When Steve McIntyre, the volunteer who deciphered the data, notified NASA of the bug, Ruedy replied. He acknowledged the problem as an “oversight” that would be corrected in the next data release. Without fanfare, NASA quietly released corrected figures and graphs on the GISS website.393 When the corrected data, shown in Illustration 124, are compared with the previous data from Illustration 2 (page 15), the changes are truly astounding.
The warmest year on record is now 1934. The year 1998, trumpeted by human-caused global warming proponents and the media as record-breaking, now moves to second place. The year 1921 takes third. In fact, 5 of the 10 warmest years on record now occur before World War II. The overall effect of the correction on global temperatures is minor, only 1-2% less warming than originally thought, but this serves as an example of why temperature data must be cross-verified. It is also a warning to those who insist on claiming global warming is demonstrated by a single, or even a cluster of warm yearly temperatures.
Though the problem reported by McIntyre seems to have been a programmatic one, there are other problems with collecting temperature measurements. Over time, the environment surrounding a weather station can change dramatically. Sites with long histories often show temperature trending upwards, but this may have nothing to do with global warming. A half century ago, a station may have been situated in a grassy field—today the same station could be located in a paved parking lot in the middle of a dense urban area. The station has not moved, but present day temperature readings will be higher due to a phenomenon called heat islands. A heat island is an area of increased temperature caused by buildings and pavement found in cities and urban areas. Often, weather station sites are not well-situated and their data are biased. This is clearly shown by the station in Illustration 125.
Illustration 125: Detroit Lakes, MN. Photo by Don Kostuch.
How important are adjustments for heat islands and sensor placement? In a 2003 study, Kalnay and Cai found effects from land-use changes and urbanization were three times greater than previously thought. When they applied their corrections to what was previously thought to be an accurate temperature history for the continental United States, they found no trace of a significant warming trend for the previous century.394 As temperature increases across the continental US have generally been much lower than in other parts of the world, this observation does not imply that global warming is not a reality. It does suggest that the rate of warming may not be as rapid as suggested by the IPCC.
Proxies and Paleoclimate
When reliable, directly measured climate data are not available, scientists must turn to proxy data. Proxies come in many forms: silt deposits in lakes and oceans, plant pollen trapped in sediment, rings in trees and coral, and the skeletons of dead organisms. Some proxy data only provide the most general indicators of climate. Other proxies provide more precise measurements.
Among the indicators used by paleoclimatologists are plant macrofossils, charcoal, diatoms, chrysophytes, phytoliths, biogenic silica and pollen analysis.395 Surprisingly, pollen turns out to be very tough stuff, able to survive where other plant matter decays and disappears. By analyzing pollen that collects in lake bed sediment layers, called varve, it is possible to gain insight into past climatic conditions. The word varve is derived from the Swedish word varv meaning “in layers.” From the pollen of trees, grasses and flowering plants, scientists can tell if past temperatures were warm or cold, the weather wet or arid.
It was the presence of pollen in northern lake bed silt deposits that first identified the onset and end of the Younger Dryas period (page 144). Dryas is the genus name of a small flowering plant that likes cold weather. When the last ice age waned, dryas pollen vanished from European lake sediments. But then, it suddenly reappeared in the pollen record, marking the return of glacial conditions during the Younger Dryas.
Climatic variations we can observe directly today, such as El Niño and the North Atlantic Oscillation, are visible in climate proxies such as tropical corals and European tree rings. Volcanic eruptions and inferred solar variability, which can have significant cooling effects, are reflected in historical records and ancient writings.
Many proxies involve measuring the ratios between isotopes, different versions of atoms belonging to the same element. Elements are identified by the number of protons present in their atomic nuclei. The number of protons is matched by the number of electrons surrounding each nucleus. Since atoms form chemical bonds by sharing electrons, it is the number of protons/electrons, the atomic number, that determines an element's chemical properties. But there is another type of subatomic particle present in the nuclei of atoms, neutrons.
