CLIMATE AND LIFE

Climate and life Coevolution of climate and life It has been well established that life and climate interact and that they have mutually altered each other over geologic history. In a sense, they have undergone coevolution, a term coined by the American biologists Paul R. Ehrlich and Peter H. Raven to describe the process whereby two or more species depend on the interactions between them. It might be said that the coevolution of life and climate during the past 4,000,000,000 years of Earth history is an expression for the incredibly complex mixture of forces causing climatic change. The developing atmosphere According to a widely held theory, the release of gases from the interior of the Earth during volcanic eruptions, along with other primitive processes, produced the planet's early atmosphere. Such processes determined the Earth's albedo (reflectivity to incoming solar radiation) and its heat-trapping greenhouse properties (see below Causes of climatic change: The greenhouse effect). Eventually carbon, hydrogen, oxygen, and nitrogenthe basic chemical constituents of lifewere synthesized into primitive organic molecules from which early life forms subsequently evolved. This probably occurred during the first 500,000,000 years of Earth history. Anaerobic bacteria (those capable of living in the absence of oxygen) thrived in this early atmosphere, which consisted primarily of water vapour and carbon dioxide, with trace amounts of ammonia, methane, and various other gases. These organisms survived during this period because they found their ecological nichei.e., an environment having suitable chemical and physical conditions. Climate and life Impact of climate on human life General observations The interest in long-term atmospheric changes induced by human activities (anthropogenic changes) sometimes overshadows the fact that the natural short-term variability of climate also must be considered in assessing the influence of climate on society. Average, or mean, conditions are not a sufficient representation of the role climate plays in environmental and human affairs. Biologic systems are strongly affected by extremes in weather and climate. Accordingly, the character of climatic variability has an important effect on both natural and human communities and on the health of those communities. A few obvious examples of the negative impact that natural climatic variability has had in the United States include the following. During the 1930s, crop and soil damage was the worst ever sustained in the Dust Bowl region of the central United States, and a forced emigration from the Great Plains reinforced the economic problems brought on by the Great Depression. In the winter of 1977, drought in the west and extreme cold in the east created economic disruptions estimated at billions of dollars. In the hot summer of 1980, the elderly, poor, and infirm in the south central states suffered abnormally high rates of mortality and morbidity. In addition, yields of such summer crops as corn and soybeans were greatly reduced. No thorough analysis of the distribution of losses and benefits resulting from climatic fluctuations of this sort has been made, but estimates put the net loss in the range of billions of dollars. Experience with natural climatic fluctuations should be drawn upon both to develop policy measures with which to minimize society's vulnerability to anthropogenic climatic changes and to take advantage of new climatic resources. Atmospheric impact assessments for dealing with long-term problems usually involve an extrapolation of experience with the shorter-term effects of natural variability. Such experience, however, suggests that population size and the distribution and availability of resources mediate how natural climatic variability already affects society regardless of the existence of long-term climatic trends. Thus, it is necessary to project how both climatic and societal factors will evolve and interact. Day-to-day variability is an important atmospheric factor because society and the living environment respond nonlinearly to such variations. For example, the freezing point is a threshold below which small changes may have major effects on vegetation. Similarly, above certain temperatures, plant, animal, and human vitality can be seriously threatened. Since present-day climatic variability could well be superimposed on long-term climatic trends, one of the most significant effects of a seemingly small shift in the mean climate could be a large change in the frequency of harmful extreme situations. Weather and technology In the latter part of the 1800s during the homesteading of the Great Plains, farmers were for the most part successful because the weather was generally favourable. Severe drought conditions in the 1890s, however, shattered the illusion that life on the plains was necessarily good and drove out many settlers. This pattern of boom-and-bust farming recurred several times: good weather and high-production years were followed by periods of drought, economic ruin, and