climate change and deep water formation During the past million years the Earth's climate has varied on two broad timescales. There have been regular interglacial-glacial cycles over thousands of years and irregular, decadal changes within these cycles. Astronomically driven changes in the intensity of the seasons might periodically trigger climate change. It is, however, increasingly clear that other factors result in the rapid changes recorded in ice-core records, most probably variations in deep water formation.

Modern oceanic deep water masses

A deep water mass is a body of water, in contact with the sea floor, which is identifiable by its distinct combination of physical and chemical characteristics. The properties most used in identifying deep water masses are potential temperature and salinity, since away from the sea surface these can be changed only by mixing with other water masses. At present, deep water is formed in the North Atlantic and to a much smaller degree in the Weddell continental shelf, in the Atlantic sector of Antarctica. These regions produce two very distinct water masses, the southward-flowing, ‘warm’ (2 °C), saline (>34.9‰, i.e. more than 34.9 parts per thousand) North Atlantic Deep Water (NADW), and the northward-flowing, colder (<0 °C), less saline (>34.7‰) Antarctic Bottom Water (AABW).

Each winter in the Norwegian-Greenland sea northward-flowing water at approximately 800 m depth rises to the surface because of wind stress. This water is chilled from 10 °C to 2 °C, which, coupled with its already high salinity, results in a water mass dense enough to sink to the ocean floor, producing NADW. The winter injection of saline water from the Arctic Ocean increases the density of the NADW. NADW ponds up north of the ridge between Greenland and Scotland (which forms a major barrier to deep water flow below 400 m), until it intermittently overflows, cascading into the North Atlantic. About 20 per cent of NADW is formed in the Labrador Sea and mixes with the NADW travelling westward from the Norwegian Sea. The formation of NADW gives off heat equal to 30 per cent of the yearly direct input of solar energy to the surface of the North Atlantic (5 × 10²¹ cal); this heat is the reason for western Europe's mild winters.

NADW flows into the Pacific, via the Indian Ocean and the Antarctic Circumpolar Current. It slowly upwells, becoming shallower, reaching the surface in the North Pacific and returning to the North Atlantic as part of the upper warm water circulation. This is termed the thermohaline circulation. The global thermohaline circulation appears to be self-sustaining. In the North Atlantic, NADW outflow is balanced by warm surface inflow. Owing to the low salinity of sea surface waters in the North Pacific no deep water is formed at present in this region. The low salinity is due to the upwelling of cold deep water with low evaporation rates and the input of fresh water evaporated from the North Atlantic.

AABW forms by (1) the seasonal removal of water to marine ice sheets, which increases sea-water salinity, and (2) the surface cooling of sea water in pockets of ocean enclosed by sea ice termed polynyas. The largest polynya recorded is 1000 km by 350 km. In the Weddell Sea cold-core eddies up to 28 km across form cold-water chimneys siphoning water down into the AABW. AABW occurs in all the Atlantic, Indian, and Pacific ocean basins. It is the densest water mass found in the oceans, and in the Atlantic it travels northwards below the NADW at a depth greater than 2.5 km. Variations in NADW and AABW properties are largely a reflection of mixing between the two water masses.

Proxies of deep water masses

As deep waters travel from their regions of formation they become enriched in total dissolved organic matter, because of the continual ‘rain’ of detritus from the photic zone. This is termed the ageing effect. This enrichment results in a water mass that (1) is more acidic and therefore more corrosive to CaCO₃; (2) has lower δ¹³C values, because δ¹²C is preferentially fixed in organic matter (δ¹³C and δ¹²C, the delta values for the isotopes carbon-13 and carbon-12 respectively, are measures of the relative differences between the values for a sample and a standard); and (3) has an increased phosphorus content; the Atlantic deep water contains on average about half as much phosphate and nitrate as deep water in the Pacific Ocean, while the Circumpolar Deep Water (CPDW) in the southern ocean reflects a mixture of the two.

Changes in deep water distribution over time are reflected in the sedimentary record. The CaCO₃ content and degree of dissolution of calcite fossils varies, reflecting the corrosiveness of the benthic water mass. In the Atlantic glacial stages, calcite dissolution often increased, suggesting the presence of older deep water. Epifaunal benthic foraminifera have been used to trace past water masses by using the levels of carbon stable isotopes and cadmium incorporated in their shells. Certain species are known to incorporate δ¹³C and cadmium into their shells in equilibrium with the surrounding sea water. Other species form shells at a constant disequilibrium to the surrounding sea water. These can also be analysed and a correction factor can be applied to obtain a value for the original sea-water composition. Cadmium is important, for in the modern ocean it shows a strong correlation with phosphorus and nitrate in sea water. Both proxies indicate, in general, an increase in the ageing effect in the NADW during glacial periods. The ¹⁴C/¹²C ratio within calcite shells can be used as a proxy for the age of the surrounding water when the shell calcified. Variations between ¹⁴C/¹²C ratios for coexisting planktonic and benthic foraminifera are used to date the benthic water mass. In the tropical Pacific the age difference today is 1600 years, whereas in the tropical Atlantic it is 350 years. It appears that during the glacial period the deep water mass in the Atlantic was at least twice this age. Sediment grain-size variations, the distribution of volcanic ash fragments, and deep-sea channels have also been used to track changes in deep water distribution and directions of bottom-water flow.

Past changes in deep water formation

Using a variety of geochemical proxies it is clear that the thermohaline circulation has more than one stable mode. There are two timescales of variability in NADW production: variations on a glacial-to-interglacial scale, and short rapid changes within glacial-interglacial periods (for example the Younger Dryas). NADW production may be a stable system for millennia and then undergo a rapid transition to a new state on the time scale of a decade. It appears that for at least the past 1.5 Ma glacial stages have had significantly less NADW production compared to interglacial periods. During the last glacial maximum the production of NADW was greatly reduced. It occurred in a region south of Iceland, and the water mass formed was convected to a shallower depth, not reaching the sea floor.

