the left are measured, ΔO 2 from Δ(O 2 /N 2 ) by assuming N 2 is nearly fixed (with
approximation for release on warming). On the right, F is reported by governmental
agencies tracking coal mining and gas and oil production. The equations are solved
for Oc and L. According to Manning and Keeling (2006) the sinks for 1993 to 2003
were: Oc = 2.2 ± 0.6 and L = 0.5 ± 0.7 GtC yr−1. In those years, F ≈ 6.5 GtC yr−1, so
the ocean took up about a third of the fossil-fuel CO 2 , and overall the amount left in
the air was somewhat over half the fuel input. The uncertainty in L is larger than the
mean estimate, apparently including the possibility that deforestation and soil opening
make land a source not a sink.
Fig. 16.7 Seasonally adjusted (to the annual means of seasonal data, Fig. 16.8) time-
series of O 2 /N 2 ratios relative to reference air at La Jolla, California, USA (33°N,
117°W). Units are millionths (meg).
(^) (After Manning & Keeling 2006.)
Fig. 16.8 Northern- and southern-hemisphere seasonal cycles of O 2 /N 2 ratios showing
the inverse seasonality of net production (photosynthesis > respiration) and net
metabolism.
(^) (After Keeling et al. 1998b.)