Predicting the Future—and Saving Lives
El Niño, Climate Change, and the Global Ocean Observing System
In previous chapters we have seen the importance of marine predictions in saving lives. Our emphasis has been on predicting an impending danger such as a tsunami, a storm surge, or a rogue wave—a danger that could arrive in a day, or in a few hours, or in the case of a tsunami generated by a nearby submarine earthquake, in less than thirty minutes. There are, however, other predictions that can be for much further into the future—predictions of catastrophes that result from environmental changes that develop over years, decades, centuries, or even longer, changes also produced or greatly influenced by the power of the sea. While perhaps lacking the drama of a huge tsunami wiping an entire town out of existence, or a storm surge flooding a city, or a rogue wave sinking an oil tanker, these more distant calamities can also lead to loss of life and cause serious economic impacts. The two most important examples are El Niño, which occurs every two to seven years, and long-term climate change, especially with respect to the effects of global warming where the most serious effects could possibly show up in only a few decades. Both are global phenomena with worldwide impacts. Both demonstrate the immense power of the sea on a planetary scale and its critical role in driving weather patterns around the world and in modifying our climate. Being able to predict the onset and effects of El Niño and being able to predict the effects of global warming can be of great economic benefit and can save lives by allowing us to prepare for future dangerous conditions. In the past, strong El Niños and dramatic changes in climate have caused millions of deaths.
Two of the largest El Niños in recorded history took place at the end of the nineteenth century, before the name El Niño was known to anyone other than Peruvian fishermen. Those fishermen first used that name for the annual warming of the ocean waters near Peru that happened every year near Christmas, El Niño meaning “the little boy” and referring to the Christ child. But every few years that ocean warming would be especially severe, and eventually the name came to be used only for those occasional extreme warming events. This was long before the El Niño off Peru was recognized as being an early stage of a global climate phenomenon that involves both the ocean and the atmosphere and whose effects are felt all over the world. This phenomenon was later called El Niño–Southern Oscillation (ENSO). Southern Oscillation was the name given to the very slow seesawing of atmospheric pressure over the Pacific Ocean (meaning that pressure is high over South America when it is low over Indonesia, and vice versa). Much later it was discovered that the Southern Oscillation was connected to the very slow rising and falling of ocean water temperatures across the tropical Pacific (El Niño being the phase when there are warm waters in the eastern Pacific near Peru).
The El Niño that began with much warmer waters off Peru in December of 1876 lasted for two years and had deadly impacts around the world.1 The primary commercial fish in Peruvian waters was the cold-water anchovy, which was so plentiful that fishermen could literally scoop them up with their hands and fishing boats could forgo nets and use suction hoses. Hundreds of thousands of tons of anchovy were exported to Europe. When the warm waters of El Niño came, the anchovy disappeared and the Peruvian fishing industry collapsed. The seemingly limitless anchovy population had also supported a seabird population so huge that their droppings (guano) grew into small mountains on islands off the coast of Peru. At that time guano was the most prized fertilizer in the world, and hundreds of thousands of tons were exported to France, Great Britain, and other nations. When the anchovy disappeared, the so-called guano birds, primarily the Guanay Cormorant, died, and the droppings stopped. The mountains of guano shrank, not only because guano exports continued while not being replenished but also because they were washed away by the heavy rains brought by El Niño. One reason the mountains of guano had built up over the years was that Peru is normally one of the driest places on Earth, a desert by the sea, with the only freshwater coming down rivers from the Andes. But during an El Niño, torrential rains came, causing floods that killed thousands of people in Peru and also washing the guano from the islands into the sea. So not only did the fishing industry collapse, but the guano fertilizer business collapsed as well. The Peruvian government, which depended on revenues from these two industries, faltered (and two years later lost a war to Chile).2
But much more terrible effects of this El Niño were still to come. Whereas El Niño brought heavy rains to a normally dry Peru, farther to the west, across the Pacific and Indian Oceans, it brought drought to lands that were normally wet and green. El Niño stopped the rain-bringing monsoons, and in 1877–1879 the resulting droughts caused famine and disease on an incredible scale. The death tolls were staggering—ten million dead in India, between ten million and twenty million dead in China, another million in Brazil, and more dead in southern Africa, the Dutch East Indies (now Indonesia), Vietnam, the Philippines, and Korea.3 Twenty years later, in 1899–1900, a second very strong El Niño again stopped the monsoons and caused droughts in India and China and Africa, killing another eight million in India, another ten million in China, and additional deaths in Brazil and Africa.4