serious soil erosion. The worst drought and resultant soil degradation occurred during the 1930s in the area of the Dust Bowl. Average wheat and corn yields fell by as much as 75 percent. Worse, millions of tons of valuable topsoil were lost. As a result of this climate-induced disaster, the federal government established the U.S. Soil Conservation Service to help farmers protect the soil. Decades later new crop strains better adapted to regional climates were developed, and irrigation and chemical fertilizers were made available to take advantage of the new genetic strains and to promote production. These advances, together with the development of such technological aids as tractors, mechanical harvesters, irrigation pumps, and agrochemicals (e.g., pesticides and herbicides), have increased the productivity (average yield of grain per harvested area) of the Great Plains by 200 to 300 percent since the 1930s. Total production (productivity multiplied by total harvested area) also has risen. In addition, the amount of year-to-year variability in yield, relative to the average yield, for most grains has decreased over time. Yet, while the relative variability (year-to-year variability as a percentage of long-term average yield) has generally declined for crops in recent years, the absolute variability (year-to-year variability in yield by itself) has increased on the whole in spite of all the technological advances. This has led to an ongoing debate about the relative role of climate versus the role of technological advances in influencing both average-yield and yield-variability trends during the 1960s and '70s. While agrotechnology appears to have been the prime factor behind the general increase in annual grain yield, there is some doubt as to its actual contribution to the favourable decrease of variability in relative yield. Some investigators, most notably the American geographer Richard Warrick, have argued that this decrease cannot be entirely attributed to modern farming practices. Only if the weather anomalies of the 1960s and '70s had been as bad as those from roughly 1930 to the late 1950s could agrotechnology be credited with the reduced impact of climatic stress on crop yields. Warrick and his associates compiled an index of the severity of summer droughts in the Great Plains in the period 193177, which indicates quite clearly that since the late 1950s weather conditions have not been bad enough to test the hypothesis that technological advances have truly reduced the annual variations of grain yield. Given this situation, it is risky to assume that modern technology in the breadbasket of North America can stably maintain yearly productivity. Climate and life Impact of human activities on climate Three distinct atmospheric problems have been debated intensely since about the mid-1970s, though two of them are quite old issues: the possible reduction of stratospheric ozone from chemical emissions; the generation of acid rain; and climatic change stemming from the greenhouse effect. What these three problems have in common is quite simple: they all (1) are complex and punctuated by large uncertainties, (2) could be long-lasting, (3) transcend state and even national boundaries, (4) may be difficult to reverse, (5) are inadvertent by-products of widely supported economic activities, and (6) may require substantial investments of present resources to hedge against the prospect of large future environmental changes. Ozone depletion Of these problems, the only one to have received any substantial public policy action is that centring on the reduction of stratospheric ozone. Ironically, it is perhaps the easiest of the problems to reverse. The importance of the stratospheric ozone layer in shielding the Earth's surface from the harmful effects of solar ultraviolet radiation has been recognized for several decades. It was not until the early 1970s, however, that scientists began actually to grapple with the fact that even relatively small decreases in the stratospheric ozone concentration can have a serious impact on human healthan increased incidence of skin cancer, particularly among fair-skinned peoples. Plans in the United States, Great Britain, and France to build a commercial fleet of supersonic aircraft triggered much heated discussion over the potential reduction of the ozone layer by the exhaust gases (e.g., nitric oxide) emitted by such high-altitude planes. The debate in turn stimulated intensive scientific research on the stratosphere, which resulted in new findings and new concerns. By the mid-1970s, various U.S. investigators had determined that chlorofluorocarbons (CFCs), widely employed as propellants in aerosol spray cans, could reduce the amount of stratospheric ozone significantly. A temporary ban was imposed on the use of certain CFCs in the United States, but only after much emotional debate among environmental and industrial scientists, reports by the National Academy of Sciences, and the development by industry of economically viable substitutes for spray-can propellants. (For a more