ÐÓ°ÉÔ­´´

The Popular view of the greenhouse effect is that it will bring about
a world some 4 °C warmer than now, where seas rise as the ice sheets
melt and ecological chaos leads to starvation and civil strife. These are
seen as the inevitable consequences of releasing into the atmosphere carbon
dioxide from the burning of fossil fuels and other gases from industry.
But is this really what the future has in store? Models of the world’s climate
and historical temperature records suggest that a greenhouse world may not
be as different from our own as feared. How much the world warms overall
may not be of prime importance if the warming is not uniform from region
to region, season by season and between day and night.

Most scientists involved in research on climate change believe that
the Earth will be warmed by an enhanced greenhouse effect resulting from
increased emissions of gases that absorb infrared radiation. The atmospheric
concentration of several such gases has increased as a result of the growth
of industry and population; almost all the current increase in infrared
absorption by atmospheric gases is associated with carbon dioxide, methane,
nitrous oxide and CFCs, in descending order of importance. As a result of
the accumulation of these gases in the lower atmosphere, a very small, but
increasing, fraction of the energy that would otherwise escape to space
is redirected towards the Earth’s surface. But scientists disagree on how
much, when and how the world will warm.

The popular picture of searing heat, floods and imminent starvation
became prominent, especially in the US, following the publication in the
1980s of the results of climate simulations which predicted that a doubling
of the amount of carbon dioxide in the air would result in an average warming
of 4.2 °C across the globe, with winter warming of as much as 18 °C
around the North Pole. This depressing vision can be accommodated by the
more extreme scenarios in last year’s report from the Intergovernmental
Panel on Climate Change, but it is not characteristic of the median range
of climate impacts suggested in that report.

Levels of carbon dioxide in the Earth’s atmosphere are certainly increasing.
Continuous measurements of carbon dioxide concentration have been made since
the late 1950s at the Mauna Loa Observatory in Hawaii. The average for 1958
was 315 parts per million (ppm); it is now near 354 ppm. Concentrations
before industries developed in the 19th century were once assumed to be
about 295 ppm. But estimates of this background concentration of carbon
dioxide, based on measurements made in ice cores that contain bubbles of
air trapped in preindustrial times and go back tens of thousands of years,
have now dropped to about 270 ppm. The rise in carbon dioxide concentration
to present-day levels should alone have increased the world’s average temperature
by about 1.5 °C, according to simple models of the greenhouse effect.
But levels of other major greenhouse gases have also increased considerably.
Although present in only tiny quantities, the combined effect of all the
gases is equivalent to 420 ppm carbon dioxide in the atmosphere – more than
halfway, in effect, to doubling the concentration of carbon dioxide.

Climate models made in the mid-1980s predict an equilibrium temperature
rise of 2.5 °C for this increase in carbon dioxide or its equivalent
in other gases. But the atmosphere is not in equilibrium at present, so
this is an overestimate of the warming that might be expected so far. The
most frequently cited records of global temperature are those published
by Phil Jones and Tom Wigley, of the Climate Research Unit at the University
of East Anglia, which show an actual rise of only 0.45 °C. Their records
also reveal some interesting timing; most of the warming in the northern
hemisphere happened before the Second World War, but most of the extra greenhouse
gases have been emitted since then. And in the southern hemisphere, which
should warm up least and slowest because of its oceans, the changes seen
so far are closer to those predicted by modelling.

So why do models say one thing and the climate data another? One big
factor is that the models calculate temperature rises expected at equilibrium,
and the Earth may take hundreds of years to reach this ideal, balanced state,
if it ever does. The circulation of water in the oceans has been cited as
one reason for overall global warming lagging behind that expected from
the increase in greenhouse gases. It takes hundreds of years for some ocean
water to reach the surface again after sinking to the sea floor. But this
time-lag cannot account for the whole difference. Generous estimates of
the effect of ocean circulation imply that the globe should still have warmed
by more than 1 °C, primarily after 1950. The net warming observed since
then is about 0.3 °C. The temperature of the surface of the sea scarcely
lags behind readings taken on land. If some interaction between ocean and
atmosphere is the major reason for the relatively small warming to date,
it must involve deep ocean circulation, something not taken into account
in most climate models.

