Impact of anthropogenic activities on global land oxygen flux

Atmospheric oxygen (O2) is one of the predominant features that enable earth as a habitable planet for active and diverse biology. However, observations since the late 1980s indicate that O2 content in the atmosphere is falling steadily at part-per-million level. Although a scientific consensus has emerged that the current decline is generally attributed to the combustion of fossil fuel, a quantitative assessment of the anthropogenic impact on the O2 cycle on both global and regional scale is currently lacking. This paper quantifies the anthropogenic and biological O2 flux over land and provides a 10 quantitative and dynamic description of land O2 budget under impacts of human activities on a grid scale. It is found that total anthropogenic O2 flux over land has risen from 35.6 Gt/yr in 2000 to 46.0 Gt/yr in 2013, while the production from land (11.5Gt/yr averaged from 2000 to 2013) displays a faint increase during the same period. High anthropogenic fluxes mainly occur in Eastern Asia, India, North America and Europe caused by fossil fuel combustion and in Central Africa caused by wildfire. Due to strong heterotrophic soil respiration under higher temperature conditions, the positive biological O2 flux in 15 the tropics is not significant. Instead, boreal forest and Tibetan plateau become the most important sources of atmospheric O2 in the Anthropocene. The anthropogenic oxygen consumption data are publicly available online at https://doi.org/10.1594/PANGAEA.899167.


Introduction
In recent decades, the strong response that global carbon cycle gives to anthropogenic forcing is addressed by voluminous literature on both global and regional scale (Huang et al., 2007;IPCC, 2013;Luyssaert et al., 2007), which is essential for understanding environmental history of our planet and predicting and guiding our future.However, the O2 cycle, another fundamental biogeochemical cycle in the earth system, is also responding forcefully to global change with its own dynamic.
The decline of O2 concentration, an average loss rate of 4 ppm per year, is comparable in magnitude to those of CO2.The essentiality of atmospheric O2 to the habitability of our planet Earth and the survival of humankind can never be overstated since an equable O2 in the atmosphere is central to life (Petsch, 2013).Therefore, systematic investigations on O2 cycle, especially under anthropogenic forcing deserves to be high on the agenda.Fossil fuel combustion which effectively causes an O2 loss has clearly not been identified as the only culprit responsible for the recent decline.Additional processes including respiration of human and livestock, wildfire, production from terrestrial and oceanic ecosystems also play a part (Huang et al., 2018).Most of the existing concepts of O2 budget are based on a static scenario, which assumes that atmospheric O2 concentration remains constant and each component of the budget is independent of time (Bender et al., 1994a(Bender et al., , 1994b)).Nevertheless, since the beginning of the Industrial Era, humans have been producing energy by combusting fossil fuels, a process that not only emits considerable amounts of CO2 into the atmosphere (IPCC, 2013), but also removes substantial amount of molecular O2 from the atmosphere, thus resulting in the disturbance of both carbon and O2 cycle comparable to the major natural flux in magnitude.Records from polar ice core indicate a decline of atmospheric O2 related to the burning of fossil fuel (Battle et al., 1996).Previous research has shown that approximately half of the anthropogenic CO2 since the Industrial Era has been absorbed by the ocean and terrestrial ecosystem (Qué ré et al., 2018), while the response of each component involving the O2 cycle to anthropogenic forcing is poorly discussed.Does more O2 transport from land and ocean to the atmosphere to compensate for the current depletion of O2? What is the contribution of land ecosystem to the supplement of atmospheric O2? How will the contribution variate in the future?These questions remain to be answered systematically.
Additionally, the descriptions of O2 budgets are generally provided in a global mean, only roughly representing the globally averaged value of each process.However, the amount of O2 consumed and produced is not uniform over land and could vary dramatically from place to place due to different levels of human activity (burned fuel type, population density etc.) and different natural conditions (vegetation cover, growing season etc.).Therefore, in some regions, the O2 consumption may far exceed the production, which requires the supplementary O2 from regions where production of O2 surpasses the consumption to maintain the local O2 level through atmospheric circulation.Where does the supplementary O2 come from and how much O2 can it provide?Will the O2 supply from these areas increase or decrease in the future?To address those interesting problems, a dynamic global O2 budget on a grid scale is necessary.
In this paper, most major types of anthropogenic O2 consumption processes are estimated globally on each grid and are combined with the estimation of terrestrial biological O2 flux to provide a dynamic O2 budget on a grid scale that varies with space and time.The dynamic O2 budget can provide us with an insightful perspective on global climate change.

