the capricious role of water
By Dr J Floor Anthoni (2010)
(This chapter is best viewed in a page-sized window and best navigated by opening links in a new tab of your browser)
Note! for best printed results, read tips
for printing. For corrections and suggestions, e-mail
-- Seafriends home -- climate index -- global issues -- Rev:20100515,20101105,20110124
Water is the strangest substance on Earth, by far. It should be a gas at Earth's temperatures, yet it occurs as a liquid, a solid, a gas and as cloud. Take for instance a related substance, hydrogen sulfide (H-S-H), which is as close as one can get to water (H-O-H). H2S has molecular weight 1+32+1=34, whereas water H2O has 1+16+1=18, and is thus substantially lighter than hydrogensulfide. Yet water occurs as a liquid and H2S only as a gas.
It is evident that water molecules like to cluster together to form bigger clumps, and this property is attributed to its unique shape and electrical polarity, shown in the drawing. Its peripheral H+ atoms attract the central O- from another molecule, and vice versa so that four molecules form a pyramid. This pyramid is not entirely symmetrical, such that it attracts to other pyramids and so on.
Other properties of water are no less spectacular:
is Earth's water found?
The table on right shows where all the water is found, its estimated volume and how high it would reach if evenly spread over the globe. Also estimated residence times are shown.
Most of the water is found in the oceans of course, followed by ice caps and ground water. The atmosphere contains a negligible amount of water, yet this water plays a most important role in regulating Earth's climate. The total amount of water is 1370 + 4 + 30 = 1400 million cubic kilometres (bold figures).
Cycle times or residence times are rather deceptive because although it would take millennia to recycle all the water in the oceans, its surface layers recycle fast - where else is atmospheric moisture coming from? Deceptive also is soil moisture, some of which (permafrost) is not recycled. However, being 5 times larger than that in air, it plays an enormously moderating role on climate (summer begins only when soils have dried up).
Note that the words warmth and heat are often used interchangeably, but there is a difference. Warmth is the energy content of a substance, whereas heat is the energy in transfer from a warm object to a cold object. Cold is identical but opposite in the way it is used. Hence our introduction of the word coolth which is equivalent to lack of energy stored, similar but opposite to warmth.
The most important physical properties of water, relevant to climate are how it changes phase with temperature and pressure, its latent and specific heat, and the importance of humidity. Latent and specific heat will be dealt with again in the next chapter.
|Water, ice and vapour
Water is the most important substance on Earth, but is rather difficult to understand. Let's begin with a diagram showing how it changes to ice and vapour, depending on temperature (horizontal) and pressure (vertical). For ease of understanding, the conventional graph is mirrored upside down such that higher pressures are found lower in the graph (as on Earth). We use the unit of pressure of bar (for barometric pressure or atmosphere), which is 1 on Earth's surface. There are three red lines that separate water from ice and vapour, coming together in the triplepoint where water exists in balance with ice and water vapour. Note that these three lines represent the boundary conditions where an equilibrium (balance) exists, such as can be done in an experiment. The thick black horizontal bar represents the surface of Earth, where barometric pressure is one atmosphere. It extends from -50ºC to +50ºC, roughly the maximum range of surface temperatures. The diagram also shows four coloured shapes, which make the picture more interesting.
The blue shape for water runs between 0 and +50ºC, narrowing towards higher pressures because it becomes colder in the deep sea. The pressure under water changes by 1 bar for every 10 metres, so that 1000 bar (1E3) is found in deep ocean trenches. Should a volcanic vent produce hot water of +200ºC (far right bottom), it can exist as steam because its location in the phase diagram is above the red curve separating water and vapour.
Although the whole ocean covers over 70% of the planet's surface, mostly
at temperatures above 0ºC, the phase diagram shows that this is in
the vapour phase, indicated by the lighter triangle named evap for
What it means is that surface water is not in balance anywhere on Earth,
and that it will always be 'eager' to evaporate, the rate of which depends
on temperature (and wind + waves). Water can exist in rivers and lakes
above sea level, but this means that it will be even more out of balance.
Fortunately, temperature decreases with altitude by 0.6ºC per 100m.
Life owes its existence to evapotranspiration because it enables plants to draw water from the soil, all the way to their leaves. In addition, all oceans are continually evaporating moisture which then rains down on the land, thus supplying terrestrial life with a steady supply of water. More about this below.
