354
mate feedback responses to the direct effect of increasing infrared absorption in the atmosphere.
This temperature increase is expected to be heterogeneous in both time and space. Winter temperatures are expected to increase more than summer
temperatures and the polar regions are expected to
warm more than lower latitudes. Significant precipitation increases are expected at high latitudes
and smaller changes (of either sign) at lower latitudes. Other expected effects of global warming include increased evaporation, stratospheric cooling,
rising sea level, changes in cloud cover, changes in
soil temperature and moisture levels, and changes
in the intensity and frequency of storms and
droughts.
One of the most critical meteorological parameters governing terrestrial ecosystem function is
soil moisture, which in tum is linked to temperature, precipitation, humidity, the pattern of snow
melt, the distribution of vegetation, and extreme
weather events. General circulation models can
yield only broad speculations regarding impacts of
increased greenhouse gases on soil, especially soil
moisture. All models predict increased soil moisture in high northern latitudes in winter. Most models predict increased aridity in midlatitude land
areas in summer, mainly due to enhanced summertime evaporation (Hansen et al. 1988a; Hansen et
al. 1988b; Manabe and Weatherald 1987). Also,
snowpacks will melt earlier and summer depletion
of the soil moisture reservoir will begin earlier
(Hartmann 1990; Kattenberg and Maske111995). In
areas with marginal soil moisture or seasonal
drought, it is likely that the probability and intensity
of drought will increase. Summer soil moisture reductions and increased probability of summer
drought, particularly in midlatitudes, would have
serious implications for agriculture and water resource allocation. Summer soil moisture in northern midlatitudes increases in some regions, however, when aerosol effects are included in the
model, in contrast to the decrease found in
greenhouse-gas-only simulations (Kattenberg and
Maskell 1995).
Global warming could have far-reaching effects
on the biosphere, and while GCMs attempt to predict the magnitude and scope of such warming, specific information regarding the potential impacts on
natural systems remains limited. Little is known
about the detailed consequences of rapid climate
Karin P. Shen and John Harte
change on ecological processes. Soil conditions and
the response of ecosystems will depend on local
changes in temperature (including its rate of
change), precipitation, soil moisture, and extreme
weather events. Predicting the consequences of a
small change in overall mean temperature for
whole-ecosystem function is therefore going to be
very difficult. Furthermore, a major difficulty in
predicting the direct ecological consequences of
global warming is that there is massive local heterogeneity of temperature within the same ecosystem at any given time. Current models are unable
to make reliable estimates of changes in the critical
parameters on the required local scales. The consequences of changes in global mean temperature
for the structure and function of terrestrial ecosystems are largely unknown.
One approach to modeling the biosphere is to
link a GCM with a land surface process model
(Dickinson et al. 1986; Harvey and Schneider
1985; Randall et al. 1996; Sato et al. 1989; Sellers
et al. 1986; Sud et al. 1990; Verseghy et al. 1993).
The land surface components of coupled climate
models range from simple schemes to complex representations of soil and vegetation (Gates et al.
1995). Developing a land surface model that can
interact and evolve with the atmosphere as it
changes (two-way interaction) is an active research
area in climate modeling. All GCM simulations to
date have ignored vegetation changes, and nearly
all current land surface schemes involve only oneway atmosphere-biosphere interaction. HendersonSellers and McGuffie (1994) made the first attempt
to include the terrestrial biosphere as an interactive
component of a global model applied to estimating
the sensitivity of the climate to a doubling of the
atmospheric concentration of CO 2 • There is also
avid interest in developing a comprehensive land
surface scheme that integrates biophysical, biogeochemical, and ecosystem dynamical processes.
Some recent land surface schemes incorporate biogeochemical and ecological knowledge and, when
coupled with atmosphere-ocean GCMs, will be capable of modeling biological and physical responses to global change (Sellers et al. 1997). For
example, the IBIS (integrated biosphere simulator)
model of Foley et al. (1996) combines land surface
biophysics with terrestrial carbon fluxes and global
vegetation dynamics.
mate feedback responses to the direct effect of increasing infrared absorption in the atmosphere.