The observational oxygen concentration data
In this study, O2 concentrations of 9 stations from the Scripps O2 Program (http://scrippso2.ucsd.edu/)are used.Since these data from the 9 stations are collected from distant locations, they are able to represent the average concentrations over a wide range of areas rather than local background values (Keeling, 2018).
The concentration of atmospheric O2 is expressed as changes in the O2/N2 ratio of air relative to a reference (air collected in the mid-1980s).The observed changes are so tiny that they are reported in per meg units (Manning and Keeling, 2006): where the subscripts 'sample' and 'reference' denote the sample air and the reference air collected in the mid-1980s, respectively.1 per meg equals 10 -6 .The conversion between volume fraction (ppm and per meg) and mass (Gt O2) can be moles is a reference value for the total number of O2 molecules in the atmosphere.

The estimation of anthropogenic oxygen flux
In this paper, the following four processes including fossil fuel combustion, human respiration, livestock respiration and wildfire are discussed.Other processes are tiny enough to be neglected, compared with the processes mentioned above, like the oxidative weathering process (Haynes et al., 2016).In terms of the wildfire consumption, some fire activities are caused naturally, while others such as agriculture waste burning, deforestation are human-induced.For convenience, we include wildfires in the calculation of anthropogenic fluxes and exclude it from the biological O2 fluxes by terrestrial vegetations.

Oxygen flux caused by consumption by fossil fuel combustion
The estimation of O2 flux by fossil fuel combustion is converted from CO2 emissions data from the Carbon Dioxide Information Analysis Center (CDIAC, http://cdiac.ess-dive.lbl.gov/).These data provide time series of annual CO2 emissions from anthropogenic sources, including fossil fuel burning, cement manufacturing and gas flaring in oil fields as well as energy production, consumption and trade data on a 1°×1° grid from 1751 to 2013.
The O2 consumption in the combustion process is estimated according to the equation as follow: Due to the different fuel mix in each country, the oxidative ratio (the number of O2 moles that are consumed per mole CO2 emitted) can vary over spatial and temporal scales (Steinbach et al., 2011).In this paper, the carbon emission by different types of fossil fuel burned in each country is also obtained from the CDIAC.Considering there are several historical events (collapse of the Soviet Union in 1991, independence of South Sudan in 2010, etc.) that change the national boundaries, we adopt CShapes (Weidmann et al., 2010), a new dataset that provides historical maps of country boundaries in the post-World War II period to map the gridded oxidative ratio on country level.Then, O2 consumption by fossil fuel on each is estimated based on the following equation: (2) where   is the annual O2 consumption (Gt O2/yr);   2 is the relative molecular mass of O2 (32g/mol); subscript i is used to indicate the type of fuel;    is the i-th species carbon emissions from fossil fuel combustion (Gt C/yr);   is the relative molecular mass of carbon (12 g/mol);   is the molar ratio of O2:CO2 when the i-th fuel is completely burned (

Oxygen flux caused by human respiration
The O2 flux by human respiration is based on the population density datasets at 0.5° × 0.5° resolution from 1980 to 2100 from Murakami and Yamagata (2016).The data can be obtained in http://www.cger.nies.go.jp/gcp/population-and-gdp.html.
The datasets also provided actual GDPs from 1980 to 2010, and estimated GDPs under different SSPs from 2010 to 2100.
We assume that in a day a man works 8 hours with oxygen consumption rate at 1.0 L O2/min and rests for the remaining 16 hours with oxygen consumption rate at 21.0 L O2/h.Then, an adult consumes approximately 1.17 kg (816 L) of O2 per day.
It is estimated that an astronaut consumes 0.84kg O2/day for survival (Jones, 2003).Our estimation (1.17kg) is relatively higher, which is reasonable because the astronauts have fewer physical activities than people on earth.We estimate the total respiration consumption based on the following equation: where CRES-H is the annual O2 consumption of human respiration (unit: Gt O2/yr), P is the total population, and Cd is the daily O2 consumption per capita (kg O2/day).