detailed discussion of this issue, see atmosphere: Depletion of stratospheric ozone.) Atmospheric humidity and precipitation Atmospheric humidity, which is the amount of water vapour or moisture in the air, is another leading climatic element, as is precipitation. All forms of precipitation, including drizzle, rain, snow, ice crystals, and hail, are produced as a result of the condensation of atmospheric moisture to form clouds in which some of the particles, by growth and aggregation, attain sufficient size to fall from the clouds and reach the ground. Atmospheric humidity At 30 C, 4 percent of the volume of the air may be occupied by water molecules, but where the air is colder than -40 C, less than one-fifth of 1 percent of the air molecules can be water. Although the water vapour content may vary from one air parcel to another, these limits can be set because vapour capacity is determined by temperature. Temperature has profound effects upon some of the indices of humidity regardless of the presence or absence of vapour. The connection between an effect of humidity and an index of humidity requires simultaneous introduction of effects and indices. Vapour in the air is a determinant of weather because it first absorbs the thermal radiation that leaves and cools the Earth and then emits thermal radiation that warms the planet. Calculation of absorption and emission requires an index of the mass of water in a volume of air. Vapour also affects the weather because, as indicated above, it condenses into clouds and falls as rain or other forms of precipitation. Tracing the moisture-bearing air masses requires a humidity index that changes only when water is removed or added. Atmospheric humidity and precipitation Precipitation Precipitation is one of the three main processes (evaporation, condensation, and precipitation) that constitute the hydrologic cycle, the continual exchange of water between the atmosphere and the surface of the Earth. Water is evaporated from ocean, land, and freshwater surfaces, is carried aloft as vapour by the air currents, condenses to form clouds, and ultimately is returned to the Earth's surface as precipitation. The average global stock of water vapour in the atmosphere is equivalent to a layer of water 2.5 centimetres (one inch) deep covering the whole Earth. Because the Earth's average annual rainfall is about 100 centimetres, the average time that the water spends in the atmosphere, between its evaporation from the surface and its return as precipitation, is about 1/40 of a year, or about nine days. Of all the water vapour that is carried at all heights across a given region by the winds, only a small percentage is converted into precipitation and reaches the ground in that area. In deep and extensive cloud systems, the conversion is more efficient, but even in thunderclouds the quantities of rain and hail released amount to only some 10 percent of the total moisture entering the storm. In the measurement of precipitation, it is necessary to distinguish between the amountdefined as the depth of precipitation (calculated as though it were all rain) that has fallen at a given point during a specified interval of timeand the rate or intensitywhich specifies the depth of water that has fallen at a point during a particular interval of time. Persistent moderate rain, for example, might fall at an average rate of five millimetres per hour and thus produce 120 millimetres of rain in 24 hours. A thunderstorm might produce this total quantity of rain in 20 minutes, but at its peak intensity the rate of rainfall might become much greaterperhaps 120 millimetres per hour, or two millimetres per minute, for a minute or two. The amount of precipitation falling during a fixed period is measured regularly at many thousands of places on the Earth's surface by rather simple rain gauges. Measurement of precipitation intensity requires a recording rain gauge, in which water falling into a collector of known surface area is continuously recorded on a moving chart or a magnetic tape. Investigations are being carried out on the feasibility of obtaining continuous measurements of rainfall over large catchment areas by means of radar. Apart from the trifling contributions made by dew, frost, and rime and by desalination plants, the sole source of fresh water for sustaining rivers, lakes, and all life on Earth is provided by precipitation from clouds. Precipitation is therefore indispensable and overwhelmingly beneficial to humankind, but extremely heavy rainfall can cause great harm: soil erosion, landslides, and flooding. And hailstorm damage to crops, buildings, and livestock can prove very costly. Origin of precipitation in clouds Cloud formation Clouds are formed by the lifting of damp air, which cools by expansion as it encounters the lower pressures existing at higher levels in the atmosphere. The relative humidity increases until the air becomes saturated with water vapour, then condensation occurs on any of the aerosol particles suspended in the air. A wide variety of these exist in concentrations