On its left in the graph, water is bounded by ice which
can exist in thickness of several kilometres, and thus high pressures.
The line down from the triplepoint, tilts slightly such that high pressures
do melt ice. A female ice skater places 50kg on a square centimetre skate,
which amounts to 50 bar, enough to melt ice of -2ºC.
Ice can be found in high mountains, where it can sublimate (evaporate direct from ice to vapour) to the air, but in order to melt, its temperature must increase to above 0ºC. Ice is about 9% less dense than water, so the vertical pressure scale is 10 bar for 11m of ice. Note that in salt water the red melt line runs somewhat to the left because salt lowers the freezing point.
From the diagram it can be seen that ice, unlike water, is not out of balance at the surface pressure of 1 bar. It is thus a very stable form of water. However, with altitude, ice sublimates as shown by the missing triangle, and cold air can still hold a sizable amount of moisture (see diagram below). However, a substantial amount of latent energy is required to do so (623 cal/g) compared to melting (80 cal/g).
pressure and temperature
Water vapour, better known as steam, is a gas that serves civilisation in its use for steam engines and steam turbines (and much more). The graph shows how steam pressure increases rapidly with temperature, to such extent that above 120 degrees, it can do work, like converting heat to electricity in power plants. But at earthly temperatures, it cannot do so. Even so, water remains the most important substance for transferring heat and for equalising the temperature of the planet (blue shape). In the next graph we'll examine how it works.
Water vapour is found only in air, which rapidly thins with altitude,
shown as the purple shape below and the light blue shape above, following
the dashed temperature line in the phase diagram. Notice that at some stage
the temperature of the air crosses into the ice phase, above which clouds
may turn into ice (snow). The troposphere ends at 0.2 bar (11km) where
the tropopause is, then warms to just below freezing point in the stratosphere
which ends at 0.01 bar in the stratopause (47km).
|Water content of air
The amount of water found as water vapour is very small, no more than an ocean of 3 cm deep (0-7cm). It is found only in air. The graph shows the amount of water in air, against temperature, in red as a percentage and in green as grams per cubic metre. Cold air holds substantially less water than warm air, and not shown, this amount decreases rapidly with altitude. Shown are also the ranges for three habitats: the ice caps extending below 0ºC; the oceans from 0 to 27 ºC (overlapping the world's green plant belt); and the deserts which could produce much water vapour if only they had water.
Also shown is water vapour's partial pressure 0-40mbar which corresponds to 0-4% (purple curve, 4% of 1 bar = 40 mbar). This curve represents the water's 'eagerness' to evaporate.
Surrounding the purple curve is drawn a light purple shape, meaning that the temperature of the ocean (purple) usually differs from that of the air. The right-hand side of this purple shape corresponds to a cold ocean under warm air, and obviously, the water is unable to fill the air's capacity for water, resulting in low relative humidity, dry air. The left-hand side of the shape corresponds to cold air over warm water, which causes high relative humidity, damp air.
Our extensive section about soils and their ecology, analyses the importance and consequences of evapotranspiration. The world map below shows where moisture is found in the air, as expected, mainly around the equator.
|Global relative humidity
Humidity of the air has been measured by weather balloons, which give accurate results. Their data shows that relative humidity has been decreasing in the upper troposphere, particularly at 0.3 bar at the uppermost boundary. But even at 0.7 bar (?km) it has decreased by 4% over 50 years. Why this happened or what it means, is not known, but since water vapour is thought to be the most potent greenhouse gas, a decrease may have contributed to cooling of the atmosphere.
density at 4 ºC
Specific Weight at 4 ºC
Specific weight/density of ice
Latent heat of melting
Latent heat of evaporation
Specific heat capacity water
Specific heat capacity ice
Specific heat capacity water vapour
Thermal expansion from 4 ºC to 100 ºC
Bulk modulus elasticity
kg/m3, 1 kg/litre
0.9167 kg/m3, 8.99 kN/m3
3.34E5 J/kg, 79.8 cal/g
2.27E6 J/kg, 543 cal/g
380 ºC - 386 ºC
221.2 bar, 22.1 MPa (MN/m2)
4.187 kJ/kg/ºK, 1.0 cal/g/ºC
2.108 kJ/kg/ºK, 0.504 cal/g/ºC
1.996 kJ/kg/ºK, 0.477 cal/g/ºC
2.15E9 (Pa, N/m2)
|Very large volume times
high specific heat capacity
Atmosphere only equates to 10m depth
Speed times mass
Ocean currents are slow and superficial. Winds are fast.
