Environmental Engineering Reference
In-Depth Information
2071-2100 ranging from 3.7°C to 6.2°C above the 1971-2000 mean, and with
the increases in annual continental precipitation between 6.5% and 13.8%—
to the Lund-Potsdam-Jena global vegetation model and found the following large
uncertainties.
Even the overall response of the biosphere was unclear, with changes ranging
from the loss of 106 Gt C to the additional storage of 201 Gt C, with three out of
i ve climate models indicating that the biosphere will be a net source of carbon
(particularly thanks to boreal forests), one yielding a neutral outcome, and one
implying a carbon sink, and with the response of the tropical rain forests of South
America and Central Africa being most uncertain. Moreover, the relative agreement
between different models was less for individual seasons than for the annual mean.
This implies not only a large relative error—assuming 650 Gt C of plant carbon in
the year 2000, the range would be +30% to -16%—but an absolute uncertainty
(about 300 Gt C) nearly as large as the most likely total of plant carbon decline
that the biosphere has experienced since the mid-Holocene.
Land for Food Production
No other human activity has been responsible for such a large-scale transformation
of terrestrial ecosystems as food production. All kinds of natural ecosystems have
been converted to permanent i elds; this category of land use is still commonly
labeled arable land, although in some countries substantial shares of it are now
managed with reduced tillage or no-till cultivation. This land claimed no less
than 1.380 Gha in 2010, or nearly 11% of all continental surfaces. The addition
of permanent plantations (growing fruits, tea, coffee, cacao, palms) extends the
total of about 1.53 Gha of agricultural land, or about 12% of the earth's ice-free
surface (FAO 2011f). For comparison, the total based on combining various
agricultural inventory data satellite-derived land coverage (at 5-minute or roughly
10 km resolution) came up with a nearly identical median value of 1.5 Gha and
a 90% coni dence range of 1.22-1.71 Gha in the year 2000 (Ramankutty et al.
2008).
These totals refer to the existing stock of farmland, not to the area of all crops
that are actually planted in a calendar- or a crop-year; in national terms that aggre-
gate can be both a bit lower and substantially higher. This is because in many
countries some arable land is always fallowed, while variable shares of farmland
are planted to more than one crop a year. Consequently, multicropping ratios—
expressing the number of crops harvested regionwide or nationwide per unit of