The most likely cause of the larger time-scale variability in NADW production are changes in the North America ice sheet, possibly driven by astronomical processes. These variation are on a 23 000-year cycle and therefore appear to be linked to the Earth's precessional cycle. Atmospheric general circulation models show that the North America ice sheet directly controls the sea-surface temperature of the North Atlantic. Strong cold winds, generated on the ice sheet's northern flanks, blow out across the ocean, cooling the surface waters. Lower sea-surface temperatures result in less evaporation and therefore less saline, less dense surface waters. Modelling has shown that a slight decrease in surface-water salinity could markedly decrease deep-water formation in the Norwegian Sea within a few decades. There is a strong correlation between sea-surface temperature (based largely on planktonic foraminiferal assemblages) and rates of NADW production on a glacial-interglacial time scale. A growth in ice sheets, causing a southerly shift in the polar front, might also inhibit the inflow of saline surface water from the South Atlantic which is required for NADW production.

Modelling has indicated that changes in wind direction along the Marginal Ice Zone (MIZ), may also be important in controlling the rate of NADW production. At present the dominant wind direction during winter along the MIZ is north-east, resulting in northward Ekman transport of water and ice. As ice drag on the underlying water is greater than that due to surface water drag, water is being removed to the north faster than it is being replaced from the south along the MIZ. Water therefore upwells along the MIZ to remedy the imbalance. This water is substantially cooled and sinks to form NADW. During glacial periods, because of the large North America ice sheet, the wind direction is westerly. This results in a southward Ekman transport, a build-up of water along the MIZ, and downwelling of water. This water is not, however, dense enough to sink to the sea floor.

Short timescale variations

The Younger Dyras represents a period of almost complete return to glacial conditions at 11 000–10 000 radiocarbon years BP in the North Atlantic during the last deglaciation. This has been related to injections of melt water from the North American ice sheets. During the Younger Dryas it appears that meltwaters ponded up on the southern edge of the ice sheets forming Lake Agassiz. This appears to have subsequently overflowed into the North Atlantic in a series of complex catastrophic events. This would result in a less dense ‘freshwater cap’ in the North Atlantic, which would favour stable stratification of the ocean. Modern observations have shown that a decrease in salinity of more than 1‰ in the surface waters is sufficient to suppress NADW formation.

There have been at least three other periods of rapid cooling of the North Atlantic climate since 14500 bp. These are all believed to have been caused by injections of melt water, reducing the rate of NADW production. The point of entry of the melt water appears to have varied between the Gulf of Mexico, via the Mississippi, and the Arctic Ocean via the Siberian river system.

The termination of the Younger Dryas was extremely rapid, possibly taking less than a decade. Greenland ice cores indicate a rapid increase in snow accumulation at the termination. Modelling suggests that a switching on of NADW production will result in increased advection poleward of warm surface waters. This is a suggested cause for the end of the Younger Dryas and the associated increased input of atmospheric moisture over the North Atlantic.

Late Quaternary sediments in the North Atlantic contain clear horizons of ice-rafted debris. These are termed ‘Heinrich events’; they represent decadal-length events of atmospheric and sea-surface cooling, reduction in the flux of planktonic foraminifera, lower sea-surface salinities, and brief, but exceptionally large, discharges of icebergs. These phenomena are the result of the slow thickening of the North America ice sheet on frozen sediment (Binge phase), giving way to rapid ice-stream surges as the water-saturated sediment defrosts (Purge phase). The injection of large amounts of ice and melt water again forms a ‘freshwater cap’, reducing the rate of NADW production and resulting in colder north Atlantic temperatures. The effect of Heinrich events is widespread and is seen in widely separated places: in pollen records from Florida and ice-core records from the Andes. Heinrich events and their effect on NADW production may well be a major cause of rapid climate changes within interglacial periods.

Global effects

Prolonged droughts in the Sahel and tropical Mexico during the past 14 000 years coincided with major injections of fresh water into the North Atlantic. A reduction in the production of NADW would slow the thermohaline circulation, and less water would be drawn into the North Atlantic, resulting in colder surface waters and a decrease in the rate of evaporation. This would in turn cause reduced precipitation on surrounding land masses. Between 1968 and 1982 colder, fresher waters in the Norwegian and Labrador seas, indicative of decreased NADW production, have been correlated with decreased precipitation over West Africa as the tropical rain belt shifted southward. A more sluggish thermohaline circulation would also reduce the strength of upwelling in the equatorial oceans. This would affect the rate of moisture input into the world's great tropical convective systems, resulting in more arid conditions at high latitudes.

Near-contemporaneous climate variations recorded in the Greenland and Antarctic ice cores are believed to have been driven by variations in NADW input into the Southern Ocean. Reduced NADW input would lower Antarctic surface temperatures because of decreased heat release from ocean to air, amplified by increased albedo due to ice growth. In addition, a northward shift of the Antarctic climate would reduce the poleward heat transport through the atmosphere. Modelling has, however, suggested that increased input of NADW could cool the southern hemisphere, as a larger volume of warm surface/intermediate water must be removed to replace that incorporated in NADW. The input of NADW could also displace warm intermediate waters northward in the Southern Ocean.

