Aerosols that originate from natural emissions may also be expected to change
in future scenarios. For example, terpene emissions depend on temperature, precipitation
and light levels, and sea salt emissions depend on wind speed and temperature.
DMS emissions depend on wind speed and temperature, and dust emissions depend
on soil moisture and wind speed. Changes to natural emissions associated with
changes to these factors were considered in scenarios SC5 and SC8.
To construct future changes in natural emissions, the NCAR CSM simulations
(Dai et al., 2001) were used to estimate possible changes in climate. This climate
simulation was one formulated to treat a "business as usual" (IS92a)
scenario and resulted in a global average surface temperature change for the
decade prior to 2100 relative to that for the decade prior to 2000 of 1.76°K.
While methods are available to estimate the effect of changes in climate on
natural emissions, given current vegetation cover, tools to define the impact
of future changes in land use and land cover on spatially disagreggated emissions
are not well developed. Thus, even though changes in land use would be expected
to affect vegetation cover and this is responsible for emissions of terpenes
and determines which areas are subject to dust uplift, these effects could not
be included. Furthermore, the understanding of how phytoplankton populations
that produce DMS may change with changes in climate is poor. Therefore, in the
following, only the effects of changes in temperature, precipitation, light
levels, and wind speeds on future emissions were considered. The changes in
wind speed were estimated from the ratio of monthly average wind speeds for
the years 2090 to 2100 to that for the years 1990 to 2000 associated with the
IS92a scenario (Dai et al., 2001). This ratio was used to adjust the wind speeds
that were used to generate current (2000) emissions to obtain future (2100)
emissions of both DMS and dust. This method gives only a first-order estimate
of possible changes, since it assumes that the distribution of wind speeds within
a month remains constant with time. For wind speeds associated with sea salt,
a somewhat different method was used (see below). The terpene fluxes were estimated
directly from the daily data that were projected by the CSM simulation for 2100.
Emissions of DMS were projected using the procedures described in Kettle et
al. (1999). Since the atmospheric concentration of DMS is negligible in relation
to what it would be if the atmospheric and oceanic concentrations were in equilibrium,
the DMS flux is assumed to depend only on the sea surface concentrations and
the wind speed. The projected DMS flux (as well as that for 2000) used the average
of the parametrizations of Liss and Merlivat (1986) and Wanninkhof (1992). A
correction of the piston velocity for sea surface temperature was made using
the Schmidt number dependence of Saltzman et al. (1993).
The DMS flux for 2000 was calculated from the monthly average sea surface
temperature data of Levitus and Boyer (1994) and the sea surface DMS concentration
data of Kettle et al. (1999). In addition, climatological wind speeds from Trenberth
et al. (1989) and climatological sea ice cover fields from Chapman and Walsh
(1993) were used. The global flux for the year 2100 was calculated as the product
of this initial field and the ratio of the piston velocity field in 2100 to
that in 2000 using the sea surface temperature and wind speed information from
the NCAR CSM.
There are several possible sources of error in these calculations. The most
serious assumption is that the DMS concentration fields do not change between
the years 2000 to 2100. DMS is produced as part of phytoplankton bloom cycles,
especially in high latitude areas. It is likely that the mean distribution of
phytoplankton blooms in the upper ocean would change between 2000 and 2100 given
any perturbation of the sea surface temperature, wind speed, and sunlight. The
other major assumption is that the monthly climatological ice cover does not
change between 2000 and 2100. Ice acts as a lid on the ocean in upper latitudes
through which DMS cannot pass.
Overall, the calculations suggest a small increase in global DMS flux between the year 2000 (with a global DMS flux of 26.0 TgS/yr) and the year 2100 (with a global DMS flux of 27.7 TgS/yr). The most noticeable features in the 2100 fields are the predicted increases in DMS fluxes in some areas of the North Atlantic, North Pacific, and some areas of the Southern Ocean immediately adjacent to the Antarctic continent. There are some localised increases predicted in the tropical and sub-tropical Pacific Ocean.
Other reports in this collection