When these two catastrophic famines occurred, no one had ever heard of El Niño, other than Peruvians for their own local event. British scientists in India began trying to understand why the monsoon had failed, and they searched for ways to predict when that would happen again. In the process the Southern Oscillation was discovered, but it would be another sixty-five years before anyone would realize the critical role that the ocean plays in this global phenomenon. Once ENSO was understood, however, historians and paleoclimatologists discovered evidence of numerous other El Niño impacts in the past. Even ancient Egypt had been affected by El Niños. Historians discovered that the pharoahs had their engineers regularly measure the flow of the Nile with instruments now referred to as Nilometers, because the annual flooding of the Nile was so important to Egypt, providing the water necessary for irrigation and the fertile sediments that caused crops to flourish. The flow of the Nile depended on rain falling in Ethiopia, which varied dramatically with the monsoons. When a significant El Niño occurred, the Nile would have a low flow, as shown by Nilometer records, and Egypt would experience drought and deaths and political unrest.5 Even in the twentieth century, political events in Africa were affected by El Niño. The 1972–1973 El Niño caused a severe drought in Ethiopia, followed by famine and turmoil, which contributed to the overthrow of Emperor Haile Selassie. Ten years later, another drought caused by the 1982–1983 El Niño led to the overthrow of the man who had deposed Selassie. During this period over a million people starved to death or were killed during the resulting political chaos. These are but a few historical examples of past effects of El Niño. Scientists have tracked El Niños back 130,000 years using paleoclimate records such as tree rings and corals and cores from the sea bottom, even being able to tell that there were fewer El Niños during the Little Ice Age (approximately AD 1300–1850) and more during the Medieval Warm Period (approximately AD 900–1300).
El Niño is the most famous part of the ENSO cycle, but there is also a period with colder than average water temperatures along the Pacific Coast of South America, called La Niña (“the little girl”). La Niña’s effects on weather around the world are generally the opposite of the effects of El Niño. The strong La Niña that began in December 2007 and continued well into 2008 contributed to the 2007–2008 winter being the coldest winter in the United States since 2001, when the last La Niña occurred.6 This same winter was the worst in fifty years in China’s eastern, southern, and central regions and included a heavy snowfall that stranded 5.8 million passengers throughout China’s railway network a week before the Chinese New Year on February 7.
An accurate prediction of drought and famine due to El Niño could have saved countless lives in the past because stockpiles of food could have been made available to help the starving.7 Today, with global relief organizations and countries willing to help when there is a natural catastrophe (as in the worldwide response to the 2004 tsunami), a similar El Niño–caused drought would probably not lead to millions of deaths, and so the prediction of El Niño would not have as dramatic a benefit as it would have had at the end of the nineteenth century. But knowing that a drought is coming can lead to preparations that would be economically beneficial, including farmers changing to more drought-tolerant crops or increasing crops in other regions not affected by the drought. If El Niño predictions are good enough to predict the locations of future floods, preparation for those floods can save lives. The largest El Niño in the twentieth century, in 1997–1998, had many effects around the world, such as torrential rains in California that caused widely reported mudslides, in which homes slid into the sea. The running joke on late-night TV in 1998 was to blame everything on El Niño. Effects of this El Niño, including the heavy rains across California, were correctly predicted by NOAA’s National Centers for Environmental Prediction six months in advance.8 As a result, overall property losses were a billion dollars less than what they had been for the previous large El Niño in 1982–1983.9 Today fairly accurate prediction of an El Niño six months to a year in advance has become possible using computer models that ingest millions of gigabytes of real-time data from instruments on buoys deployed across the Pacific Ocean.10 But even with this system we still cannot always correctly predict the specific effects of an El Niño for particular regions.
Climate, of course, has been changing on time scales much longer than that of ENSO, with the world going into and out of ice ages every 80,000 to 120,000 years, and with other climatic variations on shorter time scales, all of which involved changes in the ocean and its interaction with the atmosphere. The world has been warming since the last glacial maximum, about 20,000 years ago. Along the way there have been notable climatic oscillations in some regions, such as in Europe, where the Medieval Warm Period had temperatures almost as high as they are today and the Little Ice Age had temperatures generally lower, although they fluctuated wildly. Near the end of the last major ice age, when mile-thick ice sheets covered more than a third of the Earth’s landmass, sea level was 300 feet lower than it is today, exposing major areas of continental shelf and allowing humans to walk from Asia to North America across what is now the Bering Strait. As the climate warmed, melting glacial ice sheets that had reached to the sea pulled back, exposing a coastal route by which humans were able to migrate down the west coasts of North America and South America. The warming climate also produced changes in the Middle East that led to the beginnings of civilization. Later, during the Medieval Warm Period, centuries of drought caused the Mayan empire in Central America to disintegrate. Farther north during this warm period, the Vikings were able to explore islands and coasts throughout the North Atlantic and colonize Greenland, which at that time was indeed green because of the warmer temperatures. The British successfully cultivated vineyards in England. During the Little Ice Age that followed, however, these vineyards disappeared, and the Vikings had to abandon their North Atlantic colonies. The Baltic Sea froze over, and even Mediterranean countries had very cold winters. With shortened growing seasons there were frequent crop failures, followed by famines and many deaths.