But the mismatch between the predictions of models and actual warming
becomes worse when the temperature data are examined more closely. Towns
and cities are warmer than their surroundings, because they change the balance
of energy reflected and absorbed at the surface, perturb the wind locally
and alter the amount of cloudless sky, among other effects. Researchers
are not sure exactly how the various components of the urban effect operate,
but the difference is detectable for towns of 2500 people and more. And
as the world’s population increases and more people move to towns, the effect
grows. It would be surprising if the last years in the record were not the
warmest.

Tom Karl, of the US National Climatic Data Center, and his colleagues
have created a Historical Climate Network (HCN) of 492 stations that have
been adjusted to remove the effects of changes in the type of instruments
used, their location and the growth of neighbouring towns or cities. This
corrected record applies only to the US (apart from Alaska), but it serves
as an important check on other global records which may cover a much wider
area. Data from the HCN itself do not necessarily represent global trends;
in fact, they show a slow erratic cooling for much of the past 50 years.

Comparison between this and other records reveals that data such as
Jones and Wigley’s record contain an artificial warming of between 0.10
and 0.15 °C for the US stations. This reduces the overall warming in
their hemispheric record by an additional small amount, to between 0.35
and 0.4 °C. Analogous data from the Soviet Union show no overall urban
bias, apart from urban stations being cooler than surrounding rural ones
from 1953 to 1967. Mainland China had a similar urban cooling in the 1960s.

As further corroboration, the HCN record correlates better than any
other with temperatures measured by satellite from microwave radiation by
Roy Spencer and John Christy of NASA at Huntsville. Jones and Wigley’s record
is less than half as reliable. Since 1979, the Jones and Wigley data show
a dramatic warming relative to the satellite data in the southern hemisphere,
but no similar effect in the north. The good correlation between the satellite
data and the HCN for the US suggests that the satellite gives an accurate
record; it follows that a great deal of the warming in the last years of
the Jones and Wigley data set may be spurious and that the warming so far
may be considerably lower than 0.4 °C. The alternative is that the satellite
data are influenced by cooling in the stratosphere, especially near the
poles.

Land-based climate records also show an additional, insidious problem
– a general bias in the type of site chosen. The longest-standing records
tend to come from major towns and the predominance of water transport during
that the 19th century meant that they were usually by a river. Such sites
tend to show cold air drainage that buffers them from night-time warming.

It is possible to compensate both for population and site biases, by
using a combination of barometric readings on the ground and from weather
balloon ascents to assess temperature, data that cannot suffer from an urban
effect. In this way, Jim Angell, of the US National Oceanic and Atmospheric
Administration, found a net warming of 0.24 °C in the past 30 years
apart from transient phenomena. These results are consistent with the land
record worldwide.

Temperature records up in the air

I made a related study covering much of North America back to the beginning
of systematic balloon observations in 1948. To go back further, to 1885,
I used the frequency of cyclones, which is known to correlate well with
atmospheric pressure patterns. The variation in surface temperatures calculated
from this record was approximately three times larger than that measured
with thermometers on the ground. The rise in the first half of the 20th
century and the subsequent fall through the 1970s followed the same trends
as the ground-based record, although the magnitude was again greater. But
it is interesting to note that the warming in the early 20th century calculated
by this method, 1.8 °C, does not differ significantly from the expected
warming in the region calculated by climate models devised for the next
century.

All general circulation models of the world’s climate suggest that polar
warming will be magnified in winter. Yet there has been a substantial decline
in winter temperatures over the Atlantic Arctic since 1920, and no change
in night temperatures at the South Pole. Laurence Kalkstein and others at
the University of Delaware have recently shown that there has been no net
warming of the North American Arctic, but that the coldest air masses there,
with surface temperatures about -40 °C, have warmed some 2 °C.

