Lectures - Monday and Wednesday, 11:00 AM - 12:15 PM
Lab - Tuesday, 4:10 PM -7 PM
In this lecture we put together the knowledge we acquired in lectures 1-4 to understand the general circulation of the atmosphere and climate at and above the Earth's surface. We will learn about the global-scale motion system on Earth and how it divides the planet into different zones, distinguished from one another by their typical weather and climate. This division of the global climate system has had a significant impact on the development of civilization on Earth, because, as we shall later see, present day climate has been with us, more or less in the same way it is now, for the last 10,000 years or so. When time scales longer than that are considered (that is Earth's history on geological time scales) we find that the atmospheric general circulation and climate have undergone very large fluctuations. These were caused sometimes in response to changes in insolation, and other times because of geological changes on Earth itself (continents drifting, mountains rising, and land isthmuses opening to make way for the ocean).
Understanding the way the climate system works is a real challenge today, because anthropogenic activity on Earth is introducing significant changes in atmospheric composition and surface properties (such as the increase in CO2 and aerosol concentrations and deforestation), and we need to assess their future impact. This of course can not be fully addressed before we learn about the oceans and biomass, two other large participants in the climate system. We will return to the issue of climate change later in this class.
This section deals with the climate of the Earth's atmosphere as opposed to weather (which was discussed in last week). The difference between these two concepts in a "mathematical" sense is that weather is a function of space and time, where the time variations are taken to be on the order of a day. Climate, in the pure sense, is a function of space only. It describes the way the average state of the atmosphere varies in space. However, because seasonality is such an important part of the climate system (because of the large variation in insolation throughout the year) we always include one aspect of time behavior in the description of climate - the seasonal cycle. Thus we distinguish between the average behavior of the climate by calendar month. Usually the variability of the seasonal cycle is described using 12 monthly time points - each a long-term average of weather in one calendar month (for the sake of brevity and to capture only the major changes throughout the seasonal cycle, we will show below only charts for January and July). Note however that the division of the year into 12 months is arbitrary, and the changes in outside forcing mechanisms (that is the sun) is continuous. Thus the monthly representation may not provide an appropriate way to describe the temporal variability of climate for all applications.
It is also important is to note that climate can not be represented only by its averaged behavior, and the characteristics of climate fluctuations around this average are important. In other words, the standard deviation of climatic variables are as important a part of the description of climate as is the mean. We shall emphasize this point later on.
As we have discussed earlier, atmospheric motion is driven by the uneven horizontal distribution of net incoming radiation. On a global scale, the most outstanding part of this uneven distribution is its latitudinal dependence (Fig. 1). The atmosphere (and oceans) respond to this imbalance by attempting to move heat from the tropics and subtropics, where insolation surpasses the infrared terrestrial radiation going out to space, to the middle and high latitudes, where there is a net radiative loss of heat. Convection, the vertical process of heat transport, and advection, the horizontal process of heat and moisture transport, work together to accomplish this goal.
It seems logical then to expect that the mean atmospheric motion (so called general circulation) will go from equator to pole in both hemispheres. Could this be the case? Lets begin by observing the horizontal distribution of winds near the surface. This is shown in Fig. 2 for January and July. The most outstanding feature about this distribution is its zonally banded structure - that is, it has the pattern of east-west elongated features aligned along latitude circles. As we shall discuss hereafter, these belts define different climate and weather regimes.
Marching from equator to pole we find the following characteristic wind belts:
The surface wind distribution is not perfectly zonal. Clear difference between continents and ocean are evident in both panels of Fig. 2. Winds are much weaker over land (friction over land is larger than over the oceans, and high variability is another reasons for a low average). Along the coasts, winds deviate quite a bit from zonal symmetry. More on that in the next section.
As we saw in last Monday's lecture, the winds are related to the pressure distribution via the geostrophic balance and, at the surface, via the additional influence of friction. The global sea level pressure distribution for January and July (Fig. 3) is indeed consistent with the wind fields we saw above (Fig. 2). The zonal wind belts are matched by the more or less zonal distribution of sea level pressure. The breakdown in zonal symmetry is however much more evident here. What is causing the asymmetry?