Today, of course, the primary climatic concern is global warming. What global warming really means in this case is the increased warming of the globe that has occurred in the last century or two, believed by most scientists to be caused by an enhanced greenhouse effect due to the increased carbon dioxide in the atmosphere during that time period.11 It is critically important to be able to accurately predict the Earth’s climatic response to this higher level of carbon dioxide, which has resulted from two actions by humans. First, enormous amounts of carbon dioxide have entered the atmosphere through two centuries of burning fossil fuels. Second, the cutting down of forests over many centuries has reduced nature’s ability to take carbon dioxide out of the atmosphere. Since the resulting increased carbon dioxide is a greenhouse gas, which traps heat in the atmosphere before it can escape to space, the result has been additional warming beyond the natural warming that would have occurred since the last ice age without the additional quantity of greenhouse gases added by humankind. How much of the increased warming over the last century is due to this enhanced greenhouse effect has been debated. The Earth also responds (chaotically, in mathematical terms) to particular cycles involving orbital motions of the Earth and the sun, so it is also important to be able to predict the Earth’s response to these astronomical cycles.12
The ocean plays a crucial role in climate change and the effects of global warming. It is the greatest solar collector on Earth (since it covers 70 percent of the Earth’s surface), and it stores heat 4,000 times more efficiently than the atmosphere. It has a major role in how carbon dioxide affects climate, since it stores 500 times as much carbon as the atmosphere does and it absorbs up to half the carbon dioxide produced by burning fossil fuels. Its phytoplankton absorbs as much carbon dioxide as trees, grasses, and other plants on land do. And as an important side effect, a gas from these phytoplankton is responsible for a considerable proportion of aerosols in the atmosphere, which become condensation nuclei for cloud formation, an important aspect of climate change that is still not well understood.13 The oceans contain approximately 97 percent of the Earth’s water, although that amount varies as climate changes. During ice ages water from the sea goes into snow that accumulates as mile-high glaciers over vast areas of land, with the result that sea level drops. During the warmer interglacial periods, water that has been locked up in glaciers on land returns to the sea, and sea level rises.
Ocean currents play a major role in transporting heat from the hot equatorial regions of the Earth to the cold polar regions. The Gulf Stream, though produced primarily by winds blowing over the Atlantic, also plays a major role in the density currents involved in the Atlantic’s thermohaline circulation. The Gulf Stream moves warm salty water north along the surface of the Atlantic Ocean. When it reaches the North Atlantic, the warm salty water cools, releasing heat into the northern atmosphere, with the resulting colder salty water then sinking to great depths. This deep water then moves very slowly southward, eventually passing through the South Atlantic and the Indian Ocean on the way to the Pacific Ocean, where it finally rises again to the surface and heads back toward the Atlantic. This process is why the thermohaline circulation is also called the conveyor belt. The whole trip can take between six hundred and one thousand years. Large changes in the strength of the thermohaline circulation appear to correspond to beginnings and endings of glacial periods, while smaller changes seem to correspond to other smaller climate oscillations. Past dramatic weakenings of the thermohaline circulation may have been caused by sudden increases in freshwater in the North Atlantic, while various processes that increase salt transport from the south might have strengthened it.