Neutrons weigh slightly more than protons, but they carry no electric charge. Atoms can have different numbers of neutrons in their nuclei without changing the number of protons and electrons. For example, all oxygen atoms have eight protons and eight electrons. The isotopes 16O, 17O and 18O are each forms of oxygen with 8, 9 and 10 neutrons in their nuclei, respectively. They are the same chemically, but have slightly different weights. Of the three isotopes, 16O is the most abundant comprising over 99.7% of the oxygen on Earth. By measuring slight variations in the ratios of isotopes, scientists can uncover important clues about Earth's climate in the past.
Henry C. Fricke, of the University of Michigan, tested teeth from dead Vikings for oxygen isotopes. His study analyzed the tooth enamel from 29 human teeth excavated at three archaeological sites in Greenland and one in Denmark. Fricke explained, “the ratio of heavy (O-18) to light (O-16) isotopes in the calcium phosphate that comprises tooth enamel is directly related to this isotopic ratio in rain or snow falling on a local area, because the oxygen in this precipitation is incorporated into the tooth enamel of growing children who drink from local groundwater supplies like springs, lakes and rivers.”396 Comparing teeth from skeletons buried in 1100, with those buried in 1400, he documented a 1.5°C drop in temperatures, confirming the onset of the Little Ice Age in northern Europe.
Measuring Time
Some isotopes are radioactive, meaning they are unstable and decay into other elements over time. This fact can be used to establish the dates of some substances. Plants absorb carbon from the atmosphere when they are alive. Carbon comes in two stable, nonradioactive forms: 12C and 13C. But there is another form of carbon present in Earth's atmosphere, the unstable isotope 14C. Carbon-14 has a half-life of 5730 years and would have long ago vanished from Earth were it not for the shower of cosmic rays that bombards our planet. Cosmic rays striking nitrogen atoms in Earth's atmosphere continually replenish the amount of terrestrial 14C. Thus, the ratio of 14C to the stable isotopes of carbon remains fairly constant.397
As plants grow, they absorb all forms of carbon equally, so the level of 14C present in their living tissues remains constant. But when a plant dies, this exchange stops, and the amount of 14C gradually decreases through radioactive beta decay. Radioactive isotopes are subject to exponential decay, the rate of which is expressed in half-life time. The half-life of 14C is 5730 years, meaning that in 5730 years, half of the 14C atoms present in dead plant matter will decay into atoms of nitrogen. Atoms of 14C become 14N atoms as neutrons in the carbon nuclei become protons, and electrons are ejected. Eventually, the 14C will disappear altogether. So, by measuring the ratio of 14C to the stable forms of carbon present in dead plants, it is possible to assign a date to when they died.
This technique of radiocarbon dating was discovered by Willard Libby398 and colleagues in 1949. Because of its short half-life, Carbon-14 dating is only viable for organic matter younger than ~50,000 years before the present (BP), and usually much younger than that. Despite this restriction, carbon-14 dating revolutionized the field of archeology, and for its discovery Libby was awarded the Nobel prize in Chemistry. This technique, while very useful, has built-in uncertainties that grow larger as the material being dated becomes older. Radiocarbon labs generally report an uncertainty of around 3000 ± 30 BP, or about 1% fo
r things 3000 years old. All proxy measurements that involve isotope decay have similar built-in uncertainties.
For longer time scales, isotopes of potassium (K) and uranium (U) can be used to extend our view of the past. Atoms of 40K decay into 40Ar through beta decay much like 14C is transformed into 14N. Uranium, however, follows a different path, undergoing radioactive decay through nuclear fission. There are two naturally occurring isotopes of uranium, 235U and 238U, which decay through a series of fission transformation to become the lead isotopes 207Pb and 206Pb respectively.399 As with all of these radioisotope dating techniques, the key is knowing the material's half-life and starting with known proportions of the isotopes in question.
For uranium dating, the half-lives of 235U and 238U have been measured with great accuracy and are found to be 703,800,000 and 4,460,000,000 years, respectively. Because of these long half-lives, uranium allows dating of rock from much further in the past than carbon-14 dating of organic material—the problem is finding a good sample to perform the analysis.