Reader please note that this aspect of climate has been
overlooked or under-reported.
|Latent heat, specific heat
and heat capacity
The name latent heat (concealed heat) is used for the heat required to change from one phase to the other (e.g. melting, evaporating) without changing temperature, whereas heat capacity (=specific heat) is the amount of heat required to warm a standard unit by one degree, without changing phase. In the diagram we are using the conventional units for warmth (energy), the gram-calorie or cal (not to be confused with the food-calory which is a kilo-cal =1000cal). To warm a block of ice of 1000g from -10ºC to 0 ºC takes 5000 cal because heating ice requires 0.5 cal/g/ºC as shown in the diagram under <->. The latent heat of 80 cal/g must be provided (=80,000 cal) to melt the block, still at 0 degrees. To heat it further as water, to 100 degrees, requires 1000 x 1 x 100 = 100,000 cal because water's heat capacity is 1 cal/g/ºC. To evaporate this water requires a whopping latent heat of 540 cal/g, amounting to 540,000 cal. From there on, as pure steam without air, requires 0.5 x 1000 cal per degree.
|steps in heating 1kg ice from -10º
heat ice from -10 to 0 degrees
heat water from 0 to 100 degrees
heat steam from 100 to 150 degrees
|heat required (cal)
|Note that this example bears relevance to deep-frying items from the deep-freeze (-18ºC), and how to do it with the least loss of heat and moisture. The freezer temperature plays only a little role, and the energy to melt, then to heat to 100 degrees in fat of 190 degrees is of main concern. Worst is the heat lost unnecessarily in evaporating moisture, which could amount to 250,000 cal for 50% loss in moisture. In small fryers it is often necessary to bring the fat up to heat by removing the food after 1 minute of melting, after which the oil is then hot enough to crisp the food without appreciable loss in moisture. In this manner both half the heat and half the moisture can be saved, resulting in crisp yet moist food and also in higher through-put (and health).|
When water vapour exists in air at 2%, it adds 0.02 x 0.5 = 1% to the air's heat capacity, which is negligible. However, by changing phase from vapour to cloud, it releases a latent heat of 0.02 x 540 = 10.8 (cal), equivalent to 10.8 / 0.5 = 22 degrees of warming. Water vapour is thus a considerable player in the transfer of heat through the atmosphere.
Juan G. Roederer: "In a highly nonlinear system with large reservoirs of latent energy such as the atmosphere-ocean-biosphere, global redistributions of energy can be triggered by very small inputs, a process that depends far more on their spatial and temporal pattern than on their magnitude"
Clouds form by water vapour changing into water or ice, encouraged by cloud condensation nuclei which are 10-20% more common over the seas, than over land. Droplets over sea are also twice as large as those over land. Typically, one cubic cm (teaspoon) contains 100 droplets. Cloud droplets grow by more vapour condensing onto them (up to 30µm), but mainly by collisions and coalescense (sticking together) when moving around. As they grow, they also grow faster. As water vapour condenses, heat is released, which causes uplift of the cloud. Once droplet speed exceeds uplift, it begins to rain. The table below shows physical properties of cloud droplets.
|Note that cloud condensation nuclei CCN are typically smaller than infrared wavelength but larger than light wave lengths, sometimes a single molecule can do it. Thus raindrops begin at this size, rapidly growing to a typical size of 10 micron. As droplets grow, also their terminal velocity (hitting Earth) (speed) increases progressively. Note that raindrop volume (weight) is proportional to radius to the third power, thus increasing very progressively. During heavy rainfall, rain drops are heaviest and they are also fastest, resulting in unexpected soil damage. A 5mm rain drop is 5x5x5 = 125 times heavier than a typical 1mm rain drop, and its speed is twice faster, resulting in 600 times more kinetic energy, reason why most soil erosion happens during heavy rain storms. See soil/erosion/rain|
The way snow forms is more complicated as discovered by Bergeron-Findeison.