It wasn’t until the last fifty years that much progress was made in understanding and predicting El Niños or in understanding some aspects of long-term climate change. In both cases, this was possible only after we were able to acquire extensive oceanographic and meteorological observations from around the world and feed these data into large sophisticated dynamic models run on supercomputers. The first understanding of one aspect of the ENSO cycle goes back to 1904 when Sir Gilbert Walker, the British head of the Indian Meteorological Department, analyzed large amounts of data from weather stations around the world and recognized for the first time the seesawing of atmospheric pressure over the southern Pacific Ocean (which he called the Southern Oscillation).14 But it was not until 1957 that the Southern Oscillation’s connection to El Niño was recognized by Jacob Bjerknes, a Norwegian meteorologist at the University of California at Los Angeles working for the Inter-American Tropical Tuna Commission. Bjerknes analyzed huge amounts of oceanographic and meteorological data from the Pacific obtained by sixty-seven nations working together during the International Geophysical Year 1957–1958. He was able to determine that large water temperature anomalies in the tropical Pacific (El Niño) were connected to large atmospheric pressure anomalies over the Pacific (Southern Oscillation), and he proposed the first theory to explain how this ocean-atmosphere interaction worked.15 In the 1970s, Klaus Wyrtki, an oceanographer at the University of Hawaii, took the next step in understanding El Niño when he analyzed sea level data from tide gauges around the Pacific.16 When ocean waters warm, sea level rises because of thermal expansion. When ocean waters cool, sea level falls. Mapping his data, Wyrtki could show that a huge pool of warm water moved from the western Pacific to the eastern Pacific and then moved north to California and south to Chile.17
The scientific story of El Niño goes on to involve large oceanographic observation programs, such as TOGA (Tropical Ocean Global Atmosphere), which by 1994 had seventy instrumented buoys deployed across the tropical Pacific.18 Three years later, data from these buoys allowed the first complete description of an El Niño event and made prediction possible. Eventually, specially developed numerical oceanographic models coupled to weather models would become capable of predicting the onset of an El Niño at least six months in advance. These models can be traced back to the dynamic model first derived by Laplace for use in understanding and predicting the tide (discussed in Chapter 1). The oceanographic models added equations for salinity and heat. The weather models added equations for heat, humidity, compressibility, and other atmospheric effects. To successfully predict El Niño, an ocean model and an atmospheric model had to be coupled, which essentially means they become one model. For other ocean phenomena one could run an ocean model with energy from the atmosphere (in the form of wind stress on the ocean’s surface, or heat to warm the ocean’s surface, or rain to add freshwater at the surface) being input as what are called boundary conditions based on data. Likewise, weather models could be run with boundary conditions that included the heat from the ocean’s surface or the amount of water vapor entering the atmosphere from the ocean. But for phenomena where the ocean and the atmosphere greatly affect each other, so that both are changing significantly, then the two models must be coupled. Billions of data points from across the Pacific Ocean and around the world are used to initialize the model and run it for a while, before it is run into the future (without any data to keep it on track) to make a forecast.
It is still not clear what triggers the beginning of an El Niño, or what causes it to end. During non–El Niño conditions, with low atmospheric pressure over the western Pacific and high pressure over the eastern Pacific, the trade winds blow toward the west along the equator. This pushes warmer surface waters into the western Pacific, where they accumulate as an enormous warm pool the size of Canada, causing a great deal of evaporation that turns into heavy rains over Southeast Asia and the Indian Ocean. At the same time, these westward trade winds are pushing water away from the coast of South America, which causes the upwelling of cold nutrient-rich deep waters into the sunlit surface waters of the eastern Pacific, which in turn causes phytoplankton blooms on which the anchovy feed, as well as dry conditions on land. During El Niño, with low pressure over the eastern Pacific and high pressure over the western Pacific (the other half of the Southern Oscillation), the trade winds weaken or reverse, allowing the waters in the western Pacific to shift to the east. This essentially reverses the situation, so that heavy rains and floods now occur in eastern Pacific areas and droughts occur in western Pacific and Indian Ocean areas. In the eastern Pacific the upwelling stops, the waters warm and have less nutrients, the phytoplankton do not bloom, the anchovy are not in abundance, and the guano birds die. For a large enough event, the warm waters move north to California and south to Chile. If, when conditions reverse again and the trade winds start up, the situation becomes very strong in that direction, so that it moves past an average situation, then the waters off South America become extra cold, which is the La Niña situation.
At the start of an El Niño, the ocean affects the atmosphere and the atmosphere affects the ocean, both in the same direction. The trade winds blowing toward the west are slowed by the warm water in the east and reverse their direction toward the east, but this shift in wind direction then further pushes the warm water to the east. This positive feedback mechanism is possible only along the equator when the water movement is toward the east, because here the effect of the Earth’s rotation is to keep the water movement in alignment with the winds.19 But it is not clear whether the warm water or the trade winds make the first move, and why. Even without understanding the initial triggering mechanism, useful short-term predictions have been possible when there are enough real-time data from the buoys across the Pacific.