Fortunately, nature has provided tiny time capsules in the form of the mineral zircon. Zircon sand consists of nearly microscopic needles of silica glass. When these tiny grains form, they contain small amounts of uranium, but the physical process of formation excludes lead atoms. This results in a material that initially contains uranium and no lead to confuse the dating. Over time, the trapped uranium decays into lead and the proportion of lead to uranium reveals the elapsed time since the zircon formed. It has been through the careful analysis of uranium contained in zircon grains that the timing of the great Karoo Ice Age and the subsequent Permian-Triassic Extinction have been accurately established. Uncertainties of less than 1% can be attained, but 1% of 100,000,000 years is still 1,000,000 years.
Aside from dating rock and organic material with isotopes, other, more accurate methods exist, at least for the past million years or so. These dating methods apply to ice, mud and tree rings and involve a simple concept—counting. As trees grow, they form wood in plainly discernible bands that mark the alternating season of summer and winter. Sediments form yearly seasonal layers on both the ocean floor and the bottoms of lakes. The ice found in glaciers is formed from annual snowfalls. All these natural processes provide enterprising and patient scientists with accurate natural time lines stretching back into the past. When layer counting is combined with other proxy measurements, climate history can be revealed.
Indicators of Climate
Some natural processes create isotopes at constant or measurable rates. Radiation from the Sun and cosmic rays create beryllium (10Be) and chlorine (36Cl) isotopes in Earth's atmosphere, which then decay at known rates. Consequently, these proxy indicators can be used to determine past solar activity. Other processes discriminate against different isotopes of the same element, providing different insights. One such process, involving tiny ocean plankton called foraminifera, provides the basis for one of the main methods of tracking temperature over ages past.
Foraminifera, or forams for short, are single-celled marine organisms with hard shells. Their shells are commonly divided into chambers which are added during growth, though the simplest forms are open tubes or hollow spheres. Depending on the species, the shell may be made of organic compounds, particles cemented together, or crystalline calcite. When forams die, their shells drift to the bottom of the ocean, becoming part of the ocean floor sediment. In large enough concentrations, these tiny shells can form large beds of chalk and limestone.
Illustration 126: Benthic foraminifera. Source USGS.
Forams have been found in the fossil record as far back as the Cambrian. Early forams were much larger than their modern relatives, but their small, almost microscopic size, makes modern forams much more useful than the larger fossils. In the 1920s, it was discovered that many species of foraminifera were geologically short-lived, and others are only found in specific environments. This allowed paleontologists to determine when rock formed, and the environmental conditions at the time, by examining the specimens in a rock sample. As a result, the oil industry became a major employer of paleontologists who specialized in these microscopic fossils.
Foraminifera are divided into two primary groups based on their mode of life; planktonic marine floaters, and benthic sea floor dwellers. Benthic foraminifera are found at all latitudes and occupy a wide-range of marine environments, from brackish estuaries to the deep ocean basins. Much like plant pollen on land, they are very useful as environmental indicators. Particular species and assemblages can be used to identify ancient environmental conditions. Forams found in deep-sea core data have been utilized to identify changes in intermediate and deep water circulation, and to determine past sea level changes in coastal regions.400
Forams' role in paleoclimatology became even more prominent when biologists made the observation that the portion of 18C forams absorbed varied with water temperature. In 1947, American nuclear chemist Harold Urey401 combined that observation with newly-developed techniques that allowed precise measurements of very small quantities of atomic isotopes to create a way of determining temperatures in the ancient ocean. Cesare Emiliani, an Italian geologist who was a graduate student in Urey's lab, became interested in the problem as it applied to samples of foram shells taken from deep sea sediment cores. He is the same Dr. Emiliani who was interviewed by Time magazine about the impending new ice age in 1972 (page 42).