Oxygen flux caused by livestock respiration
O2 consumption by livestock respiration is based on the global datasets on the geographic distribution of livestock from Gridded Livestock of the World v3.0 (GLW 3) (Gilbert et al., 2018).GLW 3 provides global population densities of cattle, buffaloes, horses, sheep, goats, pigs, chickens and ducks at a spatial resolution of 0.083 degrees (approximately 10km at the equator).The dataset contains 2 versions of density pattern.In the first version of dasymetric weighting, livestock numbers are distributed within census polygons according to weights established by statistical models using high-resolution spatial covariates.In the second version of areal weighting, animal numbers are distributed homogeneously with equal densities within their census polygons.Here, the second version of data is used.
The basal metabolic rate (BMR) is the rate of energy expenditure per unit time by endothermic animals at rest.It can be reported in mL O2/min.The BMR (mL O2/h) of a mammal can be predicted by the formula given by Kleiber (1932).BMR = 3.43M 0.75 , where M is the animal's mass (g).Then, following the formula below, the annual O2 consumption of the livestock can be estimated.
where CRES-L is the annual O2 consumption of livestock (unit: Gt O2/yr), Pi is the total number of i-type livestock, and    is the average daily O2 consumption of the i-type livestock (unit: kg O2/day).
Since the data only describes the situation in 2010, we assume that the total number of all livestock is proportional to the total human population and use the spatial pattern of the data when estimating the O2 consumption in other years.

Oxygen flux caused by wildfires
The data on O2 consumption by fire are estimated based on carbon emissions data derived from the Global Fire Emissions Database (GFED) (Van Der Werf et al., 2017).Satellite information on fire activity and vegetation productivities around the world are combined together to estimate gridded burned area and pollutant emissions.The version of the dataset used in this paper is version 4, with a spatial resolution of 0.25 degrees from 1997 to 2016.The GFED data classifies fire types into the following categories: savanna, boreal forest, temperate forest, tropical forest, peat and agricultural waste.Since the combustion products are all organic, the molar ratio of O2: CO2 is 1.1:1.The O2 consumption can be estimated by the following formula: where CFIRE is the O2 consumption (Gt O2/yr); subscript i is used to indicate the type of fire;   is the mass of the i-th fire type dry matter emitted (kg DM/yr); CCi is the percentage of carbon in the i-type fire dry matter; EFi is the emission factor (kg/g) of the i-th fire type CO2; 1.1 is the molar ratio of O2:CO2 at the time of complete combustion.

Estimation of O2 consumption by fossil fuel combustion
It's widely recognized that O2 consumption by fossil fuel is the most important attribution to the recent O2 decline (Keeling and Manning, 2014;Martin et al., 2017;Valentino et al., 2008).Figure 1 shows the global pattern of O2 consumption by fossil fuel combustion and oxidative ratio in 2013.It is found that the patterns in O2 consumption (Fig. 1a) generally follow the pattern of CO2 emission.High O2 consumption areas (greater than 5.0 kg O2/m 2 /yr) are in Eastern Asia, India, Europe and the US, while low O2 consumption can be seen in South America, Africa and Australia.In the oxidative ratio map (Fig. 1b), low oxidative ratio, signifying coal as the major fuel source, are located in Eastern Asia, India and South Africa.High oxidative ratio (greater than 1.45) areas are found in Russia, Central Asia, Canada and Argentina, indicating gas as the main source of fossil fuel burning.China, United States, India and Russia are the 4 countries that consume the most O2 by fossil fuel combustion.Solid fuel, which accounts for about 70% of the total carbon consumption, is the biggest source of anthropogenic CO2 emission as well as the biggest anthropogenic sink of atmospheric O2 in China and India.In Russia, the most important energy source is gas fuel, followed by solid fuel and liquid fuel, which leads to the high oxidative ratio (1.61).Figure 3a shows the global trend of O2 consumption on a gridded level, in which the warm colors indicate an increase and cold colors a decrease in O2 consumption on each grid.The increase mainly occurs in East Asia and India, while the decrease is located in Europe.Figure 3b shows the trend of the oxidative ratio.With the increasing demand for solid fuels, the oxidative ratio displays a downward trend globally, except for regions such as Russia, Europe and Argentina.