Water does not eadily convert to ice, such that droplets occur in an under-cooled
state colder than ice. Snow crystals can then form instantaneously from
the undercooled water droplets. The formation of ice releases heat which
moderates the process. Ice crystals also grow from more water droplets
and water vapour, but also from collision and coalescence. Once their speed
exceeds the cloud's updraft, they fall out as snow, or rain in case lower
altitudes are warm.
|droplet type||found in||radius
|cloud condensation nuclei
typical cloud drop
large cloud drop
smallest rain drop
typical rain drop
heavy storm rain drop
drizzle from thin stratiform clouds
very heavy rain from cumulus clouds
Interestingly, clouds, water vapour and rain are transparent to Earth's microwave radiation in the 5.0 mm band. Using this property, satellites can measure land and sea surface temperatures, even though this signal is very weak.
From many observations, it is known that the world has experienced rapid climate fluctuations, known as ice ages. The temperature record (black) below shows how an ice age begins suddenly, only to proceed gradually with many oscillations, and that the recovery towards warm interglacials, happens more rapidly. The gases that are associated with life (CO2 and methane) follow in step. Notice the enormous variation of 8ºC as measured for Antarctica. A new ice age can begin with a sudden drop in temperature of 2-3ºC in less than a century, with disastrous consequences for life (and humans) on Earth.
|This graph, obtained from the Vostok (Antarctica) ice core, shows the capricious, yet cyclical nature of the past 4 ice ages as temperature (red) suddenly dips, then declines in large oscillations to a glacial minimum of some -8 degrees (at Vostok Antarctica). With it also the gases of life, carbon dioxide and methane, follow in step. On a more detailed time scale, the gases follow the temperature rather than the other way, which is an important observation refuting man-made global warming.|
We'll explore the glaciation mechanism by assuming that it depends on the run-away (positive feedback) effects of:
cooling => more snow => more light reflected => more cooling
and the end of an ice age:
There are a number of important factors:
this graph the average temperatures for both hemispheres are shown and
their summer and winter variations. The top shows how little the oceans
change from north to south (25 degrees) and from summer to winter (6 degrees)
and even less so near the poles and the equator. By contrast, the land
temperatures change vastly more, from north to south (50 degrees) and from
winter to summer. Notice how the polar north varies the most, due to the
large amount of land, compared to that of the oceans. In other words, radiation
(in and out) is an important part of arctic climate, reason why an ice
age has such large effect.
Notice that towards the poles, the difference between air and sea temperature becomes an important driver of precipitation and climate.
|This graph shows the incoming radiation by month for three latitudes. Note that during the summer months, the mid-latitudes and polar regions receive more sunshine than the tropics because of their long days and short nights. Notice the very short summer months for the poles. Note also in the right-hand graph how radiation changes most rapidly in the polar regions. Thus the effect of the summer months is rapidly negated by shorter summers and less warming, which is what happens in ice ages.|
Because the air can contain only a few centimetres of water, the bulk of the water is stored in oceans and ice caps. Thus mass transfers of water can occur only between these two. Ironically, ice is mainly stored on land whereas water mainly in the oceans. So an ice age can happen only by the transport of moisture through the air, from the oceans to the ice caps. It is a slow process, even though it appears to happen suddenly. The scenario behind ice ages could be as follows:
|From the Vostok (Antarctica) ice core, which goes back for 4 ice ages, it can be seen that the temperature goes through very similar swings. In this graph, the five interglacials have been superimposed, aligned by their maximum temperatures. The magenta-coloured squiggle is our present interglacial warm period, which has been remarkably level over time. It has lasted now for over 6 millennia and the next ice age can begin any time soon (give and take a millennium). The Eemian is the previous warm period, some 110,000 years ago.|
There exists a possibility that an ice age can begin only once the arctic ocean opens up and its ice sheets have melted away. It then becomes a source of moisture by evaporation, where before, little moisture existed. As the water evaporates, it is replaced by warm water from the warm Gulf Stream, thereby accelerating evaporation. As soon as the air crosses the cool continents, all its moisture precipitates as snow. The arctic then behaves as a warm Jacuzzi (hotpool), transferring water to surrounding continents at an abnormal rate. An ice age ends once the sea level drops below -100m, and the warm Gulf Stream is choked off from reaching the arctic ocean. The Arctic ocean has very large continental shelves (light blue on map), which would gradually reduce the jacuzzi effect, eventually freezing over entirely.