Long-term climate predictions also require coupled ocean and atmosphere models, but the complexity of the models and the difficulty in making accurate predictions are much greater than for ENSO. There are few phenomena more complex than the Earth’s climate system, and the greenhouse effect of anthropogenic carbon dioxide is only part of the story. Although the ocean and the atmosphere are most important, all parts of the Earth make a contribution, including the biosphere (the biological effect of all living things on Earth; we have already mentioned forests and phytoplankton), the cryosphere (all land-based ice, such as the ice sheets on Greenland and Antarctica), and the geosphere (the solid earth, including land and tectonic plates, whose effects include volcanic eruptions that put light-reflecting particles into the atmosphere).20 The biosphere obviously plays a critical role in the carbon cycle and in global warming by carbon dioxide and other greenhouse gases such as methane. Another critical effect is the variation in solar radiation (the ultimate driver of the Earth’s climate) over time, especially in terms of the variation in the amount of light that hits particular parts of the Earth. For example, this varies with cycles of 20,000, 40,000, and 100,000 years due to, respectively, changes in the orientation of the Earth relative to the sun, changes in the tilt of the Earth’s axis, and changes in the shape of its orbit around the sun. These three cycles appear to affect the timing of ice ages (and can be identified in paleo sea level records), although it is not understood how such seemingly small effects can have such a profound impact. It probably involves positive feedback mechanisms involving the oceans.
We usually test a model and the theories behind it by seeing whether predictions using that model are accurate compared with data. ENSO events occur every two to seven years, and so after each event we can see how well a model did in predicting it. This is not possible when predicting climate change a century or more into the future. The best we can do to test a climate model is to see if it can predict past climate changes, but then we have a problem with having adequate data to compare against the model predictions and adequate data to run the model. Instrument records go back only 50 to 150 years, depending on the meteorological or oceanographic parameters being measured. Beyond that we must develop proxy data sets (using tree rings, corals, ice cores, cores from the bottom of the ocean and lakes, etc.), from which we estimate the air and water temperatures, sea level heights, carbon dioxide, and other indicators of past climate. These can go back thousands of years (and even hundreds of thousands of years), but being proxies, they are indirect measurements based on things like ratios of particular isotopes of an element, with enough uncertainties that sometimes particular questions cannot be answered. Also, because of the long time periods that must be computed, climate models must be run with coarser time and space resolution than ENSO models, and so all physical processes might not be handled as accurately as we would like.
The most sophisticated climate models have been run with and without including the increasing carbon dioxide levels in the global atmosphere over the last century. Only the models that include the increase in carbon dioxide can reproduce the increase in the global air temperatures measured over that same period. Without the increased carbon dioxide levels, global air temperature still increases as it has since the last ice age ended, but not nearly as much. But while all models generally agree about the rise in temperature averaged over the globe, there is less agreement among climate models on the exact changes that will occur at specific regions of the world. Thus, there may be less certainty with respect to some of the predicted regional effects, many of which would be detrimental to human health if they occur. Of course, floods, droughts, and famines have happened without global warming, as we saw at the end of the nineteenth century when two strong El Niños stopped the monsoons and millions died from the resulting famines. While droughts, heat waves, and floods on an unprecedented scale would certainly justify the concern over global warming, until the models agree and can prove that they will take place on a worldwide scale, the skeptics will use such regional uncertainty as an excuse to ignore global warming. But there are other possible calamitous global effects.
A global effect that would certainly be calamitous if it occurs is a significant and rapid rise in global sea level as a result of global warming, since roughly half the world’s population lives close to the coast.21 As we have mentioned, during the last ice age sea level was 300 feet lower than it is today. Continental shelves now underwater were land 20,000 years ago. (The Port of New York and New Jersey, if it had existed then, would have been more than a hundred miles inland from the Atlantic Ocean.) By roughly 8,000 years ago, most of the 300-foot rise in sea level had occurred due to the melting and retreating of the ice sheets. In more recent centuries sea level rise has been primarily due to the thermal expansion of the upper water column of the ocean as it warmed. Over the last century sea level only rose approximately two-thirds of a foot. However, sea level will rise faster if the Greenland or Antarctic ice sheets began to melt more quickly. The ice sheet on Greenland if completely melted would produce roughly a 24-foot rise in global sea level, while completely melted Antarctic ice sheets would produce a 200-foot rise. It would probably take centuries for complete melting, but it is not necessary for entire ice sheets to melt to produce a dramatic rise in sea level. The ice sheets merely need to move faster toward the ocean and break off into huge icebergs, since icebergs raise sea level the same amount as their melted water would. A critical question is how quickly