The Institute of Science in Society (ISIS) Report 19/08/09
Implications for Climate Change Policies
New research shows oxygen depletion in the atmosphere accelerating since
2003, coinciding with the biofuels boom; climate policies that focus exclusively
on carbon sequestration could be disastrous for all oxygen-breathing organisms
including humans Dr. Mae-Wan
Ho Threat of oxygen depletion
Mention climate change and everyone thinks
of CO2 increasing in the atmosphere, the greenhouse effect heating
the earth, glaciers melting, rising sea levels, floods, hurricanes, droughts,
and a host of other environmental catastrophes. Climate mitigating policies
are almost all aimed at reducing CO2, by whatever means.
Within the past several years, however, scientists have found that oxygen (O2)
in the atmosphere has been dropping, and at higher rates than just the amount
that goes into the increase of CO2 from burning fossil fuels, some
2 to 4-times as much, and accelerating since 2002-2003 [1-3]. Simultaneously,
oxygen levels in the world’s oceans have also been falling  (see Warming
Oceans Starved of Oxygen, SiS 44).
It is becoming clear that getting rid of CO2 is not enough; oxygen
has its own dynamic and the rapid decline in atmospheric O2 must
also be addressed. Although there is much more O2 than CO2
in the atmosphere - 20.95 percent or 209 460 ppm of O2 compared with
around 380 ppm of CO2 – humans, all mammals, birds, frogs, butterfly,
bees, and other air-breathing life-forms depend on this high level of oxygen
for their well being  Living with Oxygen (SiS
43). In humans, failure of oxygen energy metabolism is the single most important
risk factor for chronic diseases including cancer and death. ‘Oxygen deficiency’
is currently set at 19.5 percent in enclosed spaces for health and safety ,
below that, fainting and death may result.
simultaneous decrease in ocean oxygen not only threatens the survival of aerobic
marine organisms, but is symptomatic of the slow-down in the ocean’s thermohaline
‘conveyor belt’ circulation system that transports heat from the tropics to
the poles, overturns surface layers of into the deep and vice versa, redistributing nutrients and
gases for the ocean biosphere, and regulating rainfall and temperatures on
the landmasses. This dynamical system is highly nonlinear, and small changes
could make it fail altogether, with disastrous runaway effects on the climate
 (Global Warming & then
the Big Freeze, SiS 20). More importantly, it could wipe
out the ocean’s phytoplankton that’s ultimately responsible for splitting
water to regenerate oxygen for the entire biosphere, on land and in the sea
Measuring O2 to better understand the earth’s carbon budget
CO2 records go back more than 50 years , but O2 measurement
in combination with CO2 goes back barely two decades , and is
already giving important information on the size of the carbon sink in the
ocean relative to the land. For one thing, O2
have very different solubility in seawater; while 99
percent of the O2
remains in the atmosphere, 98 percent of the CO2
is in seawater.
are exchanged in different processes on land,
each having a different O2:CO2molar exchange ratio and thus distinguishable from
one another. Fossil fuel combustion has a global average O2:CO2exchange ratio of about 1.4 moles of O2consumed per mole of CO2produced,
whereas land plant photosynthesis generates an average net ratio of about
1.1 moles of O2 for
These ratios can vary over spatial and temporal scales, depending on whether
photosynthesis produces more oxygen than is consumed by respiration, and on
the precise fossil fuel burnt (see later). The linkage between CO2
and O2 is broken, however, at the
air-sea interface, where substantial O2
fluxes may be unaccompanied by fluxes of CO2
and vice versa.
knowing the fossil fuel emissions and the exact value of the exchange ratios,
one can separate the total CO2 uptake into land and ocean components
on timescales of a few years. This has led to estimates of the land and ocean
sinks to be 0.5 and 2.2 GtC/year respectively for the periods 1993 to 2003,
with a fossil fuel increase rate of 6.5 GtC/y . (see Box 1 for details
on the calculations.)
new atmospheric station has been established on the North Sea oil and gas
production platform F3, 200 km north of the Dutch coast, which measures both
CO2 and O2 continuously using the latest fuel cell and
infrared technologies , and more precise data on sea-air exchanges are
Measuring O 2 and calculating
land and ocean carbon sinks
is difficult to measure changes in O2
because there is so much of it in the atmosphere compared with CO2.
So a proxy is used instead. Changes are measured as differences in O2/N2
ratios expressed in “per meg” units against the ratio
in a standard mixture kept at the Scripps Institute of Oceanography, La
Jolla California in the USA, which pioneered the measurement.
per meg = 106[(O2/N2)sam - (O2/N2)]ref/
difference is used to define O2 concentration:
4.8 per meg are equivalent to 1 ppm (i.e., 1 mmole
O2 per mole of dry air). By making
the assumption that atmospheric N2
concentrations are constant, this definition of O2
concentration can be applied to derive O2
fluxes as follows [9, 10].
