Chemical & Engineering News,
November 27, 1995

Copyright © 1995 by the American Chemical Society.

Climate Observations Substantiate Global Warming Models

Bette Hileman, C&EN Washington

Rising atmospheric carbon dioxide
levels and receding alpine glaciers
support projected global temperature increase

Many people believe that a great deal of controversy surrounds the science of global warming. In reality, however, scientists in the field do agree on many aspects of global warming. For example, on the basis of a variety of evidence a consensus is emerging among researchers that human beings, primarily through their burning of fossil fuels, are already perturbing Earth's climate - defined as weather averaged across years and large regions.

This consensus, and the evidence that supports it, are documented in the United Nations Intergovernmental Panel on Climate Change's (IPCC) latest report on the science of global warming. The 2,000-plus page report, written by about 500 scientists and reviewed by about 500 other experts, will be released next month. IPCC was established in 1988 under the auspices of the UN Convention on Climate Change to review the science of global warming and to advise some 70 countries on ways to mitigate and prevent it. IPCC issues a major report every five years. The report will be available at the Office of the U.S. Global Change Research Program, Washington, D.C.

Global mean temperature has been on the rise since 1880

Climate experts agree that the average global air temperature has risen 0.3 to 0.6 Celcius over the past century. This finding is substantiated by other indicators - accelerated melting of alpine glaciers, a sea-level rise of 10 to 25 cm over the past 100 years, and coral bleaching caused by anomalously high sea-surface temperatures - that are all consistent with the increase in global air temperatures. And according to present indications, the average global temperature in 1995 is likely to be as high as or higher than in any year since record keeping began around 1860.

These experts identify a number of other changes that have occurred in global and U.S. climate, some or all of which can be attributed to global warming. Nighttime temperatures have generally increased more than daytime temperatures. Climatic variability, or the frequency of extreme events, has increased in some regions, although it is not known whether these have risen on a worldwide scale. For example, heavy rainfall events in the U.S. have increased in intensity, and behavior of El Nio - the warming of the eastern equatorial Pacific that sometimes brings severe droughts to some regions and heavy rains to other regions around the world - has been unusual since 1976. These occurrences fit well with complex mathematical models of global climate change.

Almost all specialists agree that without drastic steps to curb greenhouse gas emissions, the average global temperature will increase 1 to 3.5 Celcius during the next century because effective levels of carbon dioxide are expected to double sometime between 2050 and 2100. This temperature range results from varied economic and population projections as well as climate sensitivity to greenhouse gases.

Even with a change of 1 Celcius, the global rate of warming would be greater than it has been at any time in the past 10,000 years. Only a few experts expect the atmosphere to warm less than 1 Celcius by 2100, and virtually no scientist who has studied the issue expects global temperatures to decline during the next century. Moreover, the warming is predicted to continue, reaching much more elevated temperatures over the next several centuries, unless bold measures are taken to reduce greenhouse gas emissions.

Even a 1 Celcius change would be significant. During the so-called Little Ice Age, a period lasting from 1500 to 1850 that was marked by extensive glacial advances in almost all alpine regions, the global temperature was only about 0.5 Celcius lower than it was in 1900.

Temperatures before the advent of a global thermometer network are determined from a combination of ice core, ocean sediment, tree ring, fossil, and other geologic data. Ice cores drilled in Greenland and Antarctica and on some alpine glaciers give the most useful data. The annually deposited layers in the ice cores contain bubbles of air trapped at the time the ice was formed, providing a mechanism to determine atmospheric carbon dioxide concentrations. The layers in the ice cores also contain a record of the ocean's temperature in the form of the ratio of isotopically "heavy" water (enriched in 18O) to "light" water in the ice. When the world's temperature is higher, more heavy water is evaporated preferentially. From the ocean temperature estimate, the air temperature for huge regions can be estimated.

Higher temperatures in Antarctica have led to disintegration of some ice shelves

There is also a general consensus that higher temperatures projected for the next century will cause more frequent and intense heat waves, wide-scale ecological disruptions, a decline of agricultural production in the tropics and subtropics, and continued acceleration of sea-level rise.