The surface of Earth is far from homogenous in its composition. Land masses interrupt large ocean basins, and the land itself is intersected by massive mountain ranges. The cover (vegetation, snow and ice, soil moisture) of the land surface varies from region to region, partly because of climate itself. All these zonal (and meridional) asymmetries contribute to the asymmetry of the circulation for both thermal and mechanical reasons. Consider for example the difference in thermal properties between land and ocean surfaces. On the time scales of a day, these gave rise to the sea breeze circulation (last Monday's lecture). On seasonal time scales they give rise to many of the zonal asymmetries we see in Fig. 3.
The difference in the heat absorbing properties of continent and ocean - continents warming and cooling faster - is apparent throughout the seasonal cycle. In winter the continents are colder than the surrounding oceans at the same latitude, and in summer they are warmer (see Fig. 4). These temperature differences express themselves as pressure differences (Fig. 3). The continent-ocean pressure differences (Fig. 2) drive the development of regional wind circulations, which are much different than the winds dominating that latitude zone. Such regional winds are particularly noticeable in the tropics. As with the daily breeze cycle, air will flow from where the surface is cold to where the surface is warm, thereby creating seasonal circulation systems known as monsoons (derived from an Arabic word denoting season).
The most famous monsoon circulation is that over India. In summer, southern Asia and the Indian subcontinent are warmed up by the sun, and low pressure develops over the continent (Fig. 3, July). A large system of ocean-to-land winds forms around India (Fig. 2, July) just as with the sea breeze. However, unlike the sea breeze circulation, monsoon winds are subjected to the Coriolis force which divert the winds from flowing directly from ocean to land. It is surface friction that is causing deviation from geostrophic balance, and convergence into the warm land mass. Over the oceans, divergence occurs. Aloft, the situation is reversed, and the air diverges from the land to the ocean. The cycle of winds is closed by the rising and sinking of air over land and ocean, respectively. Convergence into land and the uplift it generates brings large amounts of moisture and some of the largest rainfalls known on Earth (the uplift is also aided by the air flowing against the Himalayas and the Tibetan Plateau).
In winter the land is colder than the surrounding ocean, and surface winds diverge from the land to the ocean. The dry continental winds and the subsidence over the subcontinent, suppress clouds and rainfall over land. However, further east over Southeast Asia and northern Australia (centered on the equator and ~130°E, see Fig. 2, January), the circulation induced by the contrast in winter temperature between Asia and the Indian Ocean brings convergence, massive convection and enhanced rainfall.
Monsoons occur in other regions of the world: The Central and North American monsoon brings summer rainfall to Mexico and the US Southwest. In the winter, the American monsoon migrates to South America, where it is summer. Monsoon circulation also affects the African continent.
Differences between land and ocean are also seen in the westerly wind belt (midlatitudes). Here the deepest winter "lows" are located over the warmer oceans, analogous with the night half of the sea breeze cycle. In summer the situation reverses, and the pressure over land is lower than over the ocean.
The thermal effects of continents and oceans affects climate extremes such as the difference in temperature between winter and summer, or the average difference between daily minimum and maximum temperatures in different seasons. In Fig 5, for example, we plot the climatological difference between January and July temperature (a rough measure of the amplitude of the seasonal cycle). In part, this figure depicts the fact that amplitude of the annual cycle increases poleward. However it also clearly depicts a continent-ocean contrast: the largest differences at every latitude are over land. Notice also that over the massive continental regions of the Northern Hemisphere (Asia in particular) the largest differences occur over the eastern sides of the continent. The further poleward we go, the larger the difference. This reflects the extent of the moderating effect exerted by the oceans on the western sides of the continents within the westerly wind belt.
A note on upper level data representation: In meteorology, upper level variables are represented on pressure rather than on height surfaces. The fundamental force on an equal-pressure surface is the height gradient multiplied by the constant of gravity (a quantity referred to as the geopotential). This transformation is a direct outcome of the hydrostatic relationship. Air is initially moving from a high to a low point on a pressure surface (just as it moves from high to low pressure on a constant height surface). In doing so, the air is converting potential energy to kinetic energy. The Coriolis force enters and makes the air flow along constant height contours. The heights of any given pressure level are calculated in different geographical locations using the hydrostatic relationship with values of temperature and pressure measured by weather balloons. In the calculation, an assumption is made that the acceleration of gravity, g, is a constant everywhere on Earth. Meteorologists refer to this measure of height as geopotential height, because by multiplying it by a constant g it gives the potential energy of air on that pressure level.