Deep water and CO₂ variations

Carbon dioxide (CO₂) is a major greenhouse gas, and variations in the levels of atmospheric CO₂ have a profound effect on the Earth's climate. Substantial changes in atmospheric CO₂ between glacial-interglacial periods are recorded in ice cores. These changes are at least partly driven by variations in the rate of NADW production. NADW is known from present-day measurements to have a very low alkalinity: approximately 2.3 10⁻³ eq/kg, as compared with 2.4–2.45 10⁻³ eq/kg for all other deep-water masses. Alkalinity in the oceans may be defined as the combined concentration of bicarbonate and carbonate ions. In modern oceans the CO₂ partial pressure, _p(CO₂), is controlled by the alkalinity; higher alkalinity relates to increased _p(CO₂). In the Atlantic, changes in the deep water _p(CO₂) are largely a result of mixing of low-alkalinity NADW and high-alkalinity AABW. The NADW flows into the CPDW, part of which subsequently upwells in the Antarctic, forming the southern low-latitude surface waters. There is little production of CaCO₃ by organisms within these surface waters, which could have regulated the alkalinity of the sea water. Changes in CPDW alkalinity are therefore almost directly controlled by the input of NADW. CO₂ is transported from warm surface waters to cold surface waters. An increase in the potential of the cold high-latitude surface waters as a sink for CO₂ would therefore have an influence far beyond that indicated from its area. Ocean box models have shown that the polar oceans control the _p(CO₂) in warm surface waters and therefore also atmospheric CO₂ levels. Changes in the alkalinity and therefore _p(CO₂) of the southern low-latitude surface waters may have been involved in the decrease in CO₂ and cooling observed during glacial stages. During glacial stages NADW still reached the Antarctic, but at much reduced levels. Decreased NADW input into the CPDW resulted in surface waters with an increased _p(CO₂). It is important to note that changes in the input of NADW to the Southern Ocean can explain less than half the variation in atmospheric CO₂. The remainder is most likely driven by increases in high-latitude surface water productivity and variations in total ocean nutrient distribution. The use of boron isotopes (¹¹B/¹⁰B) and barium concentrations in the calcite of foraminifera as a proxy of sea-water _pH many well increase our knowledge of this topic. At present the Ba/Ca ratio of NADW is 1.8, as compared with the value of 4.5 for the more acidic Pacific Deep Water.

During the first pulse of deglaciation at 15 000 years bp it appears that there was a major flushing of the North Atlantic deep-sea basins with oxygen-rich water as NADW production was reinvigorated. This oxidized previously deposited organic matter, resulting in the generation of CO₂ and dissolution of carbonate in situ. The effect of this on NADW, CPDW, southern low-latitude surface water _p(CO₂), and atmospheric CO₂ is still uncertain.

Neogene glacial development

The development of the southern ice sheet and the resulting steepening in the global thermal gradient is related to deep water formation. NADW was not initiated until the late Miocene. Deep water during the early Miocene was formed in the low latitudes of the Tethys Sea through to the Indian Ocean and appears to have been warmer and more saline than NADW. Warmer deep water upwelled at high latitudes, resulting in warmer surface waters in the Southern Ocean. There was consequently a large amount of evaporation next to the cold land masses. More water in the atmosphere resulted in greater snowfall and is believed to be a cause of the development of the East Antarctic Ice Sheet (EAIS) during the middle Miocene. As the atmospheric temperature gradient increased, boundaries between climate zones strengthened, increasing the aridification of mid-latitude continental regions (Australia, Africa, and the Americas), and grasslands developed, providing an environment for the evolution of grazing mammals.

Another feature of an increased temperature gradient is stronger oceanic circulation. An increase in upwelling strength during the middle Miocene resulted in greater sea-surface productivity, which is recorded in extensive deposits of organic-rich sediments in the Pacific; for example the diatom-rich Monterey formation in California. This resulted in the removal of CO₂ from the atmosphere into sediments. A positive feedback loop was set up with a lowering of temperatures, since the removal of this major greenhouse gas was initially driven by cooling due to EAIS growth. The loop appears to have been broken when the nutrients available in the oceans were used up. This appears to have been a crucial step, cooling the Earth sufficiently for late Neogene glaciation. In general the Neogene unipolar ice sheet system was considerably less sensitive to variations in deep water temperatures than the late Quaternary bipolar ice sheet system.

Palaeocene benthic extinctions of deep-water fauna might be related to variations in deep-water masses. A switch from high-latitude cold, relatively fresh, deep waters to low-latitude warmer, saline deep water would result in a deep-sea oxygen deficiency, causing the remarkably rapid extinctions recorded in benthic foraminiferal assemblages, which took place in less than 3000 years. Even older deep-sea organic-rich sediments may reflect changes in the preservation potential of carbon caused by changes in deep water circulation. These sediments, acting as a sink for CO₂, may have had a major influence on the levels of atmospheric CO₂ and subsequent global climate change.

Stephen King

Bibliography

Broecker, W. S. and and Denton, G. H. (1990) What drives glacial cycles. Scientific American 262 (1), 42–50.
Duplessy, J. C.,, Shackleton, N. J.,, Fairbanks, R. G.,, Labeyrie, L.,, Oppo, D.,, and and Kallel, N. (1988) Deep water source variations during the last climate cycle and their impact on the global deep water circulation. Paleoceanography 3, 343–60.

climate change and palaeovolcanism Scientists have long known that major explosive volcanic eruptions, which produce large quantities of ash, often result in cooling of global climate. Such eruptions are often associated with the injection of large quantities of volcanic ash into the stratosphere, where it can persist for several years. Dust injected into the troposphere, by contrast, is soon washed out by rain. The ash that reaches the stratosphere, however, absorbs incoming solar radiation and is heated.

It is generally believed that large low-latitude eruptions are likely to cause climate cooling on a global scale, whereas those that take place at higher latitudes tend to cause cooling within one particular hemisphere. Professor Hubert Lamb attempted to compare the possible climatic effects of the veils of ash caused by different historical eruptions. His ‘dust veil index’ (DVI) has a reference value of 1000, based on the violent eruption in 1883 of the island of Krakatoa in Indonesia. Table 1 lists the major volcanic eruptions that occurred between 1680 and 1970; the DVI value attributed by Lamb to each eruption is also given. A particularly well-known and catastrophic event took place during 1815 when the eruption of Mount Tambora in Indonesia resulted in the establishment of a dust veil over much of Europe and led the poet Byron to write ‘Time of Darkness’.

Table.1 Major volcanic eruptions during the past 300 years The associated Dust Veil Index value according to Lamb is shown for each eruption. (After Lamb 1982.)

Alastair G. Dawson

Bibliography

Dawson, A. G. (1992) Ice Age Earth: Late Quaternary geology and climate. Routledge, London.
Kemp, D. D. (1990) Global environmental issues: a climatological approach. Routledge, London.
Lamb, H. H. (1982) Climate history and the modern world. Methuen, London.