One broad area of agreement, as noted in the IPCC report, is that levels of greenhouse gases in the atmosphere, primarily carbon dioxide, methane, nitrous oxide, ozone, and halocarbons, have grown significantly since preindustrial times. During this period, the CO2 level has risen 30% to nearly 360 ppm, methane 145% to more than 1,700 ppb, and nitrous oxide 15% to more than 300 ppb.

Implementation of the Montreal Protocol on Substances That Deplete the Ozone Layer has led to a slowdown in the growth of the atmospheric level of some halocarbons. The level of chlorofluorocarbon-11 has plateaued and that of methyl chloroform has declined. Ozone has increased in the troposphere of the Northern Hemisphere but has been somewhat depleted in the stratosphere, primarily in midlatitude and polar regions. The rates of buildup of CO2 and methane slowed in the early 1990s, but now both are increasing again.

There is also broad agreement that stratospheric ozone depletion and aerosols - small particulates consisting of mainly sulfates from the burning of fossil fuels - act to cool Earth. However, aerosols are different than greenhouse gases because they have very short lifetimes (mostly less than two weeks) in the atmosphere, and rather than causing uniform global cooling, they affect primarily the subcontinental pattern of climate change. Because of their short lifetimes, they cannot build up in the atmosphere and therefore, in the long run, cannot offset much of the warming from greenhouse gases. The loss of ozone in the lower stratosphere has a net cooling effect because ozone absorbs incoming solar radiation.

However, these cooling effects would be more than offset by the expected doubling (over its preindustrial level) in the next century in the level of CO2 in the atmosphere. Because many of the greenhouse gases remain in the atmosphere for a long time (CO2 and nitrous oxide persist from decades to centuries) their radiative forcing - their tendency to warm Earth - persists for periods that are long compared with human life spans. Even if nations decide to stabilize the global CO2 level at a relatively high 1,000 ppm, CO2 emissions eventually would have to be cut below 1990 levels because CO2's atmospheric lifetime is so long.

The basic theory behind the greenhouse effect is well established. Natural greenhouse gases in the atmosphere, primarily water vapor and carbon dioxide, raise Earth's average temperature about 33 Celcius higher than it would be if these gases were not present. These gases allow solar radiant energy to pass through the atmosphere to be absorbed at Earth's surface, but trap in the lower atmosphere much of the radiant heat emitted from the surface back toward space. Since the greenhouse effect is a permanent part of the climate system, warming from higher than natural levels of greenhouse gases should be called an "enhanced greenhouse effect."

The current scientific debate surrounding global warming focuses on how sensitive Earth's climate is to an enhanced greenhouse effect and whether the resulting feedbacks will strengthen or diminish the warming from elevated levels of greenhouse gases.

To determine how enhanced levels of greenhouse gases affect Earth's climate, modeling researchers employ three-dimensional numerical models that calculate global temperature and precipitation at specified levels of CO2 and other greenhouse gases in the atmosphere. The families of the more complex models are atmospheric general circulation models and atmospheric general circulation models coupled to ocean general circulation models of comparable complexity.

The atmospheric models were developed from those used for weather forecasting. "It's a fairly well-developed science," says Jerry D. Mahlman, director of the National Oceanic & Atmospheric Administration's (NOAA) Geophysical Fluid Dynamics Laboratory in Princeton, N.J. "The models use the same mathematics that drive today's numerical weather and hurricane prediction models."

Malhman: quantitative predictions difficult Hansen: models closely match real world

Conceptually, the atmospheric models divide Earth's surface into rectangles or grid points roughly 500 km on a side, within which cells are stacked about 20 layers deep. The flow of atmospheric gases from each of these cells into adjacent cells is calculated.

Equations governing the transfer of electromagnetic radiation through a heterogeneous gaseous medium then relate incoming solar radiation with changes in the content of each cell and the amount of radiation reaching Earth's surface. The exchange of radiative energy and water between the atmosphere and various surfaces - such as snow, ice, oceans, clouds, and vegetation - is included, as is a simple representation of the upper ocean. Supercomputers are used to solve the equations that estimate wind patterns, temperature, sunlight, relative humidity, and precipitation for each grid point on the globe.