are these ice sheets moving toward the sea, and could this process speed up? Melted water beneath the ice sheets serves as a lubricant for a more rapid glacial flow toward the sea.22 Based on the latest predictions for the movements of ice sheets on Greenland and Antarctica, estimates for the amount of global sea level rise by the year 2100 range from two and a half to six feet, a rise capable of producing tremendous economic impacts and loss of huge amounts of coastal land for many countries.23 Satellite gravity data indicate that there was a doubling of ice-mass loss from Greenland and Antarctica from 2002 through 2009.24 Findings from coral paleoclimate studies of the last interglacial period (about 130,000 years ago), when sea level was higher than it is today but global air temperature was only a little warmer, may indicate that there are mechanisms for sea level to rise higher than was previously thought possible.25
Global warming could also possibly affect the one-sixth of the world’s population that relies on mountain glaciers or seasonal snow packs for their water supply. As the world gets warmer, the melting of winter snow would occur earlier in the year so that maximum river flows would occur in late winter or early spring rather than in summer or autumn, when the demand for water is highest for agriculture, and unless there are sufficient storage capabilities, that water will be lost to the sea.26 Another likely global effect of increased carbon dioxide levels would be ocean acidification, whose impact on phytoplankton and other global ecosystems is not yet well understood but which has the potential for a variety of serious global effects.27
Whether we manage to reduce global warming or not, accurate climate prediction is critically important. The models must be able to reliably predict the local consequences of that warming, namely, how wind patterns and precipitation might change, where there would be droughts and where there would be floods, where water supplies will be threatened by loss of snow storage areas—the kind of thing that governments must know in order to prepare for whatever changes are going to occur. It is vital that climate models continue to improve and be supplied with needed global data sets from modern global observing systems. We also need more data from paleoclimate studies—in this case it is especially true that we have to be able to understand (and predict) the past in order to predict the future.
We also need to be able to accurately predict what would happen if someone actually goes ahead and implements one of the many proposals that have been suggested for artificially cooling the Earth.28 These proposed geoengineering solutions will have side effects and potentially strong positive feedback mechanisms that we do not understand. Climate prediction models are our only tools to assess the possible dangers. For example, at first look, a controllable method to counteract global warming might appear to be to imitate the way a volcanic eruption cools the Earth. A volcano shoots sulfur dioxide high into the stratosphere, where it forms sulfate particles that reflect sunlight back into space, thus cooling the Earth. But there are also other effects besides the cooling. A year after the volcanic eruption of Tambora in Indonesia (see Chapter 7, note 44) there was significant cooling. The year 1816 was called the “year without a summer”; snow fell in New England in June, and the southern United States had frost on July 4. The cold led to crop failures in the northeastern United States and northern Europe, which resulted in famines and death. The eruption also disrupted the monsoons over India and China. In 1991 the eruption of Mount Pinatubo not only had a global cooling effect but also decreased global precipitation and increased drought.29 Even if we manage to cool the Earth’s temperature using sulfur dioxide injections, the carbon dioxide concentration will still continue to grow higher, and its other effects will not be counteracted—for example, the acidification of the ocean will get much worse. Also, if the injection were suddenly stopped, perhaps sabotaged by a nation that did not like how its climate was being affected, the warming would quickly reach the same high temperature levels, but this rapid change would cause even more damage because there would be less time for ecosystems to adapt.30 We also do not know enough to rule out the possibility that the cooling might go too far. If a positive feedback mechanism is initiated that keeps cooling the planet even when the injection of sulfur dioxide is stopped, we might move into an ice age, which is another type of devastating climate that could cause death and serious economic problems. We need better climate models to fully understand all these possibilities.31
Another proposed geoengineering solution to global warming that has been proposed is to put tons of iron compounds into the ocean to stimulate blooms of phytoplankton and therefore use up more carbon dioxide. The increased phytoplankton also might release an increased amount of dimethyl sulfide particles that would become the condensation nuclei of clouds, thus increasing cloud cover, which could reflect more sunlight back into space. Others have proposed agitating the ocean’s surface to spray salt particles into the air, which can also serve as condensation nuclei for cloud formation, or artificially pumping nutrients from the deep parts of the ocean to the surface to create the same kind of phytoplankton blooms as the iron solution. All these geoengineering solutions were once thought to be too risky to even consider. We would, after all, be intentionally playing with the global climate of our only home. But if an extreme climate catastrophe did seem to be happening, such as if ice sheets on Greenland or Antarctica began to rapidly slide toward the sea, such methods might be tried. We must have a better understanding of the climate system and better prediction models to have a chance of accurately predicting what would happen.