ranging from only a few per cubic centimetre in clean maritime air to perhaps 1,000,000 per cubic centimetre (16,000,000 per cubic inch) in the highly polluted air of an industrial city. For continuous condensation leading to the formation of cloud droplets, the air must be slightly supersaturated. Among the highly efficient condensation nuclei are sea-salt particles and the particles produced by combustion (e.g., natural forest fires and man-made fires). Many of the larger condensation nuclei over land consist of ammonium sulfate. These are produced by cloud and fog droplets absorbing sulfur dioxide and ammonia from the air. Condensation onto the nuclei continues as rapidly as water vapour is made available through cooling; droplets about 10 micrometres in diameter are produced in this manner. These droplets constitute a nonprecipitating cloud. Atmospheric pressure and wind Atmospheric pressure Atmospheric pressure and wind are both significant controlling factors in the Earth's weather and climate. Although these two physical variables may at first glance appear to be quite different, they are in fact closely related. Wind exists because of horizontal and vertical differences (gradients) in pressure, yielding a correspondence that often makes it possible to use the pressure distribution as an alternative representation of atmospheric motions. Pressure is the force exerted on a unit area, and atmospheric pressure is equivalent to the weight of air above a given area on the Earth's surface or within its atmosphere. This pressure is usually expressed in millibars (one mb equals 1,000 dynes per square centimetre) or in kilopascals (kPa; one kPa equals 10,000 dynes per square centimetre). Distributions of pressure on a map are depicted by a series of curved lines called isobars, each of which connects points of equal pressure. Figure 2: World distribution of mean sea-level pressure (in millibars) for January, and primary and Figure 3: World distribution of mean sea-level pressure (in millibars) for July, and primary and At sea level, the mean pressure is about 1,000 millibars (100 kilopascals), varying by less than 5 percent from this value at any given location or time. Figures 2 and 3 show mean sea-level pressure for the mid-winter and mid-summer months. Since these charts represent average values over several days, pressure features that are relatively consistent day after day emerge, while more transient, short-lived features are removed. Those that remain are known as semi-permanent pressure centres and are the source regions for major, relatively uniform bodies of air known as air masses. Warm, moist maritime tropical (mT) air forms over tropical or subtropical ocean waters in association with the high-pressure regions prominent there. Cool, moist maritime polar (mP) air, on the other hand, forms over the colder subpolar ocean waters just south and east of the large, winter oceanic low-pressure regions. Over the continents, cold, dry continental polar (cP) air forms in the high-pressure regions that are especially pronounced in winter, while hot, dry continental tropical (cT) air forms over hot, desertlike continental domains in summer in association with low-pressure areas sometimes called heat lows. Figure 2: World distribution of mean sea-level pressure (in millibars) for January, and primary and Figure 3: World distribution of mean sea-level pressure (in millibars) for July, and primary and A closer examination of Figures 2 and 3 reveals some interesting features. First, it is clear that sea-level pressure is dominated by closed high- and low-pressure centres, which are largely caused by differential surface heating between low and high latitudes and between continental and oceanic regions. High pressure tends to be amplified over the colder surface features. Second, because of seasonal changes in surface heating, the pressure centres exhibit seasonal changes in their characteristics. For example, the Siberian High, Aleutian Low, and Icelandic Low that are so prominent in the winter virtually disappear in summer as the continental regions warm relative to surrounding bodies of water. At the same time, the Pacific and Atlantic highs amplify and migrate northward. At altitudes well above the Earth's surface, the monthly average pressure distributions show much less tendency to form in closed centres but rather appear as quasi-concentric circles around the poles. This more symmetrical appearance reflects the dominant role of meridional radiative heating/cooling differences. Excess heating in tropical latitudes, compared to the polar, produces higher pressure at upper levels in the tropics. In addition, the greater heating/cooling contrast in winter yields stronger pressure differences during this season. Perfect symmetry is interrupted by superimposed wavelike disturbances. These are most pronounced over the Northern Hemisphere, with its more prominent land-ocean contrasts and orographic features. Wind Relationship of wind to pressure and governing forces The changing wind patterns are governed by Newton's