Oceans have around 1000x the heat capacity of the atmosphere. If the atmosphere transferred so much heat to the oceans that the air temperature went from an average of 15°C to a freezing -15°C, the oceans would heat up by a tiny, almost unnoticeable 0.03°C. The atmosphere cannot heat the oceans, because it does not have enough heat capacity and heat wants to rise rather than sink.
Reader, please note that the above scenarios are forms
of speculation, based on physical principles. To our knowledge, the above
scenario has not been reported before, and it is here that you first read
it. Even if this scenario proves to be wrong, it still remains a good exercise
showing how many factors work together: solar radiation, cosmic radiation,
physical properties of water, physical properties of carbondioxide, curvature
of Earth, geography of land and sea, tipping continents and the presence
Ice shelves are found only where the sea water becomes cold enough to freeze over during winter, such as at both poles. Because the arctic (north pole) is an ocean surrounded by land, all arctic ice consists of ice shelves (except for Greenland, Iceland and north Canada). During summer as the sun light becomes stronger, the ice shelf warms up, particularly where melt water pools form, because water absorbs infrared warmth whereas snow reflects it. The higher temperature also causes the ice and melt water to slowly ablate (evaporate). Because ice is a good insulator and does not conduct heat easily, the melting below happens later and more slowly (top-down). However, because the ice experiences intensive contact with water, which transfers heat very effectively (bottom-up; heat moves upward), the melting below the ice proceeds rapidly, depending on sea temperature and currents. In the process, a layer of fresh water forms under the ice shelf, protecting it from salt water attack. In this condition, the melting process is very sensitive to wind, waves and currents, resulting in quite unpredictable 'erosion'. Think of sea ice as a very thin sheet, sandwiched between two worlds of air/wind/sun and water/waves/currents, both rather unpredictable.
sea ice extent is not a good measure to judge global warming by.
winter the ice shelf thickens from two sides: snow falls on top and seawater
freezes from below. As the seawater freezes, it leaves its salt behind,
creating a salt lens underneath the ice shelf. Because cold salt water
is very dense, it sinks to the bottom where it begins a long journey over
the bottom of the oceans, as part of the ocean thermo-haline conveyor belt.
The diagram shows how cold water sinks both around antarctica and the North Atlantic, because both cold and salinity make it maximally dense. Travelling across the equator inside deep ocean trenches, it resurfaces somewhere in the north Pacific and Indian Oceans, as a warm water girdle closing the loop.
Note that a layer of ice over the ocean changes its properties quite considerably. For instance, the wind can no longer make waves and surface currents, although it can move ice shoals. Also the water can no longer evaporate to bring more snow, and sunlight is reflected back into space. A solid ice shelf is thus a rather stable condition. As it melts and more and more open water appears, the wind, waves and currents can suddenly accelerate the melting while the open water warms by absorbing light and heat. In addition, the life in the sea begins to bloom, making its own contributions.
In the debate about global warming, much ado is made about the extent of sea ice, but it is not a good indicator of warming/cooling because it depends on so many factors that influence one another, particularly on ocean currents. Note that sea ice, which floats on the water, cannot cause the sea level to rise when it melts.
polar temperature important?
Much discussion and research is about whether the poles are warming or not, but is this relevant? Both poles are mainly covered in snow and ice, reflecting most of the solar radiation they receive. Half the year, they receive no sunlight at all. Their temperatures are unexpectedly much lower than elsewhere, with a steep gradient in between and a constant plateau over both poles (see graph above). With solar warmth negligible, their temperatures are determined by the air passing over, and this is subject to large variations.
In the chapter about water and ice we saw that ice has a rather low heat capacity and it is a good insulator, which means that if one warms the surface, the heat very reluctantly goes any deeper. So the temperature of the poles depends only on a very thin layer of material and consequently has a very low heat capacity, unlike that of the sea which has a very large heat capacity.
Furthermore, the air at polar temperatures contains very little moisture, and thus very little capacity to transfer heat by means of evaporation and condensation. It is essentially dry air with very little heat content, unlike the air passing over a body of water.
As a consequence, the poles should be excluded from world average temperature calculations. Unfortunately today they receive far too much importance.