Estimation of O2 consumption by human and livestock respiration
When estimating the O2 consumed by human respiration, we assume that a man works about 8 hours a day and rest for the remaining 16 hours, and ignore the human activities which consume more O2 such as sports and hard physical labor.As for the livestock, eight types of main livestock, including goats, ducks, buffaloes, sheep, horses, cattle, pigs and chickens are considered.The estimations are based on the basal metabolic rate, which neglects the metabolic enhancement caused by ambient temperature variation, exercise, etc (Cai et al., 2018).Therefore, the actual O2 consumption by human and livestock would be greater than the estimation in this study.
Figure 4 shows the global pattern and long-term trends of O2 consumed via human respiration.The highest O2 consumption occurs in India and eastern China, up to 0.7 kg O2/m 2 /yr, which are the most populated regions around the world (Figure 4a).
The population has grown worldwide, especially in India and East Asia.The global distribution of O2 consumption (Figure 5a) follows a similar pattern because the livestock industry should be well-developed to cope with food demand from high population density.Among the 5 continents (Asia and Australia are considered as one continent in this paper because Australia is relevantly small in both area and O2 consumption) illustrated in Figure 5, buffaloes, cattle, chickens, ducks, goats, pigs and sheep in Asia and Australia consumes the most O2, while horses in South America account for the most O2 consumed by livestock.
In terms of the global total volume, the O2 consumed by livestock and human respiration is roughly equivalent (Figure 6), about 3.0 Gt in 2013 and the total O2 consumed by human and livestock has increased from 3.4 Gt to 6.0 Gt during 1975-2013.Among the 8 types of the main livestock around the world, cattle consume the most O2, accounting for 55% of total O2 consumed by livestock.

Estimation of O2 consumption by wildfire
Figure 7 shows the global distribution of O2 consumption by wildfire from 1997 to 2016.The highest consumption by wildfire is mainly located in the tropical region, especially central Africa since these areas are rich in surface vegetation and the net primary productivity (NPP) is relevantly high.The burning of these areas would emit more carbon and thus consume more O2.
Earth Syst.Sci.Data Discuss., https://doi.org/10.5194/essd-2019-36 Open Global mean O2 consumption is 5.87 Gt O2/yr and shows a weak downtrend during 1997-2016, with a maximum of 8.1 Gt O2/yr in1997 and minimum of 4.8 Gt O2/yr in 2013.In terms of O2 consumption of various types of fires, savanna fires have the widest distribution and consume the most O2, accounting for 60%~70% of the total; followed by fires in tropical forests.

