400+ ppm concentration due to human activity is unprecedented in recent times although it has been over 1000 ppm at times further in the geologic past. CO2 peaked around 10,500 years ago with the trend expected to continue a slow decrease according to the model based on past interglacials – until about 8000 years ago when CO2 began to increase anomalously. The model predicted CO2 would be about 245 ppm at the beginning of the Industrial Revolution but it turns out to be 285 ppm. That makes a 40 ppm anomalous increase over about 8000 years. That means a carbon release of about 300 billion tons needed to be accounted for. He suggests that cutting of forests all over the earth to clear land for agriculture as well as slash and burn agriculture – all over this long time interval – could conceivable account for that much carbon. Lake sediments and types of pollen in different areas can indicate whether nearby land was cleared. Decreases in tree pollen are indicative. In 1089 William the Conquerer ordered a land survey in England and it was determined that 85% of the land had been deforested. The survey also included the current population. Ruddiman used this link between population and deforestation to develop a quantitative estimate for CO2 increase from human-induced deforestation. Burning of wood, peat, and coal would have added CO2 as well. His numbers show that it is quite plausible that the 40 ppm, 300 billion ton increase in CO2 could have been caused by human activities. Later he would say that humans likely added less CO2 but were aided by positive climate feedbacks to account for all of the additions. Humans had not only taken control of the climate but they began to affect environments in mostly negative ways. Irrigation increased salt in the soil so much that the land became barren after so many years. Animal grazing and cutting down forests for agriculture, fuel, and building encouraged soil erosion. Mud and silt clogged deltas and sea ports. 

The pre-Industrial increases in methane and CO2 which make up the long handle of the “hockey stick” graph are quite likely the result of humans. The resultant warming, says Ruddiman and many others, is actually masked by a long-term cooling trend so that in reality the warming effects of the increased methane and CO2 are greater than depicted. According to models there should have been some increased glaciation in northeastern Canada before the Industrial Age. Perhaps the human-caused warming staved it off, suggests Ruddiman. If he is correct then it is unlikely that glaciation will occur anytime soon with the significant further temperature increases since the Industrial Age even though average July solar radiation continues to drop. Ruddiman and Kutzbach modeled the climate system without any human influence (assuming they quantified human influences correctly) and found that the temperature would have dropped nearly 2 deg C! This would have been enough to initiate glaciation and they think that it would have been initiated in northeastern Canada (Baffin Island), the last place the ice melted from the last glaciation and the first place expected to ice up during any new glaciation. Climate scientists Larry Williams in the 1970’s and John Imbries and his son in 1980 also did climate modeling that predicted glaciation if the climate had not been influenced by humans. Those who study orbital cycles general conclude that glaciation will (or would have) begin within centuries or a few millennia. Many have actually suggested that glaciation is overdue, perhaps long overdue, having been delayed or cancelled by human activity. Some climate scientists suggested that the cooling observed in the 1960’s and 1970’s was the beginning of a glaciation but time proved them incorrect.

Ruddiman’s first paper on this subject came out in December 2003 and was immediately challenged. He goes through the various challenges to his hypothesis and gives responses. The first challenge was that he hadn’t looked back far enough to the climate record of 400,000 years ago but he was lucky enough to find out that the Vostok ice core had barely made it through this interval and confirmed that the anomalous CO2 and methane increases in the last 8000 years were indeed anomalous. He also managed to refute other evidence that suggested the next glaciation was not overdue but far ahead in the future based on the same period 400,000 years ago by positing a much earlier unseen beginning of ice growth then. He was also presented with a challenge in the CO2 budgeting, one that was not so easy to refute. His explanation has to do with ice growth, feedbacks, and dust but seems a bit unclear. The human influence on CO2 only accounts for about 1/3 of the observed CO2 increase if natural decreases of CO2 in previous interglacials are taken into account, says his detractors. However, Ruddiman is fairly confident of his logic that positive feedbacks accelerating due to human influence account for the extra CO2 observed. 