The zonally symmetric structure of the atmospheric circulation is much more clearly depicted in the upper troposphere. The winds at the 300 mb level (about 9000 meters above ground) are shown in Fig. 6 . Here three belts can be discerned in each hemisphere: a region of weak flow in the tropics, a wide belt of westerlies in the midlatitudes, and a second region of weak flow in the polar latitudes. The westerlies are much stronger at this level than at the surface - evidence for strong geopotential height gradients and less friction. Around the latitude of 30°, north and south of the equator, is a narrow region of very strong westerlies known as the subtropical jet stream. The jet stream is strongest in the winter hemisphere. It is not uniform in longitude and presents several relatively localized maxima (for example, in the western North Pacific and North Atlantic Basins). These jet maxima are related to the location of the regions of strong tropical convection (see below - the Hadley Circulation).
The geopotential height climatology of the 300 Mb surface is shown in Fig. 7. Note that there is a much larger degree of zonal symmetry here than in the sea level pressure map (Fig. 3). Further away from the Earth's surface, the 300 Mb level is not so strongly affected by the differences in continent-ocean properties. The deviation from zonal symmetry here is mainly cause by the large topographical obstacles on the Earth's surface (the Tibetan Plateau and the Himalayas, the Rocky Mountains, the Andes, etc.)
Because the atmospheric general circulation is, to first order, zonally symmetric, it is helpful to examine the variation of zonally averaged properties as they vary with latitude and height (Fig. 8). In observing this figure, we find that the temperature lines slope most strongly towards the surface in the midlatitudes - an indication for a large poleward temperature gradient close to the surface. We also find that high in the midlatitude troposphere, the westerly winds reach their maxima - this is the zonally averaged depiction of the subtropical jet stream. Are these features in the temperature and wind fields, respectively connected?
Because the atmosphere is in hydrostatic balance, pressure decreases with height more slowly in warm air than in cold air. If we started with the same pressure at sea level in two columns of air separated from one another by a difference in temperature, then as we ascend, a pressure difference will form between the two columns. The larger the temperature difference the larger the pressure difference. The higher we ascend, the larger the pressure difference. In the midlatitudes, the temperature difference is such that the air becomes colder as we get closer to the poles (that is, a poleward temperature gradient). A poleward pressure gradient is therefore established at high levels, and according to the geostrophic relationship, a westerly wind is generated. This relationship between winds and temperature gradients is known as the thermal wind relationship (or balance).
The large midlatitude poleward temperature gradient near the surface, which we see in Fig. 8 (its exact location varies with season), is therefore consistent with the jet stream that moves just above. The thermal wind balance also holds when the full three dimensional state of the atmosphere is considered, as one can verify from Fig. 4 and Fig. 6. Note that this relationship does not explain the existence of the jet stream. The wind and the temperature gradient are tied to one another and must, to a large degree of accuracy, obey this relationship.
What is then the reason for the westerlies and the jet stream? Or the reason for the wind belts in general? The second question is probably as old as the days when ships began circumnavigating the Earth. The first to suggest a reasonable answer to it was George Hadley, an Englishman who lived in the first part of the eighteenth century. Hadley was aware of the warmer tropics and cold high latitudes. He envisioned that this would create a global circulation akin to the sea breeze, with the warm tropical air rising and moving northward aloft, and the cold high latitude air flowing southward near the Earth's surface to displace the former.
Much of Hadley's original understanding is basically correct. The excess radiative heating in the tropics is balanced partially by evaporation but remains warmer than the air to the north and south. The warm moist air is conditionally unstable - as soon as a small uplift occurs, the air saturates and becomes unstable (see moist air instability in Week 3 lecture). Convection in the tropics is deep and massive. Large amounts of latent heat are released in the towering cumulus clouds of the tropical atmosphere, providing further driving for the upward motion (see satellite mosaic for a downward view of Earth and its cloud cover from space). At the top of the tropical troposphere (12-15 km), the atmosphere is stable and the upward motion is brought to a halt. This forces the air to diverge poleward on both sides of the equator. The further the poleward moving air gets from the equator, the larger the Coriolis force, and the air begins to deflect to the right (in the Northern Hemisphere) and left (in the Southern Hemisphere). The result (in both hemispheres) is an eastward flow of air (westerly winds). The further poleward we go, the stronger the westerlies become, till we reach the jet stream latitude. This situation is depicted in (Fig. 9) for the Northern Hemisphere.