Global land net O2 flux
The As for the air-sea O2 flux, the possibility that the ocean might be a long-term source of O2 in the atmosphere has already been widely recognized.Human-induced global warming and climate change reduce the solubility of the ocean (Bopp et al., 2002;Plattner et al., 2002) and thus causes the decline in the dissolved O2 in the upper ocean and O2 is released from the ocean to the atmosphere.The process above is believed to be superimposed on a natural background air-sea O2 fluxes of different time scales.It is estimated that about 1.4Gt O2 is outgassed from ocean per year during 2000~2010 (Keeling and Manning, 2014), which is very small compared to the magnitude of the other processes we estimated earlier, and it is difficult to provide a spatial distribution because of sparse coverage of measurements on the ocean (Keeling et al., 2010).
Since the intensity of human activities and vegetation coverage can vary over spatial and temporal scales, the local O2 budget in different parts of the world can be quite different.Figure 10 shows the total land O2 flux in tropical, northern temperate, northern boreal and southern temperate regions.Except for the northern boreal region, the anthropogenic O2 consumption in other areas is greater than the biological O2 flux that transport O2 to the atmosphere.In the tropics, the biological O2 flux is not high, which may be caused by strong heterotrophic soil respiration in relevantly higher temperature.In terms of the anthropogenic flux, wildfires induced O2 flux is comparable to consumption of fossil fuel combustion, but they follow opposite trends over time: the fossil fuel O2 flux experienced a steadily growing trend while the wildfire flux indicates a downward trend, which may be related to the reduction in wildfire activities in recent decades (Arora and Melton, 2018).In the temperate regions of the northern hemisphere, which is the most populated region in the world, total anthropogenic O2 flux account for more than half of the global anthropogenic O2 flux, reaching 23.5Gt/yr, and its growth rate is also the most significant among the 5 regions.In the boreal region of the northern hemisphere, wildfire activities are frequent and consume more O2 than fossil fuel combustion.Since this region is sparsely populated, the land terrestrial ecosystem is playing a leading role in the local O2 budget and thus become the source of O2 as well as a sink of CO2 in the background of global change.In the temperate of the southern hemisphere, livestock respiration consumes more O2 than human respiration and the consumption of O2 by wildfire also surpasses the consumption by fossil fuel due to the relevantly weak intensity of human activities.As the earth's third pole, the Tibetan plateau is one of the most important areas of global weather and climate change (Ma et al., 2017).Figure 10e shows the O2 budget over the Tibetan Plateau.The biological fluxes that produce O2 to the atmosphere display no significant trend while the O2 consumed by fossil fuel combustion continued to rise during the years.The Tibetan plateau is another source of atmospheric O2 under the impact of human activities.

Data availability
The anthropogenic oxygen consumption dataset from 1975 to 2013 is available on PANGAEA at https://doi.org/10.1594/PANGAEA.899167 as NetCDF files (.nc) with 1.0°×1.0°spatial resolution (Liu et al., 2019).As for the consumption by wildfire, the high O2 consumption area is distributed in the low latitude tropics.These areas are rich in surface vegetation and have a high net primary production.The burning of vegetation will emit more carbon and thus consume more O2.Savanna fires have the widest distribution and consume the most O2, accounting for 60%~70% of the total; followed by tropical rain forests.Since the beginning of the 21st century, the total amount of O2 consumption caused by the combustion process has been slowly decreasing.
When anthropogenic flux, including fossil fuel combustion, respiration of human and livestock, and wildfire are combined, several significant high O2 consumption area like East Asia, India, eastern North America, Europe and central Africa are observed.Except for central Africa, these areas are densely populated with high levels of human activities and thus consume more O2 than other areas of the world, while in central Africa, where the wildfire accounts for the vast majority of local O2 consumption, also exhibits a high level of anthropogenic O2 flux.