Around 2000 years ago there was an anomalous drop in atmospheric CO2 and subsequently there have been fluctuations. Ruddiman examined all known natural fluctuations and did not find matches in the data. Since the first anomalous increase that departed the CO2 trend from natural predictions was likely caused by humans he began to look for human causes for the anomalous decrease and the subsequent fluctuations. His tentative conclusion is that it was plagues that killed tens of millions of people in relatively short time periods that were the culprit. He even suggests that the widespread deaths in the Black Plague may have contributed significantly to the drop in CO2 that may have triggered the cooling known as the Little Ice Age from about 1250-1900, with most of it from 1600-1900. He goes through all the evidence and local effects of the Little Ice Age. Volcanic explosions likely accounted for some of the fluctuations and perhaps some natural climate cycles such as those triggered by sunspots did as well but this does not explain much of the trend. These analyses do indicate that the global climate system is complex and not easy to predict, even going back. We can know greenhouse gas concentrations from ice core bubbles, climate variations from tree rings, and pollen, and temperatures from various proxies. 

He examines the potential past effects of war, famine, droughts, and plagues (basically, the four horsemen of the apocalypse). The success of agriculture led to higher populations in smaller areas and in closer quarters with livestock which led to pandemics. Poor sanitation and refuse with attending rodents also contributed to the spread of pandemics. Based on other studies of past populations and the effects of pandemics on those populations, he basically mapped out all the plagues and estimated their effects, particularly those from the various bubonic plague outbreaks from Roman times through the Middle Ages. He matches the extended Roman outbreaks at intervals of 10-20 years likely due to immunity and new population coming along without immunity – to the first anomalous CO2 drop. The Black Plague in the 1300’s and 1400’s was the biggest outbreak and matches the CO2 drop in one of the ice cores but only partially in the other. The worst pandemic given here is after the Europeans entered the Americas and included a whole host of maladies: influenza, measles, mumps, typhus, whooping cough, viral hepatitis, scarlet fever, diphtheria, cholera, and bubonic plague. These were all introduced to a population that had no immunity. Historians once thought that Native American populations were about 10-20 million but more recent estimates indicate there may have been 50-100 million, with 85-90% dying after contact. If that was 50 million deaths it would have been 10% of the world population at the time. The time period of contact is given as 1500-1750 and this does give a good match to a CO2 drop from the ice cores. The main reason for the CO2 drops, he suggests, was not population drops but the return of agricultural land to forest. He goes through many historical records to try and document post-plague reforestation. He did calculations based on the data of others and concluded that the reforestation along with less deforestation and less coal and wood burning due to population drop could account for the 4-10 ppm drops in CO2 concentration. He also goes back to Baffin Island to show that the Little Ice Age did show a partial resumption there of glaciation. This may have been due to the threshold of glaciation being reached due to pandemics lowering the CO2 and subsequent temperatures. He also makes the important observation that if his hypotheses about past human effects on climate are true and accurate as he presents them, then natural variations of paleoclimate from the time of humans needs to be revised – the human effects need to be subtracted out in order to see the natural effects. Basically, this considerably complexifies the study of past climate since the advent of humans.