Daily temperature ranges may also differ from the forecasts of the initial
models. Karl and his colleagues examined daily maximum and minimum values
from the HCN and found that the difference between them – the daily range
– has declined precipitously since 1950. Maximum values declined while minimum
values rose. Although the replacement of the traditional mercury-in-glass
thermometer by a new electronic version introduced a measurement bias towards
a lower daily range, this only explains about one quarter of the decline.

If greenhouse warming takes place mainly at night, the greenhouse world
will be very different from the gloomy prospects of popular imagination.
Warmer nights will not speed up evaporation as much as the same warming
in the daytime, so droughts will not become as frequent as otherwise expected.
The growing season will be longer, because it is limited primarily by low
temperatures at night. More carbon dioxide in the atmosphere could even
speed up the growth of some plants.

Extra carbon dioxide appears to stimulate growth by increasing both
the rate at which plants fix carbon and the efficiency with which they use
water. Plants of the greenhouse world are likely to grow more rapidly for
each unit of moisture they use. And if the climate of that era is dominated
by night warming – which will lengthen the growing season without substantially
increasing the rates of evaporation – the planet may become considerably
greener.

There is evidence that some plants have already begun to adapt. The
British Museum has plant specimens collected in the 19th century, before
any major increase in carbon dioxide. They have more stomata than members
of the same species collected today. Stomata are the pores through which
plants exchange gases with the atmosphere and are the major pathway for
loss of moisture. Fewer stomata suggest that the plants have become more
water-efficient. Laboratory experiments also demonstrate that plants in
an atmosphere with considerable extra carbon dioxide have fewer stomata
than those in a normal atmosphere.

Plants in the real world also seem to be growing faster. Surveys of
mountain species in the western US, a virgin forest plot being carefully
monitored by the Oak Ridge National Laboratory, and the northern forests
of Scandinavia all reveal faster growth, inexplicable by any change in the
weather. This acceleration has taken place even at a time when forests are
supposedly being harmed by acid rain and snow. The alteration is not limited
to woody plants and shrubs; the Box describes changes in corn yield in the
southeastern US.

A decreasing difference between day and night temperatures is consistent
with an enhanced greenhouse effect when the observed reduction in sunshine
and 3.5 per cent increase in cloudiness across the US are taken into account.
In addition, Gerd Weber found a decline in sunshine in Germany, an effect
that is greater in the mountains – which implicates stratocu-mulus, the
low-altitude cloud type most effective at surface cooling. Warren and others
have also found increasing cloudiness in most measurements made at sea,
although the shipboard observations used are not wholly objective. Nonetheless,
the cloud type that shows most increase is again the low-level stratocumulus,
and again it has increased most in the northern hemisphere. The pattern
of a decline in ultraviolet radiation at low elevations and increased values
higher than 10 kilometres is also consistent with an increase in low-level
cloudiness.

In contrast to the story in the US, the ranges of temperature experienced
in the southern hemisphere are not declining. The most likely cause of such
an important difference would be an increase in low-level cloudiness in
the north that is not matched in the south. And a prime candidate for the
cause of such a difference is sulphate particles, produced either directly
or by the oxidation of sulphur dioxide from coal burnt in power stations,
for example. These particles reflect solar radiation and make clouds brighter
by serving as condensation nuclei. They are produced in much lower volume
in the southern hemisphere, and stay in the atmosphere for only days or
months; very little diffuses from the northern to the southern hemisphere.

If sulphate particles are helping to increase clouds that cool the climate,
this is one form of human activity that can serve to counter the greenhouse
effect. Mayewski and colleagues recently demonstrated that the sulphate
now in the atmosphere of the northern hemisphere from human activity is
equivalent to the maximum estimated to have been produced by the Tambora
volcano, which triggered a cooling of as much as 2 °C after it erupted
in 1815. Because these sulphates are produced primarily in the northern
hemisphere and tend to stay there, we may be equalling or overcompensating
for current increase in infrared absorption in the hemisphere that contains
most of the world’s population.