CLIMATE CHANGE AND HUMAN HEALTH

Human societies over the ages have depleted natural resources and degraded their local environments. Populations have also modified their local climates by cutting down trees or building cities. It is now apparent that human activities are perturbing the climate system at the global scale. Climate change is likely to have wide-ranging and potentially serious health consequences. Some health impacts will result from direct-acting effects (e.g., heatwave-related deaths, weather disasters); others will result from disturbances to complex ecological processes (e.g., changes in patterns of infectious disease, in freshwater supplies, and in food production).

WHAT IS CLIMATE CHANGE?

Global climate change is caused by the accumulation of greenhouse gases in the lower atmosphere. The global concentration of these gases is increasing, mainly due to human activities, such as the combustion of fossil fuels (which release carbon dioxide) and deforestation (because forests remove carbon from the atmosphere). The atmospheric concentration of carbon dioxide, the main greenhouse gas, has increased by 30 percent since preindustrial times.

Projections of future climate change are derived from global climate model or general circulation model (GCM) experiments. Climatologists of the Intergovernmental Panel on Climate Change (IPCC) review the results of these experiments for global and regional assessments. It is estimated that global mean surface temperature will rise by 1.5° to 3.5° C by 2100. This rate of warming is significant. Large changes in precipitation, both increases and decreases, are forecast, largely in the tropics. Climate change is very likely to affect the frequency and intensity of weather events, such as storms and floods, around the world. Climate change will also cause sea level rise due to the thermal expansion of the oceans and the melting of the mountain glaciers. Global mean sea level is anticipated to rise by 15 to 95 centimeters by 2100. Sea level rise will increase vulnerability to coastal flooding and storm surges. The faster the climate change, the greater will be the risk of damage to the environment. Climatic zones (and thus ecosystems and agricultural zones) could shift toward the poles by 150 to 550 kilometers by 2100. Many ecosystems may decline or fragment, and individual species may become extinct. The IPCC Second Assessment report concludes that climate change has probably already begun.

IMPACTS ON HEALTH

To assess the potential impacts of climate change on health, it is necessary to consider both the sensitivity and vulnerability of populations for specific health outcomes to changes in temperature, rainfall, humidity, storminess, and so on. Vulnerability is a function both of the changes to exposure in climate and of the ability to adapt to that exposure (see Figure 1).

Science classically operates empirically, via observation, interpretation, and replication. However, having initiated a global experiment, it would not be advisable to wait decades for sufficient empirical evidence to describe the health consequences. Risk assessment must therefore be carried out in relation to future environmental scenarios. The traditional "top-down" approach is to

Figure 1

answer the question, "If climate changes like scenario X, then what will be the effect on specific health outcomes?" In contrast, "bottom-up" approaches begin with the question, "How much climate change can be tolerated?"

It is important to distinguish between "climate and health" relationships and "weather and health" relationships. Climate variability occurs on many time scales. Weather events occur at daily time scale and are associated with many health impacts (e.g., heatwaves and floods). Climate variability at other time scales also affects health. In particular, the El Niño Southern Oscillation has been shown to influence interannual variability in malaria, dengue, and other mosquito-borne diseases. Climate change is the long-term change in the average weather conditions for a particular location. Climate change will become apparent as a change in annual, seasonal, or monthly means. Thus, incremental climate change will be superimposed upon the natural variability of climate in time and space.

Natural Disasters. Climate change will increase the risk of both floods and droughts. Ninety percent of disaster victims worldwide live in developing countries, where poverty and population pressures force growing numbers of people to live in harm's way—on flood plains and on unstable hillsides. Unsafe buildings compound the risks. The vulnerability of those living in risk-prone areas is perhaps the single most important cause of disaster casualties and damage.

Water Quality and Quantity. Human health depends on an adequate supply of potable water. By reducing fresh water supplies, climate change may affect sanitation and lower the efficiency of local sewer systems, leading to increased concentrations of pathogens in raw water supplies. Climate change may also reduce the water available for drinking and washing. In developed countries, the anticipated increase in extreme rainfall events, which may be associated with the outbreaks of diarrheal diseases, may overwhelm the public water supply system. Flooding is likely to become more frequent with climate change and can affect health through the spread of disease. In vulnerable regions, the concentration of risks with both food and water insecurity can make the impact of even minor weather extremes (floods, droughts) severe for the households affected. The only way to reduce vulnerability is to build the infrastructure to remove solid waste and waste water and supply potable water. No sanitation technology is "safe" when covered by flood waters, as fecal matter mixes with flood waters and is spread wherever the flood waters go.

Food Security. Current assessments of the impact of climate change indicate that some regions are likely to benefit from increased agricultural productivity while others may suffer reductions, according to their location and dependence on the agricultural sector. The IPCC has reviewed the results of many modeling experiments that project future changes in crop yields under climate change. Climate change may increase yields of cereal grains at high and midlatitudes but may decrease yields at lower latitudes. The world's food system may be able to accommodate such regional variations at the global level, with production levels, prices, and the risk of hunger being relatively unaffected by the additional stress of climate change. However, populations in isolated areas with poor access to markets may still be vulnerable to locally important decreases or disruptions in food supply.

Heat Waves and Milder Winters. Heat stress is a direct result of exposure to high temperatures. Stressful hot weather episodes (heat waves) cause deaths in the elderly, as well as heat related illnesses such as heat stroke and heat exhaustion. A change in world climate, including an increase in the frequency and severity of heat waves, would affect the quality of life in many urban centers. Heat waves are responsible for a significant proportion of disease-related mortality in developed counties such as the United States and Australia, where the impact of weather disasters has been significantly reduced. Milder winters under climate change would reduce the excess morbidity and mortality, such as the United Kingdom, the beneficial impact may outweigh the detrimental.

Air Pollution. The air is full of particles and gases that may affect human health, such as pollen, fungal spores, and pollutants from fossil fuel emissions. Weather conditions influence air pollution via pollutant (or pollutant precursor) transport and/or formation. Exposures to air pollutants have serious public health consequences. Climate change, by changing pollen production, may affect timing and duration of seasonal allergies.

