Materials collected from caves and lakebeds help us understand past climate change and how natural factors and human intervention combine to determine future climate change (photo: release)

Geochemical research enhances climate models
2015-01-28

Materials collected from caves and lakebeds help us understand past climate change and how natural factors and human intervention combine to determine future climate change.

Geochemical research enhances climate models

Materials collected from caves and lakebeds help us understand past climate change and how natural factors and human intervention combine to determine future climate change.

2015-01-28

Materials collected from caves and lakebeds help us understand past climate change and how natural factors and human intervention combine to determine future climate change (photo: release)

 

By José Tadeu Arantes

Agência FAPESP – What is the impact of medium- and long-term natural fluctuations in the earth’s climate? How has the world’s largest forest evolved in the last 50,000 years? To what extent do data from the past validate predictions of the future?

Questions like these were used as a backdrop for the research project “Variability of the South American monsoon system of the last three millennia integrating lake, speleothem and marine records”, coordinated by Francisco William da Cruz Junior, a researcher affiliated with the University of São Paulo’s Geoscience Institute (IG-USP), and Renato Campello Cordeiro at the Chemistry Institute of Universidade Federal Fluminense (IQ-UFF) in Rio de Janeiro. The project was supported by a cooperative agreement between FAPESP and FAPERJ, the Rio de Janeiro State Research Foundation.

The researchers reconstructed the climate and natural environment that existed thousands of years ago in what is now Brazil by investigating lakebeds and caves.

According to Cruz, this type of research into what is called “the paleoclimate” can help today’s climatologists. “We now know a great deal about high-frequency cycles, such as El Niño and La Niña, which recur every two to seven years,” he said. “But there are longer natural cycles can last decades, centuries or even millennia. We face statistical constraints when we try to detect these long cycles because our data-collection periods are not long enough. After all, they depend on the age of the collection stations.”

According to the authors, the project covered a much longer period than the 3,000 yearsto which its title refers. Cruz analyzed samples of mineral deposits from caves (speleothems), specifically stalagmites, which he dated using uranium-thorium dating before submitting them to geochemical analysis. Cordeiro profiled lacustrine sediments (partially hardened silt from lake beds) and dated different sections of the columns using carbon 14 or radiocarbon dating methods.

One of their findings was that the Amazon forest acquired its characteristic luxuriance during the extremely recent period of the past 4,000 years.

Cave samples

Radiocarbon dating cannot be used for speleothems because it works only with organic matter. Speleothems are inorganic, consisting of carbonates deposited in caves via rainwater infiltration through soil and rock.

“By measuring concentrations of uranium and thorium using a mass spectrometer, we can use the ratio between the two values to estimate the age of a sample with a margin of error of only 1%,” Cruz said.

After being dated, the samples were submitted for geochemical analysis. The chemical composition of each section of the sampled material is correlated with the environmental conditions that prevailed in the period when the layer was formed.

“It’s hard to translate chemical composition into millimeters of rain per year,” Cruz explained. “But it’s possible to detect fluctuations in precipitation that characterize more or less rainy periods, and use this as a basis on which to construct a map of rainfall distribution patterns on a continental scale.”

According to Cruz, the most reliable indicator in his samples is the ratio between different isotopes of oxygen, such as the ratio between oxygen 18, which is rare, and oxygen 16, which is the most abundant. In fact, this ratio has enabled researchers to determine sources of precipitation and indicate, for example, whether they were generated by humidity originating in the Amazon or the Atlantic.

Lake sediments

“Studying paleoclimate made me review some concepts in ecology,” Cordeiro recalled. “Ecology operates with the idea of climax, according to which the ecosystem reaches a height of complexity and then starts to decline. Paleoclimate science, however, shows that systems change in terms of structure and functioning as a result of climate and are always in equilibrium with climate change.”

“The Amazon forest systems, as we know them today in terms of biomass, are the product of a relatively stable climate in the past 4,000 years. During this period their biomass increased significantly and with a growing accumulation of carbon until the present day. They constitute the most important biome in the terrestrial biosphere as far as carbon stocks and carbon capture are concerned. This heritage is now seriously endangered by human action,” Cordeiro said.

The levels of the lakes that exist in Amazonia today trended down during a period between 50,000 and 18,000 years ago, according to Cordeiro, and many lakes dried up completely between 25,000 and 18,000 years ago because of the last ice age, when global water levels fell owing to the formation of glaciers.

Later – between 15,000 and 8,000 years ago – the climate again became wetter, and between 8,000 and 4,000 years ago there was another period of declining humidity known as the “mid-Holocene dry phase.”

“In this phase the climate was extremely variable,” Cordeiro told Agência FAPESP. “Periods of forest growth alternated with periods when the forests suffered intense disturbances and there were many forest fires. From about 4,000 years ago, with the advent of less intense dry intervals than in the mid-Holocene, the climate became consistently more humid. During the last 400 years, in particular, the forest biomass has increased significantly. More recent samples clearly show the impact of human presence via a sharp increase in carbonized particles.”