Over the past five years or so, several atmospheric general circulation models have been coupled to an ocean general circulation model. The ocean model represents various layers of the ocean to its full depth and circulation among and between those layers. When cold high-density saline water sinks to the ocean depths in polar regions, the ocean acts to slow down the rate of warming in the atmosphere by storing heat contained in salty surface water that moves from the tropics to sink in polar regions. Consequently, a coupled ocean-atmosphere model is able to simulate the lag in greenhouse warming caused by deep-ocean circulation.

Mahlman says that despite the criticisms leveled at the models, they work well in many respects. "They capture a lot of the fundamental physics of how the atmosphere works, how the daily temperature range works, how the seasonal variations from summer to winter work, how ice-age climates work, and how natural variability works," he says.

The models simulate the large-scale features of the current climate reasonably well, and they have accurately reproduced some climate features of the distant past. They also predicted very well the amount of cooling that resulted from the Mount Pinatubo eruption in mid-1991. The sulfates and other aerosols from Pinatubo, a volcano in the Philippines, cooled the climate noticeably in 1992 and 1993 almost exactly as predicted in the model run by James E. Hansen, director of the National Aeronautics & Space Administration's Goddard Institute for Space Studies, New York City.

When the models account for the cooling from aerosols, they also reproduce fairly accurately the global warming that has taken place over the past century. Even more important, Hansen says, when aerosols are included the geographic pattern of temperature changes more closely matches the changes in the real world. These tests of the models have given climatologists more confidence in their results.

The models all project that climate change, once begun, will continue for hundreds of years. Taken together, they indicate that temperatures will rise 1 to 3.5 Celcius by 2100, with an associated sea-level rise of 15 to 95 cm. The current temperature projections for 2100 are somewhat lower than those in the 1990 IPCC report, primarily because the models now account for the cooling effect of sulfates and other aerosols. The models indicate that warming in general will be greater over land than over the oceans and that the maximum warming will occur in high northern latitudes in winter. But in the high-latitude Southern Ocean (Antarctic Ocean) and the northern North Atlantic, there will be little warming, perhaps even a cooling due to the effect of changes in ocean circulation.

Furthermore, most models project that with global warming, the increase in mean surface temperatures will be more pronounced during the cold season; that precipitation at mid to high latitudes will increase, especially during the cold season; that droughts will be more severe and longer lasting, particularly during the warm season; that nighttime temperatures will increase more than daytime temperatures during the warm season; that a greater portion of warm season precipitation will come in heavy showers or thunderstorms rather than in gentler, longer lasting rainfalls; and that the day-to-day variability of temperatures will decline for mid to high latitudes. These are changes that are now being observed, if not globally, at least in many regions.

But after many decades, Hansen says, "models show that daytime warming will be almost as great as nighttime warming."

Thomas R. Karl, senior scientist at NOAA's National Climatic Data Center in Asheville, N.C., has analyzed weather data for the U.S. (excluding Alaska and Hawaii) over this century to see if the observed changes are consistent with what models predict for global greenhouse warming. Karl created what he calls a greenhouse climate response index (GCRI), which is a measure of how well climate data fit what models indicate for global warming. GCRI is the arithmetic average of four indicators: the percent of the U.S. with much above normal minimum temperatures, the percent of the U.S. with much above normal precipitation during the cold season, the percent of the U.S. in extreme drought in the warm season, and the percent of the U.S. with a much greater than normal proportion of precipitation derived from extreme one-day events.

Karl found that since 1976, GCRI values have been higher than the average GCRI for previous years in the century, meaning that "the late-century changes in the U.S. climate are consistent with the general trends anticipated from a greenhouse-enhanced atmosphere."

"There is only a 5 to 10% chance that the increase in GCRI results from natural variability," he says, which means there is a 90 to 95% chance that the U.S. climate is already being affected by enhanced levels of greenhouse gases.

Although the country's contiguous states cover only 2% of the globe, the changes in this region are similar to what has happened in many other countries in the Northern Hemisphere, Karl says. Scientists are now calculating precise GCRI indexes for these countries.