Social Dislocation. The growth in the number of refugees and displaced persons has increased markedly. Refugees represent a very vulnerable population with significant health problems. Large-scale migration is likely in response to flooding, drought, and other natural disasters. Both the local ecological disturbance caused by the extreme event and the circumstances of population displacement and resettlement would affect the risk of infectious disease outbreaks. Even displacement due to long-term cumulative environmental deterioration, including sea level rise, is associated with such health impacts.

Infectious Diseases. Vector-borne diseases are transmitted by insects (e.g., mosquitoes) and ticks that are sensitive to temperature, humidity, and rainfall. Climate change may alter the distribution of important vector species, and this may increase the risk of introducing disease into new areas. Temperature can also influence the reproduction and survival of the infective agent within the vector, thereby further influencing disease transmission in areas where the vector is already present. However, the ecology and transmission dynamics of vector-borne diseases are complex. The climate factors that could critically influence transmission need to be identified before the potential impact of a changing climate can be assessed.

Malaria is on the increase in the world at large, but particularly in Africa. In several locations around the world, malaria is reported in the twenty-first century at higher altitudes than in preceding decades, such as on the mountain plateaus in Kenya. The reason for such increases has not yet been confirmed but include population movement and the breakdown in control measures. Climate change may contribute to the spread of this major disease in the future in highlands and other vulnerable areas. Climate change impact models suggest that the largest changes in the potential for disease transmission will occur at the fringes—in terms of both latitude and altitude—of the potential malaria risk areas. The season transmission and distribution of many diseases that are transmitted by mosquitoes (dengue, yellow fever), sandflies (leishmaniasis), and ticks (Lyme disease, tick-borne encephalitis) may also be increased or decreased by climate change.

ADAPTATION AND MITIGATION

There are two responses to global climate change:

• Mitigation. Intervention or policies to reduce the emissions or enhance the sinks of greenhouse gases. The current international legal mechanism for countries to reduce their emissions is the United Nations Framework Convention on Climate Change (UNFCCC).
• Adaption. Responses to the changing climate (e.g., acclimatization in humans) and policies to minimize the predicted impacts of climate change (e.g., building better coastal defenses).

The key determinants of health—as well as the solutions—lie primarily outside the direct control of the health sector. They are rooted in areas such as sanitation and water supply, education, agriculture, trade, transport, development and housing. Unless these issues are addressed, it can be difficult to make improvements in population health and reduce vulnerability to the health impacts of climate change.

R. Sari Kovats

(see also: Environmental Determinants of Health; Geography of Disease )

Bibliography

Houghton, J. T.; Meira Filha, L. G.; Callander, B. A.; Harris, N.; Kattenberg, A.; and Maskell, K., eds.(1996). "The Science of Climate Change." Second Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge: Cambridge University Press.

McMichael, A. J., and Haines, A. (1997). "Global Climate Change: The Potential Effects on Health." British Medical Journal 315:805–809.

Patz, J. A.; McGeehin, M. A.; Bernard, S. M.; Ebi, K. L.; Epstein, P. R.; Grambsch, A.; Gubler, D. J.; and Reiter, P. (2000). "The Potential Health Impacts of Climate Variability and Change for the United States: Executive Summary of the Report of the Health Sector of the United States National Assessment." Environmental Health Perspectives 108:367–376.

Watson, R.; Zinyowera, M. C.; Moss, R. H.; and Dokken, D., eds. (1996). "Climate Change 1995. Impacts, Adaptations, and Mitigation of Climate Change: Scientific and Technical Analyses." Second Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge: Cambridge University Press.

climate change and lake levels Lakes occupy a mere 1 per cent of the Earth's land area, but they are of great importance in reconstructing its past environments. Just as mountains are areas of net erosion and sediment loss, so lakes are subject to net accumulation and sediment gain. The resulting lacustrine sediments preserve a record of past ecological conditions and geomorphic processes, which can be studied by techniques such as pollen analysis and magnetic measurements. Indeed, the build-up of sediment means that most lakes are relatively ephemeral features, as they eventually fill up and become terra firma.

As well as acting as a natural ‘archive’ for the surrounding landscape, lakes can also provide a valuable record of regional climate change via study of fluctuations in their water level and salinity. While all lakes respond to changes in water balance, the most responsive and easily quantified are those which are hydrologically closed. These ‘non-outlet’ lakes adjust their surface area and water depth dynamically in response to changing inputs in the form of rainfall and river flow, and to changing losses in the form of evaporation. Among the most important non-outlet lake basins are the Caspian Sea, Lake Eyre in Australia, and the Great Salt Lake in North America. Climatically induced fluctuations in these and many other smaller lakes have been documented in recent decades through historical records and remotely sensed imagery. Lake Chad, for example, has shrunk dramatically in area during the late twentieth century. This shrinkage has been associated with the Sahel droughts of the 1970s and 1980s (Fig. 1a). Similarly (although for different reasons), the surface area of the Aral Sea has shrunk by 40 per cent since 1960, with disastrous ecological and human consequences.

Lake hydrology

The water balance of any lake is determined by the inputs, namely precipitation onto the lake surface, inflowing streams, surface run-off and subsurface inflow, e.g. springs; and the outputs, namely evaporation from the lake surface, outflowing streams or rivers, and subsurface outflow, e.g. sink holes. If a lake is in hydrological equilibrium, inputs and outputs will balance. In a hypothetical case where groundwater exchanges are negligible, then the variable which adjusts to maintain equilibrium is surface outflow. As a lake water surplus increases, so the outflow discharge becomes greater to compensate for it. It is possible, however, that there will be no water surplus from the lake, in which case there will be no surface discharge and the lake will be hydrologically closed.

In a lake without an outlet, hydrological equilibrium is maintained, not by changes in outflow discharge, but through adjustment of the area of the lake and hence by net evaporation loss. Although the surface area is the important hydrological parameter for a lake, it will co-vary with lake water level. The former can be calculated from the latter by means of the area-depth hypsometric curve determined from the individual morphometry of each lake. It is this fundamental hydrological relationship which links climate to fluctuations in lake level, and enables the past water balance and climate to be reconstructed from palaeolimnological data (i.e. data from ancient lakes).