“The sedimentation rate varies from lake to lake. A one-meter column may correspond to 40,000 years in one case and only 100 in another. If the sedimentation rate is high, the sample will display high temporal resolution. On the other hand, if the sedimentation rate is low, a given column height with lower resolution lets you go back much further into the past. We have cores from Lagoa da Pata in the far north of Amazonas State near the equator, almost on the border with Venezuela, in which a 1.2m column enabled us to go back 50,000 years,” Cordeiro said.

Calculating the sedimentation rate is simple, at least conceptually. The distance between two sections of a sediment core is divided by the time interval between their respective ages, estimated by radiocarbon dating. The quotient is expressed in centimeters per year.

The ultimate goal of the research, which Cordeiro continues to develop after the end of the FAPESP-FAPERJ cooperation project, is much more ambitious. It consists of collecting sediment cores, dating the different sections, and analyzing each section for indicators that correlate with environmental conditions. “Each core requires about two years of work,” Cordeiro said.

The indicators are measured using granulometry (particle-size analysis for obtaining information about the hydrodynamics of the environment), inorganic a geochemistry (mineral composition as a basis for determining the sources of sediments), organic geochemistry (for example, whether vegetation fringing the lake in the period considered was predominantly forest or grassland), and other methods.

“We worked in cooperation with researchers who study the composition of grains of pollen deposited in sediments. If they contain pollen from arboreal vegetation, this tells us the climate was wetter at the time. Pollen from grasses and other plants adapted to conditions of water stress points to a drier climate. Pollen from cultivated species is a sign of human intervention,” Cordeiro said.

Another method he uses frequently is quantification of carbonized particles. This method offers evidence of forest disturbances during drier periods or related to human intervention. “By cross-tabulating geochemical data with pollen data, we can determine whether forest fires were spontaneous events or caused by humans,” he explained.

The past throws light on the future

In both Cruz’s speleothems and Cordeiro’s lake profiles, integration of a large mass of data enables researchers to construct scenarios regarding previous climate that make valuable contributions to today’s climate models.

“We aim to identify natural patterns so that researchers working with models can distinguish between natural occurrences and those caused by anthropic factors,” said Cruz. “Then they can try to see if anything links one with the other.”

“We began by studying the effects of long-term phenomena such as Milankovitch cycles,” Cruz explained. “These effects, which have already been described in detail in the literature, are caused by different astronomical processes,” such as changes in the distance between the earth and the sun due to the earth’s gravitational interactions with other planets and the sun, leading to precession cycles of 19,000 and 23,000 years.

“We then moved on to shorter cycles of climate change lasting a few thousand years, or centuries or even decades, like the Atlantic Multidecadal Oscillation [AMO], created by changes in ocean circulation. The challenge is to get to ever-shorter cycles, which are relatively frequent at present,” he said.

“In a very short time series, the drought that’s currently afflicting Southeast Brazil tends to be considered a unique event, but if we incorporate paleoclimate data and the composition of longer series it may be possible to detect other periods of severe drought going back hundreds of years. On that basis we can estimate how far human intervention is magnifying any natural cycles that may be involved.”

The data furnished by paleoclimate studies can not only help enhance climate models but also contribute to more effective environmental management. This is clear, for example, in the research currently being developed by Cordeiro in the Amazon to investigate the correlations between climate change and the accumulation of carbon in floodplain lakes.

“Today we know reasonably well how much carbon there is in the standing forest and how much in the soil. Colleagues have also published two important papers on carbon transport by Amazonian rivers to the sea. But as yet we have few studies on the accumulation of carbon in floodplain lakes,” Cordeiro said.

“The Amazon Basin covers some 6 million square kilometers, about 800,000 consisting of floodable areas. It’s reasonable to assume that carbon accumulation in floodplains is a very important factor in the biosphere dynamics of this fundamental element for the planet’s climate balance,” Cordeiro said.

Estimating the size of this reserve is particularly urgent in light of the plans to build a large number of dams in the Amazon region.

“The impact of these dams could be huge,” Cordeiro warned. “If they’re concrete dams, they’ll cause major changes to the hydrologic regime, practically extinguishing floodplain systems, which in aggregate act as an important carbon-accumulation compartment.”

 

  Republish
 

Republish

The Agency FAPESP licenses news via Creative Commons (CC-BY-NC-ND) so that they can be republished free of charge and in a simple way by other digital or printed vehicles. Agência FAPESP must be credited as the source of the content being republished and the name of the reporter (if any) must be attributed. Using the HMTL button below allows compliance with these rules, detailed in Digital Republishing Policy FAPESP.