But how sure are we that catastrophic climatic change will occur at some time in the future if we continue burning fossil fuels? There is of course some uncertainty in the climate model predictions, especially at the regional level. But there is uncertainty in all scientific studies, and critics are holding climate change studies to a much higher standard than other scientific work, for a variety of economic and political reasons. What we are trying to do is simply assess a risk (as we would in any risk management application), and then weigh that risk against the costs of trying to minimize that risk. We might start by looking at the possible consequences associated with the positions of the two political extremes in the global warming debate—at the one extreme the global warming deniers and at the other extreme the so-called doomsdayers. If we put a lot of money into trying to slow down global warming and the deniers are correct, the cost may be serious damage to our economy—something we have managed to do already (the 2008 economic collapse) and so something that is not that unusual. If we do not put any money into trying to slow down global warming and the doomsdayers are correct, then the cost will be unprecedented human suffering. The two sides also differ in the way they view the uncertainty. The deniers say that until we are absolutely certain that global warming is happening and is leading to catastrophe, we should not waste any money on that so-called problem. The doomsdayers say that this uncertainty is actually a reason to worry more about global warming, for if we do not fully understand the nonlinear feedback mechanisms at work in this incredibly complex climate system, the problem could be even worse than we think. The extreme of runaway climate change (like happened on Venus) might not even be out of the question. But perhaps comparing the extreme views may not be the best approach.
What actually is the uncertainty in the model predictions? This is a question that has been hotly debated. The last IPCC report said the results were 90 percent certain, but some critics said that seemed high considering the complexity of the climate system.32 In any other area of science, even a better than fifty–fifty chance of something bad happening would be enough to stimulate some type of corrective action. But critics have so far successfully blocked any significant action by saying that the cost of reducing carbon dioxide in the atmosphere, or even of keeping it from rising further, would have a very harmful effect on the economies of the United States and other nations. Today that argument should probably carry less weight. First, because of the high price of oil, alternative forms of energy are no longer that much more expensive. And second, there are other reasons besides global warming for needing to break free of our dependence on oil. Before the economic crash in 2008 the price of oil was so outrageous that the economic and political consequences were becoming obvious to almost everyone, and once the economy began to recover, that situation returned. With 70 percent of the oil it needs coming from foreign countries, including the volatile Middle East, the United States is critically vulnerable to the possibility of oil being withheld. The solution to this national security problem—reduce and ultimately eliminate U.S. dependence on foreign oil, especially oil coming from the Middle East—would also be a big step toward solving the global warming problem.
The ultimate solution to this economic/political/national security problem—and to the global warming problem—is renewable energy, such as solar, wind, geothermal, and several types of ocean energy, such as waves, tides, currents, and ocean thermal energy conversion (OTEC). In these economic times, the production of renewable energy systems has the potential to have the same beneficial effect as did the increased wartime production during World War II. Though Franklin Roosevelt had done everything he could think of to end the Depression, it took the incredible production necessary to win World War II to finally transform the weak U.S. economy into a booming economy. Pumping huge amounts of money into real production, in this case building renewable energy systems, not only would provide the stimulus needed to turn the economy around, but also would give the United States energy security, allowing us to escape our dependence on Middle East oil with all its political ramifications. The critical “side effect” would be the reduced carbon dioxide emissions wanted so badly by those worried about global warming.
Throughout this book we have seen that marine prediction has depended on two important activities. The first has been our acquisition of large amounts of data from instruments designed to measure different properties and characteristics of the ocean. The second has been our efforts to understand the physics that describes the movement of the ocean and then to develop models that mathematically represent that physics. Feeding huge quantities of oceanographic and meteorological data into these models has led to accurate descriptions and explanations of many ocean phenomena. With today’s technology, there are numerous ways to measure ocean parameters not even conceived of a century ago, with observations now made from satellites and thousands of land- and sea-based platforms. We have seen that for prediction it is important to have real-time data, data delivered as quickly as possible to the models that need them. In today’s internet age, such real-time delivery of data is no longer a problem. And with today’s ever-increasing computer power, the numerical dynamic models of the oceans, often coupled with atmospheric models, can continue to grow larger, thus allowing the greater spatial and temporal resolution that improves their accuracy.