Since the advent of the Industrial Age humans have come to influence the environment quite drastically. We are now the major environmental force on earth, says Ruddiman. Global average temperatures have increased by more than 0.7 deg C since the advent of the Industrial Age due to humans. Our impact has exceeded that of nature. The pre-Industrial greenhouse gas increases have led to a similar increase in global avg. temps with a much smaller amount of gases because they were emitted over several thousand years and the response time of the global climate system is thought to be on the order of hundreds and thousands of years – being particularly slow to affect the oceans. The “whole” atmosphere heats up much faster than the “whole” ocean. One issue with climate science is in determining the average response time of the entire climate system and calculations of climate sensitivity change with different response times. Response times for land are hours to months. Response times for ice sheets are millennia. Response times for the ocean are months to decades for the polar surface layer and for sea ice, and centuries for the deep ocean. The global average response time is decades. These are all estimates and not easy to determine with any great accuracy. It is one of several significant uncertainties with climate science. Oceans make up 70% of the earth’s surface, land 30% with ice sheets currently making up 3% of the land surface and 27% ice-free land. 4% of the ocean also has seasonally fluctuating ice cover. He gives a good explanation of thermohaline ocean circulation and how heat is distributed in the ocean. Estimates of the overall avg. ocean response time are about 25-75 years and estimates for the avg. response time of the whole global system are about 30-50 years – but even these estimates are not very certain. The big question is: How much global warming is still in the pipeline? In the coming years this will become more known.

Another influence effect is the amount of cooling aerosol particles in the atmosphere, mainly SO2, that have helped to cool it. Sulfate content has been measured in ice cores and shows the anthropogenic increase in these aerosols. One can also see the effect of the 1970’s Clean Air Act in those ice cores. Aerosol effects were probably negligible in pre-Industrial times due to lower population and more significantly, lower smokestack height for the fires for home heat and forges. Other aerosols like black carbon, or soot, have more uncertain effects. Forest fires and slash-and-burn agriculture are a major source as is burning of fossil fuels. Some estimate that industrial aerosols have cooled the climate by about 20% - or cancelled out 20% of the warming that would have occurred without their presence. The response time of removing aerosols, by burning less coal for example, would only be a few weeks.

Next he talks about oil and the specter of oil depletion and continued U.S. reliance on foreign oil. Much of this discussion is outdated due to the unforeseen advent of the fracking revolution and other innovations in oil and gas exploration and production. His prediction that oil availability will decline and then increase coal usage is not warranted with more oil available, much more gas available, penetration of renewables, and perhaps new calls for nuclear. 

He goes through the various IPCC global warming scenarios of the time (2005/2007). Two issues are most important: How much GHGs we put in the atmosphere and how sensitive is the climate. In determining sensitivity there are significant uncertainties: clouds, aerosols, deforestation and reforestation, response times, and carbon budgeting. Estimates of sensitivity to doubling atmospheric GHGs vary considerably – basically from about 1.5 deg C to over 3 deg C. Some climate change skeptics suggests even less climate sensitivity, closer to 1 deg C for doubled emissions. The effects of global warming are also difficult to predict with any certainty. Warmer temps mean more water is evaporated and there will likely be more rain and rain of more intensity will occur more often. Regional flooding and droughts are also predicted based on the effects on regional climate systems. Long-term effects of global sea level rise are predicted but estimates and timings vary. Melted ice will affect albedo – how much solar radiation is reflected back into space, creating a positive feedback for increased warming. 

Ruddiman gives a timeline of phases of climate change when humans entered the picture. Phase 1 is before 8000 years ago when humans had a negligible effect on climate. Phase 2 is from 8000 years ago to 200 years ago when humans had a small but significant effect on climate. Phase 3 is from 200 years ago to 200-300 years into the future where the effects of the Industrial Revolution and the time responses are all accounted for (with the exception of the full responses of ice sheets). This also includes our attempts to mitigate climate change to some extent. Phase 4 is after 200-300 years from now when fossil fuel depletion is immanent assuming we are still using them or when renewable and clean energy sources are widely used. Theoretically, even after we reduce GHGs from burning fossil fuels there will still be anthropogenic methane from rice farming, landfills, and livestock.