One of the more important uncertainties in climate modeling is how the cloud system reacts in response to increases in the levels of greenhouse gases. In general, high clouds act as a greenhouse and warm the climate, while low clouds, by reflecting sunlight back to space, tend to cool the system. Overall, clouds averaged together globally now have a net cooling effect of - 15 watts per square meter, says Jeffrey T. Kiehl, senior scientist at the National Center for Atmospheric Research, Boulder, Colo. With enhanced greenhouse gases, clouds could change in such a way that they cool Earth more than they do today - in other words, greenhouse warming could result in more low-level clouds - or they could change so they cool Earth less.

Another drawback of the current climate models is their inability to project climate for small regional areas. This is partly because they provide average readings for large areas of Earth's surface. Finer resolution would require much more computer power. But when it becomes possible to obtain finer resolution, it will become necessary to incorporate into the models much more information about land cover, land use changes, and vegetation. "When you get to smaller scales, the uncertainties in land surface processes and their physics and biology start to come back to haunt you," Mahlman says.

A boulder at the western margin of the Quelccaya Ice Cap in the tropical Andes or Peru in 1978. The same boulder in 1995 - the extensive retreat of the ice cap in that area has been caused by regional warming. If the current warming trend persists, the ice cap could be in danger of being lost.

In the long run, to gain more certainty in global climate predictions, much more needs to be known about how global warming will affect the biosphere, especially terrestrial carbon storage. In general, warming is expected to increase carbon storage in areas where moisture and nutrients are sufficient. In those areas, trees will grow faster. But over decades to centuries, climate change will also alter the global distribution of ecosystems. For example, forests would spread into tundra and could increase the warming and the carbon storage there. But trees could die rapidly in other areas that are subject to drought, and their deaths could introduce a large pulse of carbon into the atmosphere.

There has been a lot of speculation recently about whether more frequent hurricanes and more intense and longer lasting El Nios are related to global warming. "Until our models become a little more certain, it's difficult to conjecture whether hurricanes would increase or decrease with global warming," Karl says. "On a theoretical basis, there has been some work suggesting stronger hurricanes," he adds. A warmer sea surface is the primary feature of global warming that might cause more significant hurricanes, he explains, but ocean circulation changes may counter the effects of this added warmth.

Since 1976, El Nios have been stronger, more frequent, and more persistent than they were earlier in the century. "Some of the models suggest that stronger, more frequent El Nios would be a tendency in a warmer world," Karl says. But there is no consensus in the scientific community about how these would change. Scientists are doing a lot of research to see if they can establish a cause-effect relationship between global warming and the change in the pattern of El Nios.

In the western Caribbean off the coast of Belize, the Cook Islands, and in the Philippines, massive coral bleaching has been observed since 1983. Coral reef bleaching results from the expulsion of symbiotic zooxanthellae algae from the coral reefs. The algae provide reefs with most of their color, carbon, and ability to deposit limestone.

Raymond L. Hayes, professor of anatomy at Howard University, Washington, D.C., has investigated coral bleaching and found that it occurs when the ocean temperature exceeds 30 Celcius for more than two weeks as shown by NOAA satellites. This is only about 1 Celcius higher than the normal maximum of 28 to 29 Celcius. Once bleached, coral does not receive adequate nutrients or oxygen, Hayes says. If water temperatures return to normal the following year, coral reefs may recover, but if the thermal stress is repeated year after year, the reefs may die.

"I consider coral bleaching on tropical coral reefs as an early warning of detrimental changes attributed to global climate change," Hayes says. El Nio events could be the cause of some coral bleaching, he explains, but if Earth continues to warm from greenhouse gases, much more bleaching can be expected.

To enlarge image.

Carbon dioxide concentrations over the past 1,000 years determined from ice core records (shown as symbols) appear to have fluctuated little until 1850. Since 1958, air measurements (shown as purple line) taken at Mauna Loa, Hawaii, have supplemented the ice core data. The smooth black curve is based ona 100-year running mean. The inset of the period from 1850 onward shows CO2 emissions in gigatons (billions of metric tons) per year attributed to burning fossil fuels (shown as a blue line).

Another graphic indicator of a warmer globe is the accelerated retreat of alpine glaciers. "In the tropics, every glacier that we have any data on is retreating," says Lonnie G. Thompson, professor of geological sciences at Ohio State University, Columbus. "And where we have time-lapse data, the rate of retreat is accelerating." In Venezuela, three glaciers have completely disappeared since 1972. In the temperate zones, the majority of glaciers are retreating, even as a few do continue the traditional cyclic pattern of advance and retreat.