We have seen numerous examples of how measurement and modeling have evolved over the centuries. The first ocean measurements began with recorded observations of the tide in the Persian Gulf by Seleucus in the second century BC, but it was not until two thousand years later that tide measurement finally progressed from simple tide staffs to self-registering tide gauges. Those gauges allowed a continuous curve of the vertical movement of the sea’s surface to be drawn, thus showing other water level oscillations in addition to the tide. We then saw these gauges used to measure the size of storm surges produced by the high winds of a hurricane and to detect tsunamis generated by earthquakes or volcanic eruptions. They were able to measure the slowly changing sea level due to the warm waters caused by El Niño and thus were instrumental in developing an understanding of that phenomenon. Because many of these gauges were in operation for more than a hundred years, they provided data on how fast sea level was rising, a major issue in the discussion of climate change and the effects of global warming. Tide gauges, now more properly called water level gauges, showed that oceanographic instruments could be multipurpose. Many nations around the world installed water level observation networks, making them humankind’s first real step toward creating a global ocean observing system.
Wind waves could also be measured by water level gauges, but they were typically treated as noise to be averaged out so as to not bias the measurements of the tides or storm surges or tsunamis. Wind waves ended up being measured more accurately by other methods (for example, with accelerometers on buoys) that could also show the direction the waves were traveling. The sampling was fast enough to allow analyses that produced spectra showing the energy at each frequency. Tsunamis, though measurable by relatively rapid sampling on a water level gauge, also had special instruments devoted to them, such as bottom pressure sensors connected to DART buoys. Other oceanographic instruments were developed to measure parameters such as water temperature, salinity, and the speed and direction of currents. Such data were used to understand El Niño and the ocean’s role in climate change, the latter also requiring measurements of carbon, biological compounds, and other ocean parameters. We have seen how such measurement progressed from individual stations to multiple stations in oceanographic studies, such as TOGA, to permanently installed networks of stations providing data in real time for many purposes. And we have seen the importance of the various kinds of oceanographic data now obtainable from satellites.
The earliest hydrodynamic models began with the work of Laplace when he was trying to understand the tides. Yet it was a type of statistical model made possible by Laplace’s work that led to the harmonic prediction method, the first tide prediction “model”—a method not usable for any other ocean phenomena because only the tide has energy at specific frequencies, determined by the orbital motions of the moon, the Earth, and the sun. But Laplace’s dynamic model grew into other dynamic models that could be used to describe and predict storm surges and tsunamis, and describe and predict El Niño and climate change when coupled to weather and climate models (which also ultimately trace back to Laplace’s work). Wind wave models on the one hand had to deal with short length scales when representing wave generation by the wind but on the other hand had to be of global extent when considering sea swells traveling thousands of miles along the ocean surface. And global weather models were needed to predict the storms that would generate the swells—the waves at any particular location being a combination of locally wind-generated waves and the swells that traveled there from all over the ocean. In virtually every situation, accurate predictions required a global approach, both in the modeling and in the data needed for the models.
So it was no wonder that at some point a true global ocean observing system would be needed. It was just a matter of when the technology would develop enough to make such a system possible. By the end of the twentieth century we had satellites for synoptic ocean measurements and for rapid communication, high-powered computers for ocean modeling and for the storage and analysis of trillions of gigabytes of oceanographic data, new technologies for measuring oceanographic properties both in situ and remotely, and the internet along with other communication advances for the real-time delivery of data and predictions. The Global Ocean Observing System (GOOS) was established as an official entity in 1991 by four international agencies,33 but as we have seen, many of the instrument systems that were integrated under GOOS had been implemented over the previous decades. International programs like TOGA, for predicting El Niño, and WOCE, the World Ocean Circulation Experiment, for understanding the ocean’s role in climate change, had moved the scientific community toward real-time global monitoring, creating a foundation for GOOS.34 A permanent integrated real-time observing system like GOOS (and its U.S. component, the Integrated Ocean Observing System, or IOOS) are now clearly seen as the best way to handle the many interconnected marine phenomena. GOOS and the ocean models that it supports represent the culmination of centuries of marine scientific research and are finally beginning to provide the marine predictions needed around the world.
At the end of the last chapter of The Sea Around Us, Rachel Carson’s 1951 best seller (ten years before Silent Spring), she stated that “even with all our modern instruments for probing and sampling the deep ocean, no one can say that we shall ever resolve the last, the ultimate mysteries of the sea.”35 In the sixty years since she wrote that, our ability to probe and sample the ocean has progressed by orders of magnitude. Some mysteries do remain, including one that many think might threaten life as we know it, but such mysteries no longer appear unsolvable. Perhaps Carson’s statement was made not out of pessimism but merely as a way to extol the sea’s grandeur, that grandeur keeping the sea from being totally understood. But with our growing Global Ocean Observing System and our ever-improving dynamic computer models, we are on our way to solving the sea’s most important remaining mysteries and to being able to predict how the sea will affect us in the future. That does not in any way take away from the grandeur of the sea or decrease our appreciation for the incredible power of the sea.