An Atmospheric Potential Oxygen (APO) is defined, also in per meg units,
as the sum of oxygen as determined n eq. (1) and the oxygen that went into
producing the CO2 in
APO per meg = D(O2/N2)
+ aB 4.8[CO2]
exchange ratio for land photosynthesis and respiration; and [CO2] is
the concentration of CO2
in the atmosphere. This assumes that variations in APO can only be caused
by air-sea exchanges of O2,
and by combustion of fossil fuels.
uptake of atmospheric CO2,
however, reduces the observed upward trend in atmospheric CO2
concentrations, but has no effect on the observed downward trend in O2/N2
ratios. Thus the global budgets for atmospheric CO2and O2can
be respectively represented by eqs (3) and (4).
= F – O – B
is the globally averaged observed change in atmospheric CO2
is the globally averaged observed change in atmospheric concentration, F
is the source of CO2 emitted
from fossil fuel combustion (and cement manufacture), O is the oceanic CO2 sink, B is the net land biotic
(including biomass burning, landuse change and land biotic uptake); aF
are the global average O2:CO2
exchange ratios for fossil fuels and land biota respectively, and Z is the
net exchange of atmospheric O2 with
the ocean. All except exchange ratios are in units of moles per year.
eqs 2, 3 and 4 gives:
DAPO = (-aF
+ Z (5)
DAPO is the globally averaged
observed change in APO. Eq. 5 is used to obtain the oceanic sink, and then
eq 3 is used to obtain the land biotic sink. This gives less uncertainty,
as APO is less variable than the O2/N2 ratios.
Large decreases in atmospheric oxygen detected
Decrease in atmospheric O2
has been detected in stations around the
world for the past decade, a consistent downward trend that has accelerated
in recent years.
The largest fall in O2
was observed in the study of Swiss research team led by Francesco Valentino
at University of Bern, for data collected at high altitude research stations
in Switzerland and France. The Jungfraujoch (JFJ) station in Switzerland (3
580 m above sea level, 46o 33’N, 7o 50’E) is located
on a mountain crest on the northern edge of the Swiss Alps. The Puy de Dôme
station (1 480 m above sea level, 45o46’N, 2o 58’E)
is situated west of the Alps at the summit of Puy de Dôme.
research team confirmed the general upward trend for atmospheric CO2
and a downward trend in atmospheric O2.
But since 2003 for JFJ, and mid 2002 for at Puy, there is a significant enhancement
of O2 and
compared to previous years. At JFJ, the rate of CO2
increase shifted up from 1.08 ppm (parts per million) for the years 2001-2002
to 2.41 ppm/y for 2003-2006; while the increase in D(O2/N2)
and APO (measures of oxygen concentration, see Box 1) shifted downwards to
greater extents from –2.4 ppm/y and -1.5 ppm/y to -9.5 ppm/y and -6.9 ppm/y
For Puy, CO2 increase changed from 2.43 ppm/y for 2001-2002 to 1.07
ppm/y for 2003-2004, followed by 2.4 ppm/y for the years 2005-2006; while the
changes in D(O2/N2) and APO
were -6.1 ppm/y and -3.7 ppm/y for 2001-2002, to -10.4 ppm/y and -7.6 ppm/y
for the years 2002-2006. Averaged over all years – by removing the trends
and plotting correlations between CO2 and O2, an O2:CO2exchange
ratio of -1.9+0.7 is found for JFJ, and -1.8+0.5 for Puy; both
significantly different from the 1.1 assumed for land photosynthesis and respiration
i.e., 1.1 mole of O2generated per mole of CO2
fixed, and -1.4 for burning fossil fuels, or 1.4 mole of O2used
up when one mole of CO2 is produced.
Over time, the O2:CO2exchange ratio for JFJ, which is much less exposed
to local or regional anthropogenic influence because of its elevation and
location, was -2.1+0.1 for the years 2001-2002 and -4.1+0.1
for the years 2003-2006. At Puy, the ratio was -4.2+0.1 for the period
2001-2003, and -7.3+0.1 for 2003-2006. These ratios are completely
out of line with what could be expected from fossil fuels, and other data
indicate that there has been no significant change in fossil fuel emission
rates during the period 2003-2006.
The researchers speculated that the large
decrease in atmospheric oxygen since 2003 could have been the result of oxygen
being taken up by the ocean, either due to a cooling of water in the North
Atlantic, or water moving northwards from the tropic cooling, both of which
would increase the water’s ability to take up more oxygen. However, it would
require unrealistic cooling to account for the change in O2concentration. And all the indications are that the
ocean waters have warmed since records began .
