Cotton W.R., Pielke R.A. Human Impacts on Weather and Climate
Подождите немного. Документ загружается.
Influence of irrigation 129
excluded in computing this temporal average because a fairly strong mid-level
low stalled over the study area during that period. This synoptic environment
was responsible for the unusual depression in both the minimum and maximum
temperatures recorded in the early part of July 2000 in Nebraska (Figs. 6.21 and
6.22).
Important physical changes between the natural shortgrass prairie of this region
and the current land-use patterns include alterations in the surface albedo, rough-
ness length changes, and increased soil moisture in the irrigated areas. These
changes are capable of generating complex changes in the lower atmosphere (plan-
etary boundary layer, PBL) energy budget. For example, the simulated increase
in the portion of the total available energy being partitioned into latent heat
40.00
(a)
(b)
35.00
30.00
25.00
Temperature (°C)Dewpoint temperature (°C)
20.00
15.00
10.00
5.00
0.00
30.00
25.00
20.00
15.00
10.00
5.00
0.00
123 456
Min Temp Obs
Min Temp Model
Max Temp Obs
Max Temp Model
78
July 1997
July 1997
Mean Dewpt ObsMean DewPt Model
910
11
12 13 14 15
123 45678910
11
12 13 14 15
Figure 6.21 Observed minimum and maximum temperatures for Ainsworth
municipal compared to model 2 m temperature. Reproduced from Adegoke et al.
(2003) with permission from the American Meteorological Society.
130 Other land-use/land-cover changes
40.00
(a)
(b)
35.00
30.00
25.00
Temperature (°C)
Dewpoint temperature (°C)
20.00
15.00
10.00
5.00
0.00
30.00
25.00
20.00
15.00
10.00
5.00
0.00
123 456
Min Temp Obs
Min Temp Model
Max Temp Obs
Max Temp Model
78
July 1997
July 1997
Mean DewPt ObsMean DewPt Model
910
11
12 13 14 15
123 45678910
11
12 13 14 15
Figure 6.22 Same as 6.21 except from Omaha/Eppley Field. Reproduced from
Adegoke et al. (2003) with permission from the American Meteorological
Society.
rather than sensible heat resulted directly from the enhanced transpiration and soil
evaporation in the control run.
The analyses of domain-averaged surface temperature and surface fluxes for this
Nebraska study clearly indicate that key components of the planetary boundary
layer, particularly the partitioning of the available surface energy in this region, are
very sensitive to changes resulting from increased irrigation. RAMS simulations
forced, in part, by four different land-cover scenarios indicate a cooling in the
near-ground temperature, and significant increases in vapor and latent heat fluxes
into the atmosphere in response to this human disturbance of the landscape. These
changes are shown to be related to the conversion of the natural prairie vegetation
in this region to irrigated and dry cropland. Corroborating evidence from analyses
Desertification 131
of long-term surface climate data indicates a steady decreasing trend in mean
and maximum air temperature at locations located within heavily irrigated areas
thus supporting the irrigation-induced cooling effect suggested by the RAMS
simulations.
6.3 Dryland agriculture: Oklahoma
McPherson et al. (2004) analyzed a number of related observations and demon-
strated that Oklahoma’s winter wheat belt had a significant impact on the near-
surface temperature and moisture fields during the period when winter wheat
was growing and during the period after harvest. As they reported, as vegetation
grew across the wheat belt, maximum daily temperatures were cooler than those
measured over adjacent regions of dormant grasslands. Monthly averaged values
of maximum temperature from crop year 2000 displayed a cool anomaly over the
growing wheat from November 1999 through April 2000. Using data from a net-
work of surface stations from 1994 through 2001, the cooler temperatures over the
wheat belt were shown to be statistically significant at the 95% confidence level
for November, December, January, February, and April. As green-up of grass-
lands occurred during May, the cool anomaly over the wheat belt disappeared.
After the wheat was harvested, maximum and minimum daily temperature data
revealed a warm anomaly across the wheat belt during June, July, and August.
The warmer temperatures also were shown to be statistically significant for all
three months.
Cooley et al. (2005) also found a substantial effect on regional climate due to
dryland farming. In the southern Great Plains of the United States, where winter
wheat accounts for 20% of the land area, within the harvested region simulated
2 m air temperatures were 13
C warmer at midday averaged over the 2 weeks
following the harvest.
6.4 Desertification
6.4.1 Historical overview
Desertification has an opposite effect on climate to that of irrigation. Desertifi-
cation was originally defined by Aubreville (1949) as cited in Odingo (1990),
as land degradation in semi-arid and subhumid regions such that arid conditions
develop and persist. Vegetation loss due to human activities can, therefore, create
deserts. The Rajasthan Desert in northwest India is an example of desertification.
