BECAUSE THE ATMOSPHERIC system is dynamic, variations in temperature and precipitation regimes have occurred throughout the Earth’s history. Glaciations and deglaciations, for example, show how changeable climate can be. Detecting climate changes shows that climate differs significantly from some previous episode. Detection differs from attribution, which denotes the causes of those changes. While daily fluctuations in weather happen as a consequence of a chaotic atmosphere; climate fluctuations, based on smoothed aggregations of weather events, represent significant shifts in the longer-term averages. To discern these shifts, numerous techniques and methods are employed.
Climate changes, which occur at a variety of temporal scales, are assessed using both directly-observed and proxy data. Directly-observed information refers to meteorological and climate-related variables that have been directly evaluated, such as temperature measured by a thermometer, or cloud tops captured by satellite imagery. Directly-observed data have been used to chart recent increases in air temperature and carbon dioxide (CO2), concentration, as evidenced by Charles Keeling’s famous graph of climbing CO2 concentrations at the Mauna Loa Observatory.
Proxy data are non-meteorological data, from which climate information is inferred; they serve as a satnd-in for such climate evidence when direct observations are not always available. Examples of proxy data include ice cores, lake sediments, and historical harvest records. Among the information gleaned from proxies are temperature, moisture, sea level, and chemical composition of the atmosphere. These facilitate the reconstruction of past climates.
Although directly-observed data may be preferable, the instrumental record is often too short for all but the most recent evaluations of climate changes. For example, the longest continuous temperature record shows monthly mean temperature only since 1659 (for central England); daily records for the same location extend only to 1772. Many other records are less than 100 years old. Temperature records also require filtering because of biases that may be present in the readings. The effects of urbanization, instrument upgrades, and location changes can result in false changes. Newer technologies, such as satellites, yield even shorter records. However, much climate change research assesses variability at temporal scales up to millions of years, rendering proxy evidence a necessity.
Proxy data
Proxy or substitute data extend climate change knowledge beyond the instrumental record, although there is a decreasing resolution and confidence with increasing time. Because climate variability has been identified at various temporal scales, the resolutions of the various proxies differ. Removing the climate signal and reconstructing past climates from them requires the proxy to be related to the climate. Therefore, some link between the two needs to be established; for example, by experimentation or construction of models.
Written historical records, such as harvest yields or military records, as well as diaries and phenological records, have been used to derive information. Temporal resolutions vary, from daily to annual. Farmers’ diaries often discuss the weather. Grain harvests, for example, have been linked to precipitation. Dates of cherry blossomings in Japan can provide temperature trends. Historical records may be biased toward significant weather events (such as blizzards), so care must be taken when interpreting them.
Tree rings and pollen
The study of tree rings, known as dendrochronology, can yield annual information on climate conditions for the past millennium or so. Every year, many tree species increase the diameter of their trunks by developing concentric rings formed by a layer of wood cells underneath the bark. While each ring corresponds to one year of growth, the spacing between successive rings is usually not uniform. This unevenness is due to differences in annual growth patterns, signifying climatic changes that have either encouraged or suppressed tree growth. Different tree species are sensitive to moisture and/or temperature changes. These environmental conditions can be inferred from the spacing and thickness of the tree rings. Relatively thick rings with a large spacing between successive rings indicate a more optimal growth setting, whereas narrower, tightly-spaced rings point toward climatic stress conditions (such as drought). Dendrochronology is limited to regions that continually support tree growth.
Another form of proxy data, derived from vegetation, comes from pollen and spores produced by plants. Wind can transport pollen grains to nearby water features, where they settle to the bottom and are preserved in sediment layers. Sediment cores extracted from lake beds are analyzed for the amount and type of preserved pollen found in successive sediment layers. Because pollen grains have unique features, plant types can readily be identified and, through techniques such as radiocarbon dating, indicate the time of deposition. Plant growth and distribution is sensitive to environmental conditions, providing an indicator of temperature and moisture trends on time-scales from centuries to millennia.
Sea-floor sediments and ice cores
As with lake sediments, the layers or stratigraphy of oceanic deposits and corals can give important clues to paleoclimates. Sea-floor sediments are primarily an accumulation of calcium carbonate- (CaCO3) based shells from organisms that once lived near the ocean surface. These organisms are often sensitive to changes in temperature and salinity, proliferating under optimal conditions and declining in unfavorable conditions.
Because the ocean surface is closely connected to sea-level climate conditions, the amount and type of shells found in the ocean core layers correlates well with atmospheric conditions when the organism died. Coral are small marine organisms that are also made of calcium carbonate, recording oceanic temperature conditions through their growth patterns. Like tree rings, coral thrive under certain conditions; coral are thicker during warmer, and thinner during colder, ocean episodes.
The calcium carbonate shells from ocean cores and coral are also source material for oxygen isotope analysis. This technique examines the ratio between two different isotopes of oxygen that are recorded in the skeletal remains of marine life. Ocean water contains two forms of oxygen molecules, the more common form with an atomic weight of 16 (16O) and the atypical variety with an atomic weight of 18 (18O). Water containing the lighter oxygen (16O) isotope evaporates more readily than its heavier counterpart. This means that during periods of extensive glaciation, the amount of 18O relative to 16O in ocean water increases as more of the lighter oxygen isotope is precipitated out as snow, becoming concentrated in glacial ice. Conversely, the 18O/16O ratio decreases during warmer interglacial episodes, when surface runoff returns the 16O previously stored in the cryosphere. Hence, oxygen isotope analysis is also useful for examining the precipitation layers of ice cores.
Taken from glaciated mountaintops and ice sheets, ice cores provide information on past climates in a variety of ways. Oxygen isotope analysis and the depth of snow accumulation in successive layers can give clues on temperature and moisture conditions. Major volcanic eruptions or meteorite impacts are often recorded as a significant layer of dust between ice layers; these episodes adversely impact surface insolation and global temperatures. Ice cores can also provide information on the past chemical composition of the atmosphere from small air bubbles trapped within. Depending on the length, ice cores enable reconstruction of past climates going back thousands of years.