In a second study, atmospheric O2 and CO2 data collected
from two European coastal stations between 2000 and 2005 were analyzed .
Mace Head Ireland (53o20’N 9o54’W, 35 m above sea level),
which serves as the marine background, relatively free from local fossil fuel
consumption, and Station Lutjewad (53o24’N, 6o21’E) on
the northern coast of The Netherlands 30 km to the northwest of the city of
Groningen, which serves as a continental station receiving continental air with
northerly winds. Similar trends were detected. Over the entire period at Lutjewad,
CO2 increased by 1.7+0.2 ppm/y while oxygen decreased at -4.2+0.3
ppm/y; the corresponding figures for Mace Head were 1.7+0.1 ppm/y and -4.0+0.3
ppm/y. O2 is decreasing faster than can be accounted for by the rise
in CO2. Furthermore, the decrease is not uniform throughout the entire
period; instead it is much steeper between 2002 and 2005 at both stations, and
is not accompanied by any change in the trend of CO2 increase. This
sharp acceleration in the downward trend of atmospheric O2 from 2002-2003
onwards in Ireland and The Netherlands is in accord with the findings in Switzerland
and France . And this cannot be explained by a realistic increase in fossil
fuel use, or oxygen uptake by cooler ocean waters; if anything, oxygen level
in the oceans has also been falling . So where and what is this oxygen sink
that is soaking up oxygen?
Mystery of the oxygen sink
One distinct possibility that has been considered is
that an extra oxygen sink has opened up on land as the result of human activities.
James Randerson at University of California
Irvine was lead author on a report published in 2006  pointing out that
a decrease in atmospheric O2
could result if carbon within the land biosphere becomes more oxidized (sequestering
more oxygen) through disturbance of natural ecosystems. This has changed the
natural land cover, replacing it with plants that effectively remove more
oxygen from the atmosphere.
Atmospheric exchange of O2 with
land ecosystems is commonly expressed in terms of a net carbon flux from the
atmosphere to the ecosystem (Fnet) and the net O2:CO2exchange ratio (Rnet):
dO2/dt = - Rnet Fnet
By convention, positive sign indicates release into
the atmosphere and negative sign sequestered in the land biosphere. The net
rate is really a difference between two processes, one moving from atmosphere
to biosphere, and the other in reverse, from biosphere to atmosphere, so eq.
(6) can be written as follows.
dO2/dt = - (Rab Fab + RbaFba)
where Fab is the atmosphere to biosphere
carbon flux (the same as net primary productivity, NPP), Rab is
the oxidative ratio related to NPP (moles O2 released per mole
CO2 fixed), Fba is the biosphere to atmosphere return
flux (a combination of respiration, fires and other losses), and Rba
is the oxidative ratio related to the return flux (moles O2 consumed
per mole CO2 released).
In an ecosystem at steady state
(in dynamic balance), Fab and Fba will have the same
magnitude. But the carbon in Fba is always offset in time from
newly assimilated carbon in Fab because of carbon storage in the plant, dependent on plant
tissue lifetimes, rates of litter and soil organic matter decomposition, and
so on. Changes in Rab and Rba have the potential to
cause relatively large changes in atmospheric O2, basically because
of the time delays between fixation and the return flux due to carbon storage.
The longer the carbon storage (turnover) time, the larger the effective offset
between Fab and Fba; so O2 is consumed at
a slower rate, and more of it remains in the atmosphere
Randerson and colleagues hypothesize that
increasing levels of disturbance across natural ecosystems in recent decades
has decreased Rab. This includes wide-spread deforestation and
replacement of woody vegetation with pastures and crops in the tropics, an
increase in fire activity and tree mortality and increasing the abundance
of deciduous tree species and herbaceous plants in the boreal (northern) regions.
Globally, this includes an increase in invasive species and increased disturbance
of agricultural soils by plowing and grazing during the 20th century.
All these activities increase the oxidation state of carbon in plant and soil
organic matter. The increases in oxygen content of the resultant biomass causes
a small sink for atmospheric O2 that has not been accounted for
in atmospheric budgets.
Within a plant, lipids and lignin compounds
have carbon that is more reduced,
i.e., with more hydrogen and less oxygen; they have and large R values of
1.37 and 1.14 respectively, and are energetically more costly to build than
compounds such as cellulose and starch, which have less hydrogen, more oxygen,
and R value of 1.0. Thus, the expansion of agriculture and grazing during
the 20th century has probably caused a decrease in the oxidative
ratio of the plant biomass within these disturbed ecosystems. Using several
simple models, the researchers showed that, indeed, small changes in Rab
could lead to substantial decreases in atmospheric O2.
