Environmental Engineering Reference
In-Depth Information
rel ectance, information on biome type, the fraction of photosynthetically active
radiation that is absorbed by vegetation (a variable that changes with plant growth
and senescence), and daily surface climate conditions made it possible to quantify
continuous l uxes of terrestrial GPP and to make spatially explicit GPP/NPP esti-
mates that could be compared on seasonal and annual bases, and hence made it
possible, for the i rst time, to discern any longer-term trends in global primary pro-
ductivity (Heinsch et al. 2003; Running et al. 2004).
In 2000, during the i rst full year of MODIS operation, the GPP and NPP totals
were, respectively, 108.42 and 56.06 Gt C/year (NTSG 2006), and the means for
the next two years were 109.29 and 56.02 Gt C/year (Zhao et al. 2005). A MODIS-
based GPP has been routinely calculated for eight-day periods since February 2000,
and archived information is available either as global and regional maps with 1 km
resolution or as data sets (USGS 2010). Matching the 1 km resolution of the sensors
with plot-scale measurements on the ground makes the validation of MODIS-
derived NPP values difi cult. Comparisons of the MODIS-derived GPP with the GPP
calculated from tower-based measurements of CO
2
for forest and grassy ecosystems
showed that the satellite-based values tended to overestimate GPP at sites with
low productivity and underestimate the l uxes at high-productivity sites (Zhao et
al. 2005; Turner et al. 2006).
Beer et al. (2010) approached the problem by combining eddy covariance l ux
data (high-frequency vertical wind vector and CO
2
and H
2
O concentration measure-
ments obtained using a using 3-D anemometer and an infrared gas analyzer mounted
on platforms high above the vegetation canopy) and process-oriented biosphere
models to quantify terrestrial gross CO
2
uptake, its global distribution, and its
covariation with climate. They found that the GPP over more than 40% of the
vegetated land is associated with precipitation (up to 70% in grasslands, shrublands,
and croplands; only 30% of GPP variability in tropical and boreal forests, where
temperature plays a greater role) and that tropical ecosystems (forests and savannas)
account for 60% of total GPP of 123
8 Gt C/year.
When Ito (2011) surveyed the relevant literature from 1862 to 2011, he found
251 estimates of total terrestrial NPP, whose mean, standard deviation, and median
were respectively 56.2, 14.3, and 56.4 Gt C/year. His original intent was to see
whether the estimates of global NPP were converging, but he found that even the
estimates published after 2000 had substantial uncertainty (coefi cient of variation
±
±
15%). This uncertainty has only increased with the publication of a new paper on
interannual variability in the oxygen isotopes of atmospheric CO
2
driven by El Niño
(Welp et al. 2011). The paper showed how the decay time of the El Niño anomaly