In that area, in the millennium before the Christian era, a well-developed civi-
lization existed. Currently the region is a tropical desert with only archeological
ruins remaining. It has been suggested that overpopulation in the area denuded
132 Other land-use/land-cover changes
the vegetation, as the populace used firewood and cleared land for agriculture.
While the change in weather could have been part of the natural evolution of
climate, major landscape changes by humans on the scale of this desert region
have occurred elsewhere.
In the Middle East, Neumann and Parpola (1987), for example, have docu-
mented that the political, military, and economic decline of Assyria and Babylonia
in the twelfth through tenth centuries BC coincide with a notable period of warm-
ing and drying in the region which started around 1200 BC and lasted until
about 900 BC. While natural large-scale fluctuations in climate could have caused
this aridity, as implied by their paper, desertification provides another explana-
tion. The return of weather to wetter conditions could have occurred in response
to a rejuvenation of vegetation as human stress on the landscape was reduced,
and the larger-scale weather patterns which influence the region were not more
permanently displaced as apparently has occurred in northwest India.
6.4.2 North Africa
Charney (1975) proposed a mechanism of desertification over northern Africa in
which the removal of vegetation increased the albedo of the land. Since more
solar radiation was reflected back into space, the result was increased subsidence
in the lower atmosphere in order to compensate for the loss of heat energy. In a
later study, Charney et al. (1977) used the Goddard Institute for Space Studies
(GISS) model to conclude that local evaporation rates are as important as albedo
in influencing rainfall patterns in semi-arid regions, with their relative effects
dependent on location. This study found a more complicated interaction between
the land surface and the atmosphere than was originally hypothesized in the
Charney (1975) paper, with significant research work supporting this conclusion
repeated in Claussen et al. (1999). They found that rapid natural desertification in
the mid-Holocene could only be reproduced in a model if atmospheric–vegetation
feedback dynamics were included.
The process of desertification continues today in the Sahel regions of Africa
as illustrated in Fig. 6.23 in which a portion of Chad is shown. The darker region
in the figure corresponds to an area in which the government controlled graz-
ing while the adjacent areas were extensively utilized by cattle and goats. The
significantly higher albedo observed in the overgrazed area is evident in the figure.
6.4.3 Western Australia
In western Australia, as reported by Lyons et al. (1993), the use of fencing to
prevent the movement of rabbits into an agricultural area is evident from satellite
Desertification 133
Figure 6.23 Landsat imagery of Chad.
imagery. In this region, 130 000 km
2
of native perennial vegetation has been
transformed into winter-growing annual species. Lyons reported a 30% decline of
winter rainfall in this region which he attributes to the change of vegetation types.
Presumably, the winter crop is harvested such that subsequent transpiration of
water into the atmosphere is lost, whereas the perennial plants would be reaching
their peak growth (and transpiration) in the latter part of the winter and in the
spring.
6.4.4 Middle East
Otterman (1974) and Otterman and Tucker (1985) also used an area of con-
trasting land use in the Negev Desert–Sinai Peninsula region to demonstrate
how desertification works. Figure 6.24a presents satellite imagery of this region,
in which the political boundary between Egypt and Israel is clearly shown. In
Israel grazing was controlled such that dark plant debris and limited clumps
of living vegetation reduced the albedo of the surface. In the Egyptian area,
in contrast, overgrazing by goats permitted few plants to survive so that the
soil became more mobile and covered the vegetation debris. The result was a
higher albedo from the light bare soil. Figure 6.24b documents using a numeri-
cal model simulation that surface ground temperature differences over 5
C can
result with the desertified area being cooler. Otterman suggests that these cooler
temperatures result in less annual precipitation than otherwise would occur due
to the reduced surface supplied buoyant energy for showers. Thus it would be
less likely for the vegetation to regenerate even if the grazing pressure were
removed.
134 Other land-use/land-cover changes
(a)
316
x
y
Predicted surface temperature at 1300 LST
312
(b)
314
Figure 6.24 (a) This MODIS view shows the denuded high albedo regions
of the Sinai and Gaza Strip, in contrast to the darker western Negev. Sensor:
Terra/MODIS; Datastart: 2000-09-10; Visible Earth v1 ID: 5606; Visualization
date: 2000-10-12. Courtesy of NASA Visible Earth and Jacques Descloitres,
MODIS Land Science Team. See also color plate. (b) Predicted surface temper-
ature at 1300 LST for typical fall meteorological conditions over the same area.
From Mahrer and Pielke (1978).
Deforestation 135
Mahrer and Pielke (1978) performed a model simulation of this region to
show that the gradient of surface temperatures due to the albedo differences
will cause preferential ascent (and thus more likely showers when the atmo-
sphere is favorable for cumulus convection over the Negev side of the bound-
ary). Otterman et al. (1990) suggest that natural vegetation in a semi-arid or
arid region could have a more positive impact on rainfall than the tarring
of high albedo sandy desert areas as proposed by Black and Tarmey (1963).