Figure 2
Figure2shows the long-term trends of global total O2 consumption by fossil fuel combustion and global averaged oxidative ratio from 1975 to 2013.Global total O2 consumption has increased from 17.0 Gt to 35.0 Gt during 1975~2013.The growth could be explained by the rise of solid fuel to a large extent.The oxidative ratio has experienced a significant decrease after the 21 st century and reached 1.34 in 2013, which can be explained by a robust increase in the coal burning as well as the cement production in countries such as China and India.This oxidative ratio calculated in this paper is close to the review for journal Earth Syst.Sci.Data Discussion started: 10 April 2019 c Author(s) 2019.CC BY 4.0 License.
global distribution of land O2 flux from 2000 to 2013 is illustrated in figure8a.In figure8a, total anthropogenic flux, including fossil fuel combustion, respiration of human and livestock, and wildfire are drawn.The distribution of total O2 consumption generally follows the pattern of O2 consumption by fossil fuel combustion in most part of the world, since, among the four O2 consumption processes, combustion of fossil fuels consumes the most O2 and are mainly distributed in areas with a high density of livestock and population.There are several significantly high O2 consumption area around the world: East Asia, India, eastern North America, Europe and central Africa.Except for central Africa, these areas are densely populated with high levels of human activities and thus consume more O2 than other areas of the world, while in central Africa, where the wildfire accounts for the vast majority of local O2 consumption, also exhibits a high level of anthropogenic O2 flux.The net biological O2 flux over land are estimated based on NEE (Net Ecosystem Exchange) simulated by the CASA model, with fire activities excluded.Fire activities are considered as the anthropogenic fluxes in this paper.Negative flux (brown regions) indicate places where uptake of O2 from the atmosphere occurs.Positive flux (green colors) indicate places where the production of O2 occurs.Globally, most regions show positive flux, which means the terrestrial ecosystem is transportingO2 to the atmosphere in most parts of the world.It is worth noting that in some tropical regions such as Southeast Asia and Amazon forest, the biological O2 flux is negative, but the total O2 flux over tropical land areas (including South American Tropical, Tropical Asia and North Africa) shows positive flux of 2.1Gt/yr.In Northern Temperate (Temperate America and Temperate Eurasia), total biosphere flux amounts to 3.0Gt/yr (Figure10).The spatial distribution of the difference between anthropogenic O2 flux and biological O2 flux, which is defined as the land net O2 fluxes, averaged from 2000 to 2013 is shown in figure8c.The brown regions, mainly located in eastern Asia, Europe, North America, and northern South America, covering more than half of the land, denote the areas where anthropogenic O2 flux exceeds the biological O2 flux.In these areas, humans are consuming more O2 than the ecosystem can provide.The area shaded with green color, including northern Canada and Siberia, represents the region that can still emit O2 into the atmosphere under the impact of human activities.However, it should be noted that since the positive O2 flux is comparably smaller, the color bar, which exhibits larger negative values and smaller positive values, has been modified to enhance visualization, so that readers can observe the positive O2 flux from the land more closely.Due to the increasing fossil fuel combustion, overgrazing and population growth, current O2 consumption over land is far greater than O2 production from the terrestrial ecosystem, breaking the atmospheric O2 balance and causing the decline of O2 concentration in the atmosphere.During the period of 2000~2013, it is estimated that total anthropogenic O2 flux over land Earth Syst.Sci.Data Discuss., https://doi.org/10.5194/essd-2019forjournal Earth Syst.Sci.Data Discussion started: 10 April 2019 c Author(s) 2019.CC BY 4.0 License.has risen from 35.6 Gt/yr in 2000 to 46.0 Gt/yr in 2013, while the production from land (11.5Gt/yr averaged from 2000 to 2013) displays a faint increase during the same period (Figure 9).
Earth Syst.Sci.Data Discuss., https://doi.org/10.5194/essd-2019forjournal Earth Syst.Sci.Data Discussion started: 10 April 2019 c Author(s) 2019.CC BY 4.0 License.5 Conclusions and DiscussionsWe have presented a global gridded dataset of O2 consumption by anthropogenic processes, including fossil fuel burning, wildfire and respiration of human and livestock based on fossil fuel carbon emission from CDIAC, carbon emission from GFED and population and livestock density data.Combining this data with biological O2 flux over land, the global O2 budget on land is estimated from 2000 to 2013 on a 1°×1° grid.The O2 consumption is converted from CO2 emission via the estimation of the oxidative ratio of different countries.Low oxidative ratios, signifying coal as the major fuel source, are located in eastern Asia, India and Southern Africa, while a high oxidative ratio (greater than 1.45) areas are found in Russia, Central Asia Canada and Argentina, indicating gas fuel as the main source of fossil fuel burning.The oxidative ratio shows a downward trend from 2000, possibly due to the growing contribution from coal burning with low oxidative ratio and cement production that does not consume O2.High consumption areas are mainly located in eastern Asia, India, North America and Europe.In terms of total volume, global total O2 consumption by fossil fuel has increased from 17.0 Gt to 35.0 Gt during 1975~2013.The O2 consumed by livestock and human respiration is roughly equivalent, about 3.0 Gt in 2013 and the total O2 consumed by human and livestock has increased from 3.4 Gt to 6.0 Gt during 1975-2013.Cattle consume the most O2 among the 8 types of the main livestock calculated, accounting for 55% of total O2 consumed by livestock.
from the terrestrial ecosystem, breaking the atmospheric O2 balance and causing the decline of O2 concentration in the atmosphere.During the period of 2000~2013, total anthropogenic O2 flux over land has raised from 35.6 Gt/yr in 2000 to 46.0 Gt/yr in 2013, while the production from land (11.5Gt/yr averaged from 2000 to 2013) displays a faint increase during the same period.The regions where anthropogenic O2 flux exceeds the biological O2 flux and humans are consuming more O2 than the ecosystem can provide are distributed in eastern Asia, Europe, North America, and northern South America, covering more than half of the land.Earth Syst.Sci.Data Discuss., https://doi.org/10.5194/essd-2019forjournal Earth Syst.Sci.Data Discussion started: 10 April 2019 c Author(s) 2019.CC BY 4.0 License.