In Hadley's original explanation, the winds continued all the way to the pole, and then sank to return all the way to the equator (like a global sea breeze). Hadley was not aware of the law of conservation of angular momentum and thought the poleward moving air would conserve its east-west velocity. But because the poleward moving air is getting closer to the axis of the rotating Earth, it spins up continuously until it reaches maximum in the midlatitudes. This is where the fast moving air breaks down into large eddies (or turbulence, see below). The breakdown of the flow is also accompanied by a sinking motion towards the surface just south of the jet stream axis. This subsidence is the reason for the horse latitudes. When it reaches the surface, the air spreads equatorward and poleward. The equatorward spreading air acquires a westward movement due to Coriolis and the trade winds are formed. It continues to move westward and equatorward, gathering moisture, and converging at the Intertropical Convergence Zone.
The zonally averaged poleward motion in the tropics is not in geostrophic balance with a zonally averaged pressure force (there is no mean zonal pressure gradient, can you explain why?). It is a wind driven by upper tropospheric divergence originating in tropical uplift and convection. In the same way, part of the surface flow in the trade belt is not in geostrophic balance, and creates a return flow towards the equator (here friction "helps" the convergence) feeding the huge circulation cell named after Hadley. It defines the most vigorous part of the atmospheric meridional mean circulation (see Fig. 10). The Hadley Cell accomplishes a large part of the task of carrying sensible heat (warm air) toward the higher latitudes. While the Hadley cell also carries latent heat (water vapor) from higher latitudes toward the equator, the sensible heat transport is larger.
The latitudes of the strongest temperature gradient, north and south of the equator, are the latitudes of the strongest westerlies aloft - they are in thermal wind balance as we saw above. When the temperature gradient and upper level winds are too large, pure eastward motion can not be sustained and the flow breaks down into large-scale eddies (turbulence). That this is so, is evident form looking at a daily upper level chart (Fig. 11). Here there is no longer the "almost perfect" zonal symmetry of the flow seen in the climatology (Fig. 6). The climatology is thus misleading because it is a smoothed depiction of the atmosphere, created by averaging together many charts with turbulent flow patterns.
The breakdown into eddies is a phenomenon referred to in the scientific jargon as baroclinic instability. Baroclinic instability favors large horizontal temperature gradients and strong winds aloft. The stronger the jet, the more likely it is to be unstable.
Baroclinic eddies (the eddies resulting from baroclinic instability) are the cause of what we in the midlatitudes call weather. They are trains of alternating low and high pressure systems moving eastward in the westerly flow. They are the source of the frontal systems, with their clouds, rainfall and storminess, of the midlatitudes (see satellite mosaic for a downward view of Earth and its cloud cover from space). They break the Hadley Circulation down, and stop the equatorial air from moving clear to the poles. However, in their circular motion, baroclinic eddies push warm air poleward and cold air southward, cooling the subtropics and warming the polar latitudes (see Fig 12, note how the air is always colder on the westward - or upstream - side of the lows, and warmer on the eastward - or downstream - side). Thus the process of poleward heat transport is kept going all the way to the poles.
The meridional mean circulation associated with the midlatitudes is the Ferrel Cell (Fig. 10). It is defined by the sinking motion in the subtropics along the southern flank of the jet, and rising motion along the average position of the polar front. It is made up of the average of the turbulent motions associated with baroclinic eddies and is thus weak and not that important for the transport of heat or moisture in the climate system.
The climatological distribution of annual mean land rainfall (Fig. 13) provides a backdrop for a summary of the Earth's climate zones.
Precipitation (Fig. 13) vs. Evaporation (Fig 14): The balance between precipitation (P) and evaporation (E) is important both for the atmosphere and the oceans. Globally and annually, a balance must exists between P and E because evaporation is the only source for atmospheric moisture. However, in the Tropics: P > E while in the subtropics: E > P. Thus, a latitudinal atmospheric water imbalance exists at the Earth's surface. Water vapor must be transported into the ITCZ from the subtropics to maintain global balance. This is achieved by the return branch of Hadley Circulation and the trade winds.
Text by Yochanan Kushnir, 2000.