In reality, few lakes fulfil the assumptions set out above in acting like giant ‘rain gauges’, particularly because of groundwater flows. A lake can be closed in terms of surface hydrology, but may not be isolated from subsurface inflows and outflows. Indeed some lakes, such as those in the Corangamite region of western Victoria in Australia, are largely fed by subsurface inflows and act essentially as groundwater ‘windows’.

Lake levels and salinity

When a lake moves from being open to closed, it usually also changes from being freshwater to saline. Instead of being washed away down the outflow stream, solutes (salts) are retained within the lake water and are progressively concentrated by evaporation. As lake levels fall, the remaining water initially becomes brackish, later fully saline, and eventually hyper-saline like the Dead Sea. For this reason most arid-zone lakes are salt lakes. If, on the other hand, the level of a lake rises to the point of overflowing, salts are then flushed out of the system and the lake water becomes fresh. Past salinities are recorded in the chemical and isotopic composition of lake sediments. In a lake with an outflow, sediments such as freshwater diatomite or gyttja (organic mud) will be deposited, but if a lake moves from positive to negative water balance and the water becomes chemically concentrated, carbonates of various types will be laid down. The subsequent chemical evolution normally follows one of two main pathways, depending on the initial composition of the water. One pathway leads towards carbonate being the dominant anion and the eventual precipitation of salts such as trona (a form of hydrous sodium carbonate); the other leads towards chloride or sulphate dominance, or both, and the formation of compounds such as gypsum. The presence of these kinds of evaporite in a sedimentary sequence would be indicative of a hyper-saline, ‘playa’-type lake environment. Evaporative concentration also influences the stable isotope ratio of the sediments, and, in particular, the oxygen-isotope ratio has proved valuable in palaeolimnological reconstruction.

This alternation between fresh and saline conditions not only influences the character of the sediments deposited in the lake, but also largely controls the organisms living within it. When preserved after death in the lake-bottom sediments, their remains can be used to reconstruct past salinities and water depths, and hence also climate. Among the most useful of these indicator organisms are diatoms, ostracods, and molluscs. Figures 1(b) and 1(c) show a historically reconstructed water-level curve from Lake Naivasha in Kenya, together with the abundance of head capsules of a salt-tolerant chironomid (midge larva) from a short sediment core taken in Oloidien Bay. This arm of the main lake is isolated and saline during periods of low water level, as occurred before 1887 and again in the middle decades of the twentieth century, and at these times this chironomid species was relatively abundant. Much recent research on biological indicators has gone on to use statistical methods to reconstruct past lake salinity, especially from diatoms.

Where the regional bedrock is highly permeable, salts may not be retained in the lake, and in this case water-level fluctuations will not be accompanied by significant changes in lake salinity. This is true, for example, of the groundwater-fed lakes of the Parkers Prairie Sand Plain in Minnesota and those of the limestone massifs in Morocco's Middle Atlas mountains. At these sites, fluctuations in water depth rather than in salinity provide the main record of past changes in water balance and climate. Former high lake levels are also indicated by geomorphological evidence, notably as shoreline terraces left above the present-day water surface.

Lake-based climate reconstruction

Various methods have been used to calculate past precipitation levels for individual lake basins. Two of the principal methods are the simple water balance and combined water- and energy-balance approaches. In the first of these, the former lake area is used to calculate total evaporation losses from the water surface, which can then be balanced against the input of precipitation plus runoff from the catchment. The chief assumption that this approach makes is that past temperatures—the primary control over evaporation rates—are known. In the second approach, the evaporation term is eliminated between the water- and energy-balance equations, but it is necessary to estimate certain surface properties such as albedo and the Bowen ratio (a function of vapour pressure and temperature, which is used in the assessment of evaporation). These properties are related to the former vegetation. Both these and area-based approaches have been used to provide estimates of past precipitation in regions such as inter-tropical Africa. They indicate an average annual increase in rainfall of at least 250 mm for the Sahara, East Africa, and South Asia about 9000 (radiocarbon) years ago.

It was originally believed that the pluvial (wet) conditions responsible for climatic changes in lower latitudes matched up with the glacial (or cold) periods of high latitudes—and in a few subtropical areas, notably the American south-west, this appears to have been the case (see pluvial lakes). However, for many other regions, particularly those within the tropics, radiocarbon dating has now shown that the last wet period occurred during the Late Pleistocene and the first half of the Holocene (between 12 500 and 5000 years ago). In areas such as the Sahel, lakes which today are saline or dry contained fresh water during the early Holocene, and were able to support rich and varied aquatic faunas. The increased wetness and higher lake levels of the early Holocene appear to have been brought about by an intensification and northward displacement of the Indian Ocean monsoonal circulation, itself related to orbital forcing associated with Milankovich cycles.

Because of their typically rapid response time, lake levels also provide important palaeoclimatic data on shorter-term ‘Sub-Milankovich’ timescales. Many of the more detailed lake-level records show significant fluctuations lasting between 100 and 1000 years. These mark abrupt climatic events produced by phases of major explosive volcanism, reorganization of oceanic circulation, or similar factors. At least one of these events, the Younger Dryas, appears to have been global in nature, and is marked in the tropics by a sharp fall in lake levels, indicating a period of climatic aridity. Lakes—or at least their remains in the form of lacustrine deposits—therefore provide valuable evidence for the changing climate and hydrology of the world's dry lands over a hierarchy of timescales during the late Quaternary.

Neil Roberts

Bibliography

Almendinger, J. E. (1993) A groundwater model to explain past lake levels at Parkers Prairie, Minnesota, USA. The Holocene, 3, 105–15.
Street-Perrott, F. A. and and Harrison, S. P. (1985) Lake levels and climate reconstruction. In Hecht, A. D. (ed.) Paleoclimate analysis and modeling, pp. 291–340. John Wiley and Sons, New York.
Street-Perrott, F. A. and and Roberts, N. (1983) Fluctuations in closed-basin lakes as an indicator of past atmospheric circulation patterns. In Street-Perrott, A., Beran, M., and Ratcliffe, R. A. S. (eds) Conference on variations in the global water budget, pp. 331–45. Reidel, Dordrecht.