Figure 1 :
Figure 1: Global distribution of O2 consumption by fossil fuel combustion and oxidative ratio from in 2013.O2 consumption by 10 review for journal Earth Syst.Sci.Data Discussion started: 10 April 2019 c Author(s) 2019.CC BY 4.0 License.

Figure 2 :
Figure 2: Global total O2 consumption by fossil fuel combustion and oxidative ratio from 1975 to 2013.Long-term trends of global total O2 consumption by fossil fuel combustion (a) and global averaged oxidative ratio (b) from 1975 to 2013.
review for journal Earth Syst.Sci.Data Discussion started: 10 April 2019 c Author(s) 2019.CC BY 4.0 License.

Figure 3 :
Figure 3: Global patterns of the long-term trend of O2 consumption by fossil fuel combustion and oxidative ratio from 1975 to 2013.Global pattern of long-term trends of global total O2 consumption by fossil fuel combustion (a) and Global pattern of longterm trends of the global oxidative ratio (b) from 1975 to 2013.
review for journal Earth Syst.Sci.Data Discussion started: 10 April 2019 c Author(s) 2019.CC BY 4.0 License.

Figure 4 :
Figure 4: Global pattern of O2 consumption by human respiration and its long-term trends.O2 consumption by human (a) for the year of 2013 and long-term trends (b) from 1975 to 2013.

Figure 5 :
Figure 5: Global pattern of O2 consumption by livestock respiration.O2 consumption by livestock (a) for the year of 2013 (b) Bar chart of O2 consumed by 8 types of main livestock in each continent (Unit: Gt/yr).

Figure 6 :
Figure 6: Global total O2 consumption by human and livestock respiration from 1975 to 2010 (Unit: Gt/a).The blue bar denotes the consumption by human and the orange bar denotes the consumption by livestock.

Figure 7 :
Figure 7: Global O2 consumption wildfire from 1997 to 2016.(a) O2 consumption by human wildfire for the year 2013.(b) Longterm trends of global total O2 consumption by different fire types from 1997 to 2016.

Figure 9 :
Figure 9: Global total natural and anthropogenic O2 flux from 2000 to 2013, with each component illustrated.

Figure 10 : 5 Earth
Figure 10: Global land net O2 flux in different regions from 2000 to 2013.(a) Tropical region.(b) Northern Temperate region.(c) Northern boreal region.(d) Southern temperate region.(e) Tibetan Plateau regions with an altitude greater than 3000m.The maps at the corner of each figure represent the calculated area.