This temperature increase is expected to be heterogeneous in both time and space. Winter temperatures are expected to increase more than summer
temperatures and the polar regions are expected to
warm more than lower latitudes. Significant precipitation increases are expected at high latitudes
and smaller changes (of either sign) at lower latitudes. Other expected effects of global warming include increased evaporation, stratospheric cooling,
rising sea level, changes in cloud cover, changes in
soil temperature and moisture levels, and changes
in the intensity and frequency of storms and
droughts.
One of the most critical meteorological parameters governing terrestrial ecosystem function is
soil moisture, which in tum is linked to temperature, precipitation, humidity, the pattern of snow
melt, the distribution of vegetation, and extreme
weather events. General circulation models can
yield only broad speculations regarding impacts of
increased greenhouse gases on soil, especially soil
moisture. All models predict increased soil moisture in high northern latitudes in winter. Most models predict increased aridity in midlatitude land
areas in summer, mainly due to enhanced summertime evaporation (Hansen et al. 1988a; Hansen et
al. 1988b; Manabe and Weatherald 1987). Also,
snowpacks will melt earlier and summer depletion
of the soil moisture reservoir will begin earlier
(Hartmann 1990; Kattenberg and Maske111995). In
areas with marginal soil moisture or seasonal
drought, it is likely that the probability and intensity
of drought will increase. Summer soil moisture reductions and increased probability of summer
drought, particularly in midlatitudes, would have
serious implications for agriculture and water resource allocation. Summer soil moisture in northern midlatitudes increases in some regions, however, when aerosol effects are included in the
model, in contrast to the decrease found in
greenhouse-gas-only simulations (Kattenberg and
Maskell 1995).
Global warming could have far-reaching effects
on the biosphere, and while GCMs attempt to predict the magnitude and scope of such warming, specific information regarding the potential impacts on
natural systems remains limited. Little is known
about the detailed consequences of rapid climate
Karin P. Shen and John Harte
change on ecological processes. Soil conditions and
the response of ecosystems will depend on local
changes in temperature (including its rate of
change), precipitation, soil moisture, and extreme
weather events. Predicting the consequences of a
small change in overall mean temperature for
whole-ecosystem function is therefore going to be
very difficult. Furthermore, a major difficulty in
predicting the direct ecological consequences of
global warming is that there is massive local heterogeneity of temperature within the same ecosystem at any given time. Current models are unable
to make reliable estimates of changes in the critical
parameters on the required local scales. The consequences of changes in global mean temperature
for the structure and function of terrestrial ecosystems are largely unknown.
One approach to modeling the biosphere is to
link a GCM with a land surface process model
(Dickinson et al. 1986; Harvey and Schneider
1985; Randall et al. 1996; Sato et al. 1989; Sellers
et al. 1986; Sud et al. 1990; Verseghy et al. 1993).
The land surface components of coupled climate
models range from simple schemes to complex representations of soil and vegetation (Gates et al.
1995). Developing a land surface model that can
interact and evolve with the atmosphere as it
changes (two-way interaction) is an active research
area in climate modeling. All GCM simulations to
date have ignored vegetation changes, and nearly
all current land surface schemes involve only oneway atmosphere-biosphere interaction. HendersonSellers and McGuffie (1994) made the first attempt
to include the terrestrial biosphere as an interactive
component of a global model applied to estimating
the sensitivity of the climate to a doubling of the
atmospheric concentration of CO 2 • There is also
avid interest in developing a comprehensive land
surface scheme that integrates biophysical, biogeochemical, and ecosystem dynamical processes.
Some recent land surface schemes incorporate biogeochemical and ecological knowledge and, when
coupled with atmosphere-ocean GCMs, will be capable of modeling biological and physical responses to global change (Sellers et al. 1997). For
example, the IBIS (integrated biosphere simulator)
model of Foley et al. (1996) combines land surface
biophysics with terrestrial carbon fluxes and global
vegetation dynamics.