Ruddiman claims complete neutrality in any political debates about climate change. He claims his approach is strictly scientific. He advocates for balance on the issues so as not to negatively affect the ongoing science. He sees extremists on both sides as most problematic. He faults both alarmists and deniers for media distortions of climate change effects. He acknowledges some benefits to climate change and global warming but also thinks the potential harms outweigh those potential benefits, which is reasonable. He also acknowledges that humans are not likely to accept draconian measures to reduce global warming. He favors investing in technologies to mitigate climate change but so far those technologies offer little power and much cost and as well there is some stigma associated with geoengineering. He makes the interesting observation that one could argue that people in the Iron Age and the Late Stone Age had a greater per capita impact on the environment than do people today. Industry advocates often argue that the earth can quickly heal any impact we can put on it but this is quite clearly not the case. Ruddiman’s research has been used to bolster both sides of the debate. He notes the unscientific nonsense being propagated on the pro-industry side. He is worried about the disinformation coming from both extremes of the debate and the lack of support for the scientists. He also worries about those scientists who take strong opinions in such debates. He notes that industry funds more biased science than environmental groups. I say however that since most of us acknowledge that is the case it could make the biases on the environmentalist side bigger as they try to make up for the imbalance. There is also the belief among many of the alarmists that we have a responsibility to affect policy, even lying if necessary, as a moral obligation. The late climate scientist Stephen Schneider openly advocated such an approach. 

Ruddiman sees resource exhaustion as the most important environmental problem in the near-term (next 50-75 years) than climate change. Limits to fossil fuels, minerals, groundwater resources, surface water resources, topsoil and other non-renewable resources or those that renew too slow in certain areas will likely be more problematic, especially if population continues to increase at currents rates.

In the Afterword to the latest edition he notes some new developments. One is new ice cores reaching back 800,000 years that confirm that the CO2 and methane departures from the trends predicted by orbital cycles are indeed unique to the Holocene – where humans likely influenced their concentrations. He also notes that climate models need to be adjusted to include his scenario of modest anthropogenic GHGs in the last 8000 years. If not adjusted the initial assumptions might well be incorrect and thus the whole foundation of the modeling. Another critique against his model has been the correlation between land use and population, which some consider to be linear. He invokes many studies which say otherwise and notes that farming was inefficient and wasteful back then and burning to clear land was widely practiced. Per-capita land-use was much higher in the past, he says. Farming styles also involved farming land for short periods and moving on to clear other land for new plots. He points out that CO2 emissions would depend on the rate of land clearance rather than the total amount. The amount of methane produced by rice paddies is not easily quantifiable, he admits, since the sizes and amounts of rice paddies and the population are not known. However, it is known that the land where rice is grown in China was cleared effectively quite early, suggesting a large population. He conceded to critics that deforestation itself would not have led to the 35-40 ppm increases in CO2 but only about 10 ppm. However, he wrote in a subsequent paper that the positive feedbacks from both deforestation and coal burning may have been responsible for the balance. The effects of the increased methane reacting to form additional CO2 would have added to it as well. The main feedback, he contends was the slow responding deep ocean. This uncertainty among climate scientists shows the uncertainty inherent in carbon budgeting. With his colleagues he also ran climate model simulations to try and determine whether the early anthropogenic GHGs (or the more recent ones) prevented a glaciation. The results were positive but not conclusive. More recent detailed ice core data has also corroborated his ideas about pandemics and the lulls between them in affecting CO2 emissions as the matches turned out to be very good. Improved knowledge about per-capita forest footprints and population density would be needed to further firm up his case there. The fact that the mass deaths of natives in the Americas resulted in a permanent population drop and correlates to a significant CO2 drop also helps his case. He also notes that new information about past climate cycles still suggests that natural variations in CO2 and methane concentrations unlikely account for the anomaly that he says was introduced by humans.

This was a fascinating book and a great scientific detective story. Even though his case is not at all conclusive, it does need to be contended with and it adds significantly to the climate story. Unfortunately (or fortunately depending on who you ask), it also underscores the uncertainty in predicting the amounts of future climate change and the effects they may cause. Even so, he does acknowledge that the case for Industrial Age anthropogenic climate change and its predicted affects has gotten stronger since the book was first published.