Glaciers are rivers of ice, found on the coldest slopes of mountains. They grow in thickness by snow, which becomes compressed to ice and then to hard ice. At a depth of about 50 metres, the hard ice deforms into metamorphic ice, capable of distorting slowly. In this manner, the glacier indeed 'flows'. For instance, at its bottom and sides, it flows around obstacles, but downward of the obstacle, the negative pressure turns the flowing ice again into hard ice, 'plucking' at obstacles. Glaciers flow faster near the surface than near the bottom. Below the summer frost line, glaciers can melt in summer, and below the winter frost altitude, they melt continuously. Because ice has a high latent heat (coolth) content, melting happens slowly, and because ice is a good thermo-insulator, it stays cold near the bottom and inside.
A glacier's erratic behaviour can be thought of as follows:
thicker ice => faster speed => extends => thinner => slows down => thicker ice
Much ado is being made of glaciers shrinking due to global warming, but glaciers are not really good indicators because their size mainly depends on the amount of snowfall and thus the moisture content of the air. Earlier in this chapter we saw that land use change has a major influence on the water cycle, causing land-locked glaciers to shrink but leaving coastal glaciers unaffected. Ironically, most glaciers are land-locked, seemingly supporting the global warming argument.
When glaciers end in the sea, they may still be 'locked' onto the bedrock
because the ice can float only when 90% is submerged. Such glaciers can
spawn ice bergs by breaking off small pieces, small enough to float.
melting due to global warming?
The melting of glaciers has been cause for concern, fanning the fear of global warming. This graph pictures both humanity's use of fossil fuels and glacier shortening which began almost a century earlier and which is steadily progressing. Because glaciers shortening began before the industrial age, human emissions of carbondioxide and their associated global warming, cannot have caused their shortening. In this graph also the increase in sea level is shown, tracking along the glacier-shortening curve.
http://www.appinsys.com/GlobalWarming/GW_4CE_Glaciers.htmAn excellent summary of the world's glaciers.
Could it be that global cooling causes heat-waves? Surely not! Yet it does! So let's investigate this further, but first a number of principles that we've explained before:
|Deluges of snow and rain.
The world has been hard-hit by abnormal weather in late 2010 and early 2011. Many articles have been written and opinions voiced. Some even blame global warming, but what really happened?
It is indeed an unusual co-incidence of natural climate processes, which boil down to the following scenario:
The top 10 grain producers in the World, in order, are: China India USA Russia France Canada Germany Ukraine Australia and Pakistan. Of these top 10 producers the following countries have experienced catastrophic failures in 2010: China due to unprecedented flooding, India due to flooding, Canada due to drought, Russia due to unprecedented drought, Ukraine due to unprecedented drought, Australia due to drought and locust plague, and Pakistan, whose loss of crops due to never before seen flooding is near 100%. Russia has banned all exports of grain. World's grain supply has fallen to 72 days of consumption, their lowest level in 37 years. The Green Revolution, whose technologies had delivered the last great surge in global food production in the 1970s and 1980s seemed to be fizzling out, a view supported by the disturbing slide in crop yield advances. Yields of the major crops of wheat, maize, and rice had once increased by as much as 5 and even 10 percent a year — now they were increasing by 1 percent or nothing at all.
To sum it all up, the challenge facing the world’s 1.8 billion women and men who grow our food is to double their output of food — using far less water, less land, less energy, and less fertilizer. They must accomplish this on low and uncertain returns, with less new technology available, amid more red tape, economic disincentives, and corrupted markets, and in the teeth of spreading drought. Achieving this will require something not far short of a miracle.
"The first foreshocks were discernible soon after the turn of the millennium. In the years from 2001 to 2008 [when global cooling set in] the world steadily consumed more grain than it produced, triggering rising prices, growing shortages, and even rationing and famine in poorer countries. The global stockpile of grain shrank from more than a hundred days’ supply of food to less than fifty days". .
Commented Joachim von Braun , the head of the International Food Policy Research Institute: “High energy prices have made agricultural production more expensive by raising the cost of cultivation, inputs — especially fertilizers and irrigation — and transportation of inputs and outputs. In poor countries, this hinders production response to high output prices."
"Food was becoming the new gold. Investors fleeing Wall Street’s mortgage-related strife plowed hundreds of millions of dollars into grain futures, driving prices up even more. By Christmas (2007), a global panic was building,” reported the Washington Post.
 Julian Cribb (2010):
Coming Famine. University of California Press, 2010.