second law of motion, which states that the sum of the forces acting on a body equals the product of the mass of that body and the acceleration caused by those forces. The basic relationship between atmospheric pressure and horizontal wind is revealed by disregarding friction and any changes in wind direction and speed to yield the mathematical relationship where u is the zonal wind speed (+ eastward), v the meridional wind speed (+ northward), f = 2w sin f (Coriolis parameter), w the angular velocity of the Earth's rotation, f the latitude, r the air density (mass per unit volume), p the pressure, and x and y the distances toward the east and north, respectively. This simple, non-accelerating flow is known as geostrophic balance and yields a motion field known as the geostrophic wind. Equation (1) expresses, for both the x and y directions, a balance between the force created by horizontal differences in pressure (the horizontal pressure-gradient force) and the force that results from the Earth's rotation (the Coriolis force). The pressure-gradient force expresses the tendency of pressure differences to effectuate air movement from higher to lower pressure. The Coriolis force is a more complicated concept that arises because the air motions are observed on a rotating, nearly spherical body. The total motion of a parcel of air has two parts: (1) the motion relative to the Earth as if the Earth were fixed, and (2) the motion given to the parcel by the planet's rotation. If an observer viewed the atmosphere from a fixed point in space, the rotation of the Earth would be apparent to him and he would observe the total motion. An observer on the ground, however, sees and measures only the relative motion and, as he also is rotating, cannot see directly the rotational motion applied by the Earth. Instead, he sees the effect of the rotation as a deviation applied to the relative motion. The quantity that causes this deviation is the so-called Coriolis force. More specifically, what the observer sees is a deflection of the relative motion to the right in the Northern Hemisphere and to the left in the Southern Hemisphere. Of particular significance in this simple model of windpressure relationships is the fact that the geostrophic wind blows in a direction parallel to the isobars with the low pressure on the observer's left as he looks downwind in the Northern Hemisphere and on his right in the Southern Hemisphere. Furthermore, the wind speed increases as the distance between isobars decreases (or pressure gradient increases). Curvature (i.e., changes in wind direction) can be added to this model with relative ease in a flow representation known as the gradient wind. The basic windpressure relationships, however, remain qualitatively the same. Of greatest importance is the fact that actual observed winds tend to behave much as the geostrophic- or gradient-flow models predict in most of the atmosphere. The most notable exceptions are in low latitudes where the Coriolis parameter becomes very small and thus equation (1) cannot be used to provide a reliable wind estimate, and in the lowest kilometre of the atmosphere where friction becomes important. The friction induced by flow over the underlying surface reduces the wind speed and alters the simple balance of forces such that the wind blows with a component toward lower pressure. Atmospheric pressure and wind Monsoons Figure 2: World distribution of mean sea-level pressure (in millibars) for January, and primary and Figure 3: World distribution of mean sea-level pressure (in millibars) for July, and primary and A close examination of Figures 2 and 3 reveals particularly strong seasonal pressure variations over continents. Such seasonal fluctuations, commonly called monsoons, are more pronounced over land surfaces because these surfaces are subject to more significant seasonal temperature variations. Monsoons blow for approximately six months from the northeast and six months from the southwest, principally in southern Asia and parts of Africa. Summer monsoons have a dominant westerly component and a strong tendency to converge, rise, and produce rain. Winter monsoons have a dominant easterly component and a strong tendency to diverge, subside, and cause drought. Both are the result of differences in annual temperature trends over land and sea. The Indian monsoon At the Equator, the area near India is unique in that dominant or frequent westerly winds occur at the surface almost constantly throughout the year; the surface easterlies reach only to 20 N in February, and even then they have a very strong northerly component. They soon retreat northward, and drastic changes take place in the upper-air circulation. This is a time of transition between the end of one monsoon and the beginning of the next. Late in March the high-sun season reaches the Equator and moves farther north. With it go atmospheric instability, convectional (rising, turbulent) clouds, and rain. The westerly subtropical jet stream (see below Jet streams) still controls the flow of air across northern