Peru is particularly concerned about the accelerating melting of glaciers in the Cordera Blanca region of the Andes, Thompson says. The glaciers there provide irrigation water to the coastal desert and provide water for rivers that are dammed for hydroelectric power. Loss of glaciers would harm both power production and agriculture for the country.

It used to be thought that as climate changed in the past - when Earth came out of ice ages, for example - it happened very slowly. But over the past decade, and especially during the past five years, evidence has been accumulating that climate sometimes has made sudden jumps as regional temperatures shifted 5 to 10 Celcius over less than two decades. Measurements of oxygen isotopes and dust in Greenland ice cores and studies of glaciers in the Chilean Andes and in New Zealand's Alps show that some of these sudden temperature shifts were felt globally.

Wallace S. Broecker, a geochemist at Lamont-Doherty Geological Observatory at Columbia University, believes the sudden shifts were caused by changes in the ocean circulation, primarily in the North Atlantic. He speculates that the engine for these leaps was an abrupt change in deep-ocean circulation - what he calls the ocean's great conveyor system - which is governed by the temperature and salt content of the water. This deep-current system moves 20 times more water than all the world's rivers combined. For example, in the Atlantic, warm salty upper waters flow northward reaching the vicinity of Greenland, where the Arctic air cools them and they sink to the ocean depths.

Broecker hypothesizes that as Earth was coming out of the last glacial period, water from the fast-melting Laurentide ice sheet that covered northern North America came down the St. Lawrence River into the North Atlantic. This water shut down deep-water convection in the ocean water between Norway and Iceland that takes saline water at the surface to the depths as it cools. Heat energy carried by the Gulf Stream is the major reason northern Europe is warm compared with other regions at the same latitude. So a shutdown in this circulation would plunge Europe into sudden cold and London would have a climate like Siberia. "Even though a shutdown of the Gulf Stream is a low-probability event," Broecker says, "its consequences would be so massive that it is something we should think about."

There is some evidence that this North Atlantic circulation has begun to slow. Compared with the 1970s, the salinity of the Norwegian Sea has decreased and deep-water production there has virtually ceased. However, because there are no records of the extent of deep-water production before 1970, there is no way to tell whether this slowdown is part of a natural variation in the climate or whether it is caused by greenhouse warming.

There is a broad perception that the science of global warming is much less certain than the science of stratospheric ozone depletion. However, Mahlman, who has done research in both areas, says the level of uncertainty surrounding the ozone problem is not much different than it is for global warming.

What distinguishes the two problems is the number of responsible entities. "For ozone depletion, you can blame about 12 companies, more or less," Mahlman says, "but for global warming you can blame about 5 billion people." Consequently, because it is much more difficult to do something about global warming, "people demand more certainty than they required of the science of ozone depletion."

More research is needed but is unlikely to be funded, at least by the U.S. In fiscal 1995, federal agencies spent a total $1.8 billion on global change research. A somewhat larger amount was requested for 1996, but Congress has proposed substantial cuts in the programs of some agencies. For example, the House passed a 25% cut in NASA's Mission to Planet Earth, a major part of its research effort. For NOAA climate and air quality research, the House passed a 41% cut in the request, and the Senate proposed a 33% cut. At press time, it was still not clear how large the final reductions would be because some of the appropriations bills were still in conference and were still subject to a presidential veto. Furthermore, the Environmental Protection Agency, whose research has focused on the effects of global change, has decided to drastically reduce its participation in the U.S. Global Change Research Program.

In a recent talk, John H. Gibbons, director of the White House Office of Science & Technology Policy, quoted the British novelist and historian C. P. Snow: "A sense of the future is behind all good politics. Unless we have it, we can give nothing - either wise or decent - to the world." Gibbons went on to say that "increasing our understanding of the climate system and the impact of human activities is a necessary part of this endeavor to gain a sense of the future." Without research, mankind will not understand the impact of its activities or how to stop its "uncontrolled global experiment" - elevating the levels of greenhouse gases in the atmosphere.

[ACS Home Page] [ACS Publications Division Page]