So, is there any local increase in the reflectivity or ‘albedo’ of low-level
clouds near sources of sulphates? This is where concentrations should be
greatest, because the particles will have had little chance to disperse
or decay. Satellite data show an increase in brightness in stratocumulus
near the ocean surface, especially near Asia and North America, which are
abundant sources of sulphate. Reflectivity increases by as much as 8 per
cent, and the effect persists thousands of kilometres downwind. A ‘clean’
swath of the South Pacific ocean served as a control.

Mitigation of greenhouse warming by particulates such as sulphates would
account for what we have observed: increased night warming and a decreased
daytime warming because there are more clouds, decreased daily temperature
range, different effects in the northern and southern hemispheres, and brightening
of low-level clouds near sources of sulphates. This makes the emission of
such particles by industrial activity the most likely explanation for the
lack of major warming despite the other factors working to bring about greenhouse
warming.

Do these results bolster the argument for an accelerated timetable to
reduce or eliminate the combustion of fossil fuel, or do they suggest that
other courses of action are called for? Carbon dioxide stays in the atmosphere
far longer than sulphates, so a substantial potential warming could already
be built into the system, hidden by sulphates. Reduce sulphur dioxide emissions,
and we might see the full force of the warming pulse. Perhaps we have inadvertently
discovered an effective control technology against greenhouse warming. At
the least, we may have fortuitously bought enough time to develop a solution
that is less disruptive than the 60 or 80 per cent reductions in carbon
emissions that are generally considered necessary to stabilise the concentrations
of carbon dioxide in the atmosphere.

Patrick Michaels is associate professor of environmental sciences at
the University of Virginia, where David Stooksbury is a doctoral candidate.

* * *

Is climate change already giving us greater maize yields?

Maize is a vital food crop for people in many areas of the world. Conventional
wisdom and a voluminous scientific literature support the notion that, in
general, the greater the summer rainfall, the greater the yield. Rain is
especially vital during the period of pollination. Summer temperatures above
normal are thought to decrease the yield. These ideas have been confirmed
by research in major maize producing regions, which generally have a humid
continental climate.

Do these climatic factors apply in the same way to maize growing in
other regions? In the past, scientists have said yes. But the answer may
not be so simple.

The climate of the southeastern US is humid subtropical. It is 2 °C
warmer during summer than the major maize production area of the US, the
‘corn belt’ of the eastern Great Plains and the Midwest. Rainfall during
the summer months is similar and comes from thunderstorms. From an agricultural
point of view, the major difference between these two regions is the length
of the growing season.

To find out how climate alters the yield of maize in this humid subtropical
area I developed a family of statistical models, similar to those used by
the US Environmental Protection Agency. These models predict maize yield
for areas covering thousands of square kilometres. I considered a range
of climate variables, such as average daily maximum and minimum temperatures,
average temperature and total precipitation, each month by month. Only climate
variables that affect maize yield are retained in the final models.

The model does an excellent job of predicting maize yield (see Figure).
Summer precipitation turns out to be relatively unimportant for maize yields
in the southeastern US; it is overshadowed by the effects of temperature.

In the major maize producing region of the southeastern US, the area
between 34 degrees and 38 degrees North, the maximum temperature in July
is the most important climatic variable. The higher the maximum July temperature,
the more the maize yield drops.

The second most important variable is average minimum temperature in
July. The higher the minimum temperature, the greater the maize yield. The
greatest maize yield in this region would come if the daily temperature
range in July decreased. This is exactly what is predicted in a greenhouse
climate and what is being seen in the climate records of the southeastern
US.

The results from the southeast may also have implications for maize
yield in the US corn belt. The most important change expected there will
be an increase in the length of the growing season. Planting times will
then correspond generally with the current planting times in the southeast.
Earlier planting will allow the Midwestern maize to pollinate before the
extremely hot days of late July. Cooler weather at pollination will boost
maize yields. The other climatic variables will have minimal effect. The
major change in maize production in a greenhouse world will not be in yield
but in planting time; farmers will be able to plant the crop earlier.

David Stooksbury