India and the surface winds are northeasterlies. As the high-sun season moves northward during April, India becomes particularly prone to rapid heating because the highlands to the north protect it from any incursions of cold air. In May the southwesterly monsoon is well established over Sri Lanka. There are three distinct regions of relative upper tropospheric warmthnamely (1) above the southern Bay of Bengal, (2) above the highlands of Tibet, and (3) across the still, dry trunks of the peninsulas. The relatively warm area above the southern Bay of Bengal occurs mostly at the 500100 millibar level. It does not appear at a lower level and is probably caused by the release of condensation heat (associated with the change from water vapour to liquid water) at the top of towering cumulonimbus clouds along the advancing intertropical convergence. In May the dry surface of Tibet (above 4,000 metres) absorbs and radiates heat that is readily transmitted to the air immediately above. At about 6,000 metres an anticyclonic cell arises, causing a strong easterly flow in the upper troposphere above northern India. The subtropical jet stream suddenly changes its course to the north of the anticyclonic ridge and the highlands, though it may occasionally reappear southward of them for very brief periods. This change of the upper tropospheric circulation above northern India from westerly jet to easterly flow coincides with a reversal of the vertical temperature and pressure gradients between 600 and 300 millibars. On many occasions the easterly aloft assumes jet force. It anticipates by a few days the burst, or onset, of the surface southwesterly monsoon some 1,500 kilometres farther south, with a definite sequential relationship, although the exact cause is not known. Because of India's inverted triangular shape, the land is heated progressively as the Sun moves northward. This accelerated spread of heating, combined with the general direction of heat being transported by winds, results in a greater initial monsoonal activity over the Arabian Sea (at late spring time), where a real frontal situation often occurs, than over the Bay of Bengal. The relative humidity of coastal districts in the Indian region rises above 70 percent and some rain occurs. Above the heated land the air below 1,500 metres becomes unstable, but it is held down by the overriding easterly flow. This does not prevent frequent thunderstorms in late May. During June the easterly jet becomes firmly established at 150 to 100 millibars. It reaches its greatest speed at its normal position to the south of the anticyclonic ridge, at about 15 N from China through India. In Arabia, it decelerates and descends to the middle troposphere (3,000 metres). A stratospheric belt of very cold air, analogous to the one normally found above the intertropical convergence near the Equator, occurs above the anticyclonic ridge, across southern Asia at 3040 N and above the 6,000-metre (500-millibar) level. These upper air features that arise so far away from the Equator are associated with the surface monsoon and are absent when there is no monsoonal flow. The position of the easterly jet controls the location of monsoonal rains, which occur ahead and to the left of the strongest winds and behind them to the right. The surface flow, however, is a strong, southwesterly, humid, and unstable wind that brings humidities of more than 80 percent and heavy, squally showers that are the burst of the monsoon. The overall pattern of the advance follows a frontal alignment, but local episodes may differ considerably. The amount of rain is variable from year to year and place to place. Most spectacular clouds and rain occur against the Western Ghats, where the early monsoonal airstream piles up against the steep slopes, then recedes, and piles up again to a greater height. Each time it pushes thicker clouds upward until wind and clouds roll over the barrier and, after a few brief spells of absorption by the dry inland air, cascade toward the interior. The windward slopes receive from 2,000 to 5,000 millimetres of rain in the monsoon season. Various factors, and especially topography, combine to make up a complex regional pattern. Oceanic air flowing toward India below 6,000 metres is deflected in accordance with the Coriolis effect. The converging, moist oncoming stream becomes unstable over the hot land and is subject to convectional turmoil. Towering cumulonimbus clouds rise thousands of metres, producing violent thunderstorms and releasing latent heat in the surrounding air. As a result, the upper tropospheric warm belt migrates northwestward from the ocean to the land. The main body of air above 9,000 metres maintains a strong easterly flow. Later, in June and July, the monsoon is strong and well-established to a height of 6,000 metres (less in the far north), with occasional thickening to 9,000 metres. Weather conditions are cloudy, warm, and moist all over India. Rainfall varies between 400 and 500 millimetres, but topography introduces some extraordinary differences. On the southern