To obtain a sediment core, a long tube is sunk into the sea floor and then extracted with the layers of sediment intact. The date of each layer is established by counting down from the top layer or other methods. Test material is recovered from a single layer of sediment by carefully picking out a few hundred pinhead sized shells. The shells are cleaned, ground to powder and then heated to release CO2 gas. The ratio between carbon isotopes is then measured using a mass spectrometer. From this ratio, known as δ18O,402 the temperature of the sea water that the forams lived in is estimated.
By 1955, Emiliani had applied this technique to many sea bed cores, compiling a record of ocean temperature change going back nearly 300,000 years. It was from these data that Milankovitch's theory of astronomically driven climate change rose, phoenix-like, from the ashes. At that time, the geological mainstream had rejected the Croll-Milankovitch cycles and settled back into their traditional beliefs regarding glacial cycles. Consensus among geologists said that there had been only four long glacial episodes in the Pleistocene Ice Age, separated by long warm periods. When Emiliani looked at his data he found a dozen cycles, all in close agreement with Milankovitch's predictions.
Of course, this new case for Milankovitch's old theory was not easily accepted. Debate raged for more that 10 years, with traditional geologists defending their turf vigorously. Emiliani's data was rejected when it was shown that water containing 18O and 16O evaporated at different rates depending on temperature. The lower evaporation rate for H2(18O) would tend to bias the δ18O from the foram shells. Additional evidence provided by the mixture of foram species present in the core samples helped validate Emiliani's original result. To finally end the debate, corroborating data from ancient coral beds collected by Wally Broecker settled the argument.403
Interestingly, the argument that nearly derailed Emiliani's analysis, the different temperature-dependent evaporation rates of water containing different oxygen isotopes, led to the development of the other major source of paleoclimate temperature data—glacial ice cores.
Ice Cores, Gas Bubbles and Isotopes
The other major source of paleoclimate data combines several different proxies, including measurements of multiple isotopes and layer counting. That source is ice core data retrieved from mountain glaciers and the long-term ice sheets of Greenland and Antarctica. Samples are often used to establish temperature and CO2 levels in times past. To do this, three values must be accurately identified; date, atmospheric CO2 level, and temperature. Each of these data points are determined as follows.
There are four distinct methods f
or determining the ages of ice cores. Three are direct experimental tests and the fourth rests on somewhat uncertain theories. The methods for dating ice core layers are; counting annual layers, using predetermined ages as markers, radioactive gas dating, and calculating ice flow rates. Each of these methods is complex and presents unique problems. We will concentrate on the most popular method, annual layer counting.
The annual layer counting method looks for items in the ice that vary with the seasons in a consistent manner. There are a number of approaches to this as well. Among these are looking for indicators that depend on the temperature (colder in the winter and warmer in the summer) and solar irradiance (less sunlight in winter and more in summer). Regardless of the marker used, going back tens of thousands of years requires identifying a proportional number of yearly transitions in the ice core sample—this can be very time consuming.
Both the temperature and irradiance methods are based on measuring isotopes of certain elements; hydrogen or oxygen for temperature, and beryllium or chlorine for irradiance. Since we are interested in how temperature is determined anyway, we will concentrate on the temperature method using oxygen isotopes.
Of the temperature dependent markers, the ratio of 18O to 16O is the most important, though measurements based on isotopes of hydrogen are also used (δ2H or δD404 ). Water molecules composed of H2(18O) evaporate less rapidly and condense more readily than water molecules composed of H2(16O). Thus, water evaporating from the ocean starts off H2(18O) poor. As the water vapor travels towards the poles, it becomes increasingly poorer in H2(18O) since the heavier molecules tend to precipitate out first. This depletion is a temperature-dependent process so in winter the precipitation is more enriched in H2(16O) than is the case in the summer. Therefore, each annual layer starts 18O rich, becomes 18O poor, and ends up 18O rich. Once the H2O is captured in the crystalline ice of snowflakes, the δ18O ratio is fixed.
The Resilient Earth: Science, Global Warming and the Fate of Humanity Page 26