Global Climate Change

Energy from the Sun passes through the atmosphere as light and is absorbed by soil, rock, and water at the surface of Earth. The energy is reradiated as heat and absorbed in the atmosphere by greenhouse gases, including carbon dioxide (CO₂), water vapor, methane, ozone, nitrous oxide, and the human-made chemicals chlorofluorocarbons (CFCs). This atmospheric warming is called the greenhouse effect; without it Earth's average global temperature would be about –18 degrees Celsius (0 degrees Fahrenheit). Greenhouse gases are added to the atmosphere by natural events including volcanic eruptions, the decay and burning of organic matter, and respiration by animals. They are also removed from the atmosphere. CO₂ is absorbed by seawater and stored in plant tissue. When plants die and gradually are transformed into fossil fuels—coal, oil, natural gas—deep in the earth, their CO₂ is stored with them. The removal of greenhouse gases from the atmosphere keeps the planet from overheating.

Climate History

Besides the concentrations of greenhouse gases in the atmosphere, other factors affect global climate including Earth's orbital behavior, the positions and topography of the continents, the temperature structure of the oceans, and the amount and types of life. During much of Earth's history the climate was warm and humid with ice-free poles; global average temperatures were about 5 degrees Celsius (9 degrees Fahrenheit) higher than today. Several times glaciers covered the higher latitudes, most recently during the Pleistocene (1.6 million to 10,000 years ago), when up to 30 percent of the land was covered by ice. During the four glacial advances of the Pleistocene, average global temperature was 5 degrees Celsius lower than today and 10 degrees Celsius (18 degrees Fahrenheit) lower than the ancient global average. During the three interglacial periods, global temperature was a degree or two warmer than today. Many scientists think that Earth is in an interglacial period, and the ice sheets will return.

Since the peak of the last glacial advance 18,000 years ago, average global temperature has risen 4 degrees Celsius (7 degrees Fahrenheit), including 1 degree Celsius (1.8 degrees Fahrenheit) since the beginning of the Industrial Revolution. It is difficult to know how much of the recent warming is the result of the end of the Pleistocene and how much is the result of human activities that add greenhouse gases to the atmosphere. CO₂ is the most abundant greenhouse gas, a by-product of burning fossil fuels and modern forests. In the early twenty-first century, there is greater than 30 percent more CO₂ in the atmosphere than in 1850. There have also been significant increases in methane and CFCs. Some projections show a doubling of CO₂ over preindustrial levels by 2050 and additional increases in methane. (CFCs are being phased out by international agreement because they destroy Earth's protective ozone layer.)

Adding greenhouse gases to the atmosphere is like throwing another blanket on Earth; the consequent rise in global temperature is known as global warming. Since climate is a complex system and climate models are difficult to construct, scientists can only speculate on the effect large increases in greenhouse gases will have on global climate. Some models show average global temperature increasing as much as 5 degrees Celsius by 2100. Any temperature increase will not be uniform. Since ocean water absorbs more heat than land, the Southern Hemisphere (which has more water) will warm less than the Northern. Atmospheric circulation patterns will bring the greatest warming, as much as 8 to 10 degrees Celsius (14 to 18 degrees Fahrenheit), to the poles.

Possible Consequences

A rapid increase in global average temperature could have profound effects on social and natural systems. Warmer temperatures would cause ocean water to expand and polar ice caps to melt, increasing sea level by as much as 50 centimeters (1.6 feet) by 2100. This would flood coastal regions, where about one-third of the world's population lives and where an enormous amount of economic infrastructure is concentrated. It would destroy coral reefs, accelerate coastal erosion, and increase salinity to coastal groundwater aquifers. Warmer temperatures would allow tropical and subtropical insects to expand their ranges, bringing tropical diseases such as malaria, encephalitis, yellow fever, and dengue fever to larger human populations. There would be an increase in heat-related diseases and deaths. Agricultural regions might become too dry to support crops, and food production all over the world would be forced to move north; this would result in a loss of current cropland of 10 to 50 percent and a decline in the global yield of key food crops of from 10 to 70 percent.

Wild plant and animal species would need to move poleward 100 to 150 square kilometers (60 to 90 miles) or upward 150 meters (500 feet) for each 1 degree Celsius rise in global temperature. Since most species could not migrate that rapidly and since development would stop them from colonizing many new areas, much biodiversity would be lost. The decrease in the temperature difference between the poles and the equator would alter global wind patterns and storm tracks. Regions with marginal rainfall levels could experience drought, making them uninhabitable. Overall, since warmer air holds more moisture, an increase in global air and sea temperatures would increase the numbers of storms. Higher sea surface temperatures would increase the frequency and duration of hurricanes and El Niño events.

Many scientists believe that global warming is the most serious threat to our planet. By 2025 the world's energy demand is projected to be 3.5 times greater than in 1990, with annual CO₂ emissions nearly 50 percent higher. Thus far, attempts at international agreements to curb the emissions of greenhouse gases (for example, the Kyoto Protocol) have failed. This is due to several factors: (1) the scientific uncertainty of the role humans play in global warming; (2) the lifestyle changes necessary to reduce fossil fuel consumption in developed nations; (3) the possible slowdown in the economic development of developing nations; and (4) the need for true international cooperation. A high-technology alternative to decreasing greenhouse gas emissions is to sequester CO₂. Experiments are underway to inject liquid CO₂ deep into the earth, thereby effectively removing it from Earth's carbon cycle.

see also Biogeochemical Cycles; Carbon Cycle; Ecological Research, Long-Term; Ecosystem; Extinction; Tundra

Dana Desonie

Bibliography

Drake, Frances. Global Warming: The Science of Climate Change. Edward Arnold, 2000.