slopes of the Khasi Hills at only 1,300 metres, where the moist airstreams are lifted and overturned, Cherrapunji has an average rainfall of 2,730 millimetres in July, with record totals of 897 millimetres in 24 hours in July 1915, more than 9,000 millimetres in July 1861, and 16,305 millimetres in the monsoon season of 1899. Over the Ganges Valley the monsoon, deflected by the Himalayan barrier, becomes a southeasterly air flow. By then the upper tropospheric belt of warmth from condensation has moved above northern India, with an oblique bias. The lowest pressures prevail at the surface. It is mainly in July and August that waves of low pressure appear in the body of monsoonal air. Fully developed depressions appear once or twice a month. They travel from east to west more or less concurrently with high-level easterly waves and bursts of speed in the easterly jet, causing local strengthening of the low-level monsoonal flow. The rainfall consequently increases and is much more evenly distributed than it was in June. Some of the deeper depressions become tropical cyclones before they reach the land, and these bring torrential rains and disastrous floods. A totally different development arises when the easterly jet moves farther north than usual because the monsoonal wind rising over the southern slopes of the Himalayas brings heavy rains and local floods. The weather over the central and southern districts, however, becomes suddenly drier and remains so for as long as the abnormal shift lasts. The opposite shift is also possible, with mid-latitude upper air flowing along the south face of the Himalayas and bringing drought to the northern districts. Such dry spells are known as breaks of the monsoon. Those affecting the south are similar to those experienced on the Guinea coast during extreme northward shifts of the wind belts (as later discussed), whereas those affecting the north are due to an interaction of the middle and low latitudes. The southwest monsoon over the lower Indus Plain is only 500 metres thick and does not hold enough moisture to bring rain. On the other hand, the upper tropospheric easterlies become stronger and constitute a true easterly jet stream. Western Pakistan, Iran, and Arabia remain dry (probably because of divergence in this jet) and thus become the new source of surface heat. By August the intensity and duration of sunshine have decreased, temperatures begin to fall, and the surge of southwesterly air diminishes spasmodically almost to a standstill in the northwest. Cherrapunji still receives over 2,000 millimetres of rainfall at this time, however. In September dry, cool, northerly air begins to circle the west side of the highlands and spread over northwestern India. The easterly jet weakens and the upper tropospheric easterlies move much farther south. Because the moist southwesterlies at lower levels are much weaker and variable, they are soon pushed back. The rainfall becomes extremely variable over most of the region, but showers are still frequent in the southeastern areas and over the Bay of Bengal. By early October variable winds are very frequent everywhere. At the end of the month the entire Indian region is covered by northerly air and the winter monsoon takes shape. The surface flow is deflected by the Coriolis force and becomes a northeasterly flow. This causes an OctoberDecember rainy season for the extreme southeast of the Deccan (including the Madras coast) and eastern Sri Lanka, which cannot be explained by topography alone because it extends well out over the sea. Tropical depressions and cyclones are important contributing factors. Most of India thus begins a sunny, dry, and dusty season. The driest period comes in November in the Punjab; December in Central India, Bengal, and Assam; January in the northern Deccan; and February in the southern Deccan. Conversely, the western slopes of the Karakoram and Himalayas are then reached by the mid-latitude frontal depressions that come from the Atlantic and the Mediterranean. The winter rains they receive, moderate as they are, place them clearly outside the monsoonal realm. Because crops and water supplies depend entirely on monsoonal rains, it became imperative that quantitative, long-range weather forecasts be available. For a forecast to be released at the beginning of June, it is necessary to use, in April, South American pressure data and Indian upper-wind conditions (positive correlation) and, in May, rainfall in Zimbabwe and Java and easterly winds above Calcutta (negative correlation). Climatic classification General considerations The climate of an area, as previously noted, is the synthesis of the weather conditions that have prevailed there over a long period of time (usually 30 years). This synthesis involves both averages of the climatic elements and measurements of variability (such as extreme values and probabilities). Climate is a complex, abstract concept involving data on temperature, humidity, precipitation type and amount, wind speed and direction, atmospheric pressure, sunshine, cloud types and