Otterman et al. concluded that land management has a significant impact on
climate.
6.5 Deforestation
6.5.1 Historical perspective
Deforestation is another type of landscape change. Deforestation occurs when
lumbering and fires remove extensive areas of trees from a region. Such defor-
estation occurred, for example, in the eastern United States in the nineteenth
century and resulted in an almost total removal of the original uneven aged, cli-
max forest. Only in the Great Smoky Mountain Forest, where about 40% of the
virgin trees remain, can one appreciate the large diameter deciduous and ever-
green trees and sparse understory vegetation that originally covered the area prior
to European settlement. During the last 50 to 75 years, as small farms became
less economically viable, large areas of the forest has returned as the nutrient-
rich organic soils in the midlatitude region permitted regrowth from root sprouts,
natural seedings, and forestry planting of seedlings. Thus while the deforesta-
tion was extensive, a new forest, albeit of a different species composition and
age, has evolved in non-agricultural and non-urban areas in the eastern United
States.
6.5.2 Amazon
More recent deforestation has occurred in the Amazon region and the influence
of this removal of trees on global climate has been the subject of considerable
interest (Dickinson, 1987; Lean and Warrilow, 1989). There are regional effects
of this deforestation as well. Currently, it is estimated that the water vapor
for about 50% of rainfall within the Amazon Basin is from local evaporation
and transpiration (Salati et al., 1978). Figure 6.25 shows that the removal of
trees is not completed as one vast clear-cutting but is accomplished through
a series of strips of several kilometers in width and in large clear-cut areas.
Local wind circulations develop in response to these landscape alterations
136 Other land-use/land-cover changes
Figure 6.25 Examples of clear-cutting of the tropical forest in two areas of the
Amazon. Photos provided by Carlos Nobre of the Center for Weather Prediction
and Climate Studies – CPTEC, INPE, Brazil. See also color plate.
Deforestation 137
that result in an enhancement of local rainfall and cloudiness (Avissar and
Liu, 1996).
A major difference between the midlatitude forests and forests in the humid
tropics such as the Amazon is the absence of substantial organic material and
trace metals in the soils in the tropical environment. The heavy rains in the region
leach nutrients to below the root layers with the result that much of the nutrients
required for vegetation development and growth are within the living and recently
dead plants. A rapid recycling of nutrients occurs as new vegetation generates
from the rapidly decomposing plants. In the absence of this rapid recycling, rains
would deplete the root zones of the needed plant food. Thus, the removal of trees
by lumbering short-circuits the recycling. In the absence of these nutrients, even
an increase of rainfall due to the patchiness of the timber cutting may not benefit
the regeneration of the forest.
Several field studies have been undertaken in the Amazon in order to better
understand the influence of this natural vast forest area on climate and weather. In
1985 and 1987, the Amazon Boundary Layer Experiment (ABLE) was completed
(Garstang et al., 1990). Among the results it was confirmed that local wind
circulations apparently develop in response to large rivers in the area being
adjacent to the forests (Miller et al., 1988). A similar atmospheric response would
be expected when large areas are cleared of trees. The Anglo-Brazilian Amazonian
Climate Observation Study (ABRACOS) was initiated at the end of 1990 to
investigate the influence of clear-cutting in the region. Results from the field study
(see Fig. 6.26) demonstrate that the forest captures more radiative energy than
an area which is pasture. During the daytime, the albedo, on average, is about
2.5% larger over the cleared area with the resultant wind speed, temperature, and
moisture profiles shown in Figs. 6.27, 6.28, and 6.29.
These measurements show that the temperature change is twice as great over
the cleared area with even a greater variation in wind speed. Early morning fog
or mist is observed in the clearings, while such moisture saturation is almost
unknown in the forest.
6.5.3 Africa
The influence of deforestation in another tropical area, southern Nigeria, is dis-
cussed in Ghuman and Lal (1986). Figure 6.30, from that study, is consistent
with the results from the Amazon in that the diurnal range of temperature is
substantially greater over the cleared ground. This figure also shows that the
influence of forest results in somewhat cooler temperatures deeper in the soil
(i.e., 50 cm).
138 Other land-use/land-cover changes
Clearing
Forest
24
22
20
18
16
14
12
10
61218
Time (h)
Albedo (%)
(a)
01224
Time (h)
Clearing
Forest
900
600
300
0
–100
(b)
Net radiation (W m
–2
)
Figure 6.26 (a) Mean hourly albedo at the clearing and forest sites in the
Amazon for the study period from 12 October to 10 December 1990. (b) Mean
hourly net all-wave radiation at the clearing and forest sites for the same study
period. From Bastable et al. (1993).
6.6 Regional vegetation feedbacks
Eastman et al. (2001a,b) have shown that land-use change, grazing, and the
biogeochemical effect of increased carbon dioxide can significantly alter the
regional climate system in the central Great Plains of the United States. Figure 6.31