Another research team has raised the possibility that reactive nitrogen produced in
making artificial fertilizers for agriculture could also be tying up more
oxygen in plant tissue, soil organic matter and oceans in the form of nitrates
The importance of oxygen accounting in climate policies
Change in land use, and increased oxidation of nitrogen
could explain the long term steady decline in atmospheric O2, and
may well also account for the sharp acceleration of the downward trend since
2002 and 2003.
These years happen to coincide
with record rates of deforestation. In Brazil, 10 000 square miles were lost
mainly to pasture land, soybean plantations and illegal logging, a 40 percent
rise over the previous year . Massive deforestation has continued in the
Amazon and elsewhere, spurred by the biofuels boom ; it is estimated that
nearly 40 000 ha of the world’s forests are vanishing every day.
The crucial role of forests
and phytoplankton  in oxygenating the earth shows how urgent it is to take
oxygen accounting seriously in climate policies. Reductionist accounting for
CO2 alone is insufficient, and even grossly misleading and dangerous.
A case in point is the proposal of the International Biochar Initiative (IBI).
‘Biochar’ is charcoal produced to be buried in the soil that IBI has been promoting
worldwide over the past several years  as a means of sequestering carbon
from the atmosphere to save the climate and enhance soil fertility. It involves
planting fast growing tree and various other crops on hundreds of millions of
hectares of ‘spare land’ mostly in developing countries, to be harvested and
turned into charcoal in a process that could produce crude oil and gases as
low grade fuels. There are many excellent arguments against this initiative
, but the most decisive is that it will certainly further accelerate deforestation
and destruction of other natural ecosystems (identified as ‘spare land’). In
the process, it could precipitate an oxygen crisis from which we would never
recover  (Beware
the Biochar Initiative, SiS 44).
Valentino FL, Leuenberger
M, Uglietti C and Staburm P. Measurements and trend analysis of O2,
CO2 and D13C
of CO2 from high altitude research station Junfgraujoch, Switzerlnd
– a comparison with the observations from the remote site Puy de Dôme, France.
Science of the Total Environment 2008, 203-10.
Sirignano C, Neubert REM,
Jeijer HAJ and Rodenbeck C. Atmospheric oxygen and carbon dioxide observations
from two European coastal stations 2000-2005: continental influence trend
changes and APO climatology. Atmos Chem Phy Discuss 2008, 8, 20113-54.
Tohjima Y, Muai H, Machida
T, Nojiri Y. Gas-chromatographic measurements of the atmospheric oxygen/nitrogen
ratio at haterumna island and Cape Ochi-ishi, Japan. Geophys Res Lett 2003, 30, 1653, doi:10.1029/2003FLO17282
Joos F. Trends
in Marine Dissolved Oxygen: Implications for Ocean Circulation Changes and
the Carbon Budget. EOS 2003,
AC, Keelilng RF, Paplawsky WJ, Katz LE, McEvoy EM and Atwood CG. Atmospheric
oxygen in the 1990s from a global flask sampling network: trends and variability
pertaining to the carbon cycle. Draft 29 January 2003, http://bluemoon.ucsd.edu/publications/mip/manning.pdf
11. Battle M, Fletcher SM,
Bender ML, Keeling RF, et al. Atmospheric potential oxygen: new observations
and their implications for some atmospheric and oceanic models. Global Biogeochemical Cycles 2006, GB1010.
12. Randerson J T, Masiello
C A, Still C J, Rahn T, Poorter H and Field C B. Is carbon within the global
terrestrial biosphere becoming more oxidized? Implications for trends in atmospheric
O2, Glob Change Biol
2006. 12, 260–71.
13. Ciais P, Manning A C, Reichstein
M, Zaehle S, and Bopp L. Nitrification amplifies the decreasing trends of
atmospheric oxygen and implies a larger land carbon uptake, Global Biogeochem
Cy 2007, 21, GB2030, doi:10.1029/2006GB002799, 2007.
14. Rain Forest is losing
ground faster in Amazon, photos show”, Tony Smith, The New York Times, 27
June 2003, http://www.mongabay.com/external/record_amazon_deforestation_2002.htm#1
15. “Environment: Biofuels
boom spurring deforestation”, Stephen Leahy, IPS, 21 May 2007, http://ipsnews.net/news.asp?idnews=37035
16. IBI Programs and Projects,
International Biochar Initiative, accessed 3 August 2009, http://www.biochar-international.org/
17. Ernsting A and Rughani
D. Climate geo-engineering with ‘carbon negative’ bioenergy, climate saviour
or climate endgame? Biofuelwatch, November 2008, http://www.biofuelwatch.org.uk/docs/cnbe/cnbe.html
18. Ho MW. Beware
the Biochar Initiative. Science in Society 44 (to appear).