coverage, and such weather phenomena as fog, thunderstorms, and frost and the relationships among them. As such, no two localities on Earth may be said to have exactly the same climate. Nevertheless, it is readily apparent that, over restricted areas of the planet, climates vary within a limited range and that climatic regions are discernible within which some uniformity is apparent in the patterns of climatic elements. Moreover, widely separated areas of the world possess similar climates, which tend to recur in similar geographic relationships to each other. This symmetry and organization of the climatic environment suggests an underlying worldwide regularity and order in the phenomena causing climate (e.g., patterns of radiation, atmospheric pressure, winds, fronts, and air masses), which were discussed in earlier sections. Climate classification is an attempt to formalize this process of recognizing climatic similarity, of organizing, simplifying, and clarifying the vast amount of weather data collected by the meteorological services of the world, and of systematizing the long-term effects of interacting climatic processes to enhance scientific understanding of climates. Users of climate classifications should be aware of the limitations of the procedure, however. First, climate is a multidimensional concept, and it is not an obvious decision as to which of the many observed weather variables should be selected as the basis of the classification. This choice must be made on a number of grounds, both practical and theoretical. For example, using too many different elements opens up the possibilities that the classification will have too many categories to be readily interpreted and that many of the categories will not correspond to real climates. Moreover, measurements of many of the elements of climate are not available for large areas of the world or have been collected for only a short time. The major exceptions are temperature and precipitation data, which are available almost universally and have been recorded for extended periods of time. The choice of variables also is determined by the purpose of the classification (e.g., to account for distribution of natural vegetation, to explain soil formation processes, or to classify climates in terms of human comfort). The variables relevant in the classification will be determined by this purpose, as will the threshold values of the variables chosen to differentiate climatic zones. A second difficulty results from the generally gradual nature of changes in the climatic elements over the Earth's surface. Except in unusual situations due to mountain ranges or coastlines, temperature, precipitation, and other climatic variables tend to change only slowly over distance. As a result, climate types tend to change imperceptibly as one moves from one locale to an adjacent one. Choosing a set of criteria to distinguish one climatic type from another is thus equivalent to drawing a line on a map to distinguish the climatic region possessing one type from that having the other. While this is in no way different from many other classification decisions that one makes routinely in daily life, it must always be remembered that boundaries between adjacent climatic regions are placed somewhat arbitrarily through regions of continuous, gradual change and that the areas defined within these boundaries are far from homogeneous in terms of their climatic characteristics. Most classification schemes are intended for global- or continental-scale application and define regions that are major subdivisions of continents hundreds to thousands of kilometres across. These may be termed macroclimates. Not only will there be slow changes (from wet to dry, hot to cold, etc.) across such a region as a result of the geographic gradients of climatic elements over the continent of which the region is a part, but there will exist mesoclimates within these regions associated with climatic processes occurring at a scale of tens to hundreds of kilometres that are created by elevation differences, slope aspect, bodies of water, differences in vegetation cover, urban areas, and the like. Mesoclimates, in turn, may be resolved into numerous microclimates, which occur at scales of less than 0.1 kilometre, as in the climatic differences between forests, crops, and bare soil, at various depths in a plant canopy, at different depths in the soil, on different sides of a building, and so on. These limitations notwithstanding, climate classification plays a key role as a means of generalizing the geographic distribution and interactions among climatic elements, of identifying mixes of climatic influences important to various climatically dependent phenomena, of stimulating the search to identify the controlling processes of climate, and, as an educational tool, to show some of the ways in which distant areas of the Earth are both different from and similar to one's own home region. Approaches to climatic classification The earliest known climatic classifications were those of classical Greek times. Such sche