Stevens, William K. The Change in the Weather: People, Weather, and the Science of Climate. Delta, 2001.

climate change. The earth's climate continually fluctuates as a result of the earth's elliptical orbit around the sun. When the earth is closer to the sun it receives more radiant energy. Long-term analyses of climate based on deep ocean deposits and ice cores from the Antarctic and Greenland ice caps show there are climatic cycles lasting 25,000, 40,000, and 100,000 years, which have resulted in the oscillations between glacial and interglacial periods. There are also millennial scale oscillations: the Vikings were able to colonize Greenland in their longboats during a mild climatic period about a thousand years ago, and from there visited Vinland, whereas there was a mini-ice age in the early 18th century when goose fairs were regularly held on the frozen Thames. Recently, 70-year cycles have been identified in the North Atlantic Oscillation. When pressure is high over the Azores, and low over Iceland, the flow of the Gulf Stream increases, keeping north-western Europe's climate mild. In the opposite phase the flow reduces and the climate of Europe is cooler. Similar oscillations have been identified in the North Pacific and both may be linked with El Niño events.

The gases in past atmospheres can be measured in the gas bubbles trapped deep in the ice caps. These show that as climate has oscillated so have the amounts of carbon dioxide in the atmosphere. When CO₂ concentrations are high, more of the incoming radiant heat from the sun is trapped in the atmosphere—the so-called greenhouse effect. Other gases such as methane and CFCs (chlorofluorocarbons) are many more times more effective in trapping the heat, but their concentrations are rising more slowly. This link between carbon dioxide and climate is well established, but exactly what causes it is a matter of controversy. Since the start of the Industrial Revolution in the 18th century the burning of fossil fuels has increased the carbon dioxide content of the atmosphere from pre-industrial levels of 240 parts per million (ppm) to present levels of over 370 ppm. If we continue to burn oil and coal at the present rates it will rise to 500–600 ppm by the end of the century. As more and more of the sun's energy gets trapped in the atmosphere, the earth's climate will warm, but not uniformly across the planet; polar regions will warm more than the tropics. Although some countries will benefit from such shifts most will lose out. Sea levels will rise as the polar ice caps melt and in 2003 Australian scientists produced fresh evidence showing that the area of sea ice surrounding Antarctica has shrunk by 20% in the last 50 years. Weather patterns will become less predictable with bigger and more destructive storms. Shifts in rainfall will increase flooding in some places, but cause droughts and desertification in others. The pattern of ocean currents will shift, and close links between ocean currents and local climate are contributing to the uncertainties of predicting the outcome of global climate change and are discouraging governments from taking the social and economic sacrifices needed to halt the rise in carbon dioxide emissions.

Already ocean plankton off the west coast of Britain has become more typical of warmer seas, and there have been sharp decline in the abundances of keystone species, such as the crustacean copepod (Calanus finmarchicus), which has been linked to the catastrophic decline in cod stocks. Climate will continue to fluctuate because many of the causes are planetary, but the implications of the changes will become increasingly threatening as world population grows and industralization increases globally, one of the most serious environmental issues facing us today.

M. V. Angel

climate change and laminated sediments Laminated sediments can often be used to infer short-term (e.g. seasonal to annual) climatic or oceanographic fluctuations, since each lamina is comprised of sediment deposited under unique conditions. For example, annual winter rains lead to the deposition of silty laminae in the Santa Barbara Basin, off the coast of California. A first step in deciphering past climate records from laminated sediments is thus to identify the sedimentary components of successive laminae. This can be done in situ by using data from sediment traps. Sediment traps continuously collect descending sediment (e.g. biogenic or terrigenous material), collection periods sometimes being for a year or more. Over such periods of time, any seasonal changes in sedimentation patterns will be observed, such as increasing silt deposition due to increased rainfall or glacial discharge, or abundant diatoms due to algal blooms. Such observations can be compared with the laminated sedimentary record observed directly from the sediments. Advances in scanning electron microscope (SEM) back-scattered electron imaging techniques make it possible to examine micron-scale lamina variations in rock sections; these can provide detailed records of sub-annual deposition. Data from both sediment-trap studies and SEM examination of core intervals can be compared with his-torical climate records. The interpretations provided by such studies greatly aid both the study of laminated sediments that predate historical records and the understanding of longer-term variations in palaeoclimate and palaeoceanography.

Many studies of hand or microscopic specimens analyse specific features of laminated sedimentary sections; for example, colour or grey-scale changes. The occurrence through time of any variations can be documented, and their statistical significance investigated. The periodicity of any significant frequencies is in some instances related to climatic or oceanographic phenomena; for example, periodicities between 2 and 7 years correspond to the periodicity of the El Niño-Southern Oscillation (ENSO).

Laminated sediments occur in various depositional environments, but primarily where water is poorly oxygenated at the sea or lake bed, that is, where water is either dysaerobic (oxygen-poor) or anoxic (no oxygen present at all). Under these conditions, the activity of benthonic organisms is severely or entirely curtailed, and individual laminae are preserved. Typical locations where this occurs are: (1) basins with limited deep water replacement; for example, Californian Borderland Basins; (2) areas of intense upwelling, which experience high organic-carbon sedimentation rates owing to increased primary productivity, for example, the Gulf of California, and off the coast of Peru; and (3) lakes, where stratification of the water column often prevents oxygenation of the deeper waters; for example, Arctic lakes and high alpine lakes. More recently, deep-sea laminated diatom oozes have been recovered from the Pacific and Atlantic Oceans, making it possible to investigate short time-scale oceanographical processes such as Miocene- Pliocene El Niño-like events in the eastern equatorial Pacific. In the ancient sedimentary record, laminated sediments are preserved in marine sediments from periods of ocean anoxia, and similarly from freshwater and saline lakes, providing insights into ancient climate and ocean change.

R. B. Pearce

Bibliography

Kemp, A. E. S. (ed.) (1996) Palaeoclimatology and palaeoceanography from laminated sediments. Geological Society Special Publication No. 116.

climate records

cli·mate change • n. long-term, significant change in the climate of an area or of the earth, usually seen as resulting from human activity. Often used as a synonym for global warming.