The global carbon cycle and possible disturbances due to man's interventions

Global atmospheric CO₂ concentration has increased since the beginning of reliable monitoring in 1958 at a mean rate of about 0.9 ppm CO₂/yr. Now, atmospheric CO₂ concentration is at 330 ppm. From about 1860 up to 1974, man's intervention in the global carbon cycle caused a likely increase of 76.6 × 10¹⁵ gC, corresponding to 36 ppm CO₂ in the atmosphere, if a preindustrial content of 294 ppm CO₂ or 625.3 × 10¹⁵ g C is adopted to be valid. A further rise of atmospheric CO₂ seems to be inevitable and probably will be responsible for a climatic warming in the next several decades; therefore, a global examination of carbon reservoirs and carbon fluxes has been undertaken to determine their storage capacity for excess carbon which orginated mainly from burning fossil fuels and from land clearing. During 1860–1974 about 136 × 10¹⁵ g C have ben emitted into the atmosphere by fossil fuel combustion and cement production. At present, the emission rate is about 5 × 10¹⁵ g C/yr. The worldwide examination of carbon release, primarily by deforestation and soil cultivation since 1860, is estimated to be about 120 × 10¹⁵ g C. The net transfer of carbon to the atmosphere owing to man's interference with the biosphere is now believed to be about 2.4 × 10¹⁵ g C/yr. An oceanic uptake of roughly 179 × 10¹⁵ g C since 1860 is open to discussion. According to the chemical buffering of sea surface water only about 35.5 × 10¹⁵ g C could have been absorbed. It is argued, however, that oceanic circulations might have been more effective in removing atmospheric excess carbon of anthropogenic origin.     

Carbon dioxide and climate: an astrophysical perspective

In this survey the earth is viewed from the astrophysical perspective, i.e. using global mean values of environmental parameters. The role of carbon dioxide is described in the processes of energy transfer from the earth's surface to space, which determine “global climate” as measured by the mean surface temperature. Analogies and differences between the problems of the terrestrial atmosphere and those of the solar and stellar atmospheres are examines, both in the computation of model atmosphere and in remote sensing of atmospheric temperature and composition. Subsequently, the temporal astrophysical perspective, with a review of the evolution of CO₂ abundance and climate on astrophysical or geological time scales, on earth as on Venus (the runaway greenhouse) and on Mars is introduced. Variation of CO₂ may have been critical to the maintenance of an environment in which life could originate and evolve, and may itself have been affected by life. On human time scales, the recent and continuing increase in atmospheric CO₂ raises new problems, which are briefly surveyed. It is argued, that the differential greenhouse effect of increased CO₂ in the earth's atmosphere is essentially identical to the “blanketing effect” of spectral lines on the temperature structure of stellar atmospheres. The methods used by astrophysicists in such studies are reviewed and compared with those used to evaluate the differential greenhouse effect of CO₂ in radiative-convective models of the earth's atmosphere. The latter methods remain relatively crude, but recent results by different authors are in reasonably good agreement; however, the astrophysical perspective, i.e. the use of one-dimensional global mean models, remains a gross simplification of the real complexity of the earth's climate system, which is also true in stellar atmospheres.     

Variations of anthropogenic CO2 in urban area deduced by radiocarbon concentration in modern tree rings

Radiocarbon concentration in the atmosphere is significantly lower in areas where man-made emissions of carbon dioxide occur. This phenomenon is known as Suess effect, and is caused by the contamination of clean air with non-radioactive carbon from fossil fuel combustion. The effect is more strongly observed in industrial and densely populated urban areas. Measurements of carbon isotope concentrations in a study area can be compared to those from areas of clear air in order to estimate the amount of carbon dioxide emission from fossil fuel combustion by using a simple mathematical model. This can be calculated using the simple mathematical model. The result of the mathematical model followed in this study suggests that the use of annual rings of trees to obtain the secular variations of ¹⁴C concentration of atmospheric CO₂ can be useful and efficient for environmental monitoring and modeling of the carbon distribution in local scale.     

Energy Use and CO2 Production in Tropical Agriculture and Means and Strategies for Reduction or Mitigation

Carbon dioxide emissions due to fossil fuel consumption are well recognized as a major contributor to climate change. In the debate on dealing with this threat, expectations are high that agriculture based economies of the developing world can help alleviate this problem. But, the contribution of agricultural operations to these emissions is fairly small. It is the clearing of native ecosystems for agricultural use in the tropics that is the largest non-fossil fuel source of CO₂ input to the atmosphere. Our calculation show that the use of fossil energy and the concomitant emission of CO₂ in the agricultural operational sector - i.e. the use of farm machinery, irrigation, fertilization and chemical pesticides - amounts to merely 3.9% of the commercial energy use in that part of the world. Of this, 70% is associated with the production and use of chemical fertilizers. In the absence of fertilizer use, the developing world would have converted even more land for cultivation, most of which is completely unsuitable for cultivation. Current expectations are that reforestation in these countries can sequester large quantities of carbon in order to mitigate excessive emissions elsewhere. But, any program that aims to set aside land for the purpose of sequestering carbon must do so without threatening food security in the region. The sole option to liberate the necessary land for carbon sequestration would be the intensification of agricultural production on some of the better lands by increased fertilizer inputs. As our calculations show, the sequestration of carbon far outweighs the emissions that are associated with the production of the extra fertilizer needed. Increasing the fertilizer use in the developing world (without China) by 20%, we calculated an overall net benefit in the carbon budget of between 80 and 206 Mt yr^?1 dependent on the carbon sequestration rate assumed for the regrowing forest. In those regions, where current fertilizer use is low, the relative benefits are the highest as responding yield increases are highest and thus more land can be set aside without harming food security. In Sub-Saharan Africa a 20% fertilizer increase, which amounts to 0.14 Mt of extra fertilizer, can tie up somewhere between 8 and 19 Mt of CO₂ per year (average: 96 t CO₂ per 1 t fertilizer). In the Near East and North Africa with a 20%-increased fertilizer use of 0.4 Mt yr^-1 between 10 and 24 Mt of CO₂ could be sequestered on the land set aside (40 t CO₂ per 1 t fertilizer). In South Asia this is 22–61 Mt CO₂ yr?1 with an annual additional input of 2.15 Mt fertilizer (19 t CO₂ per 1 t fertilizer). In fact, carbon credits may be the only way for some of the farmers in these regions to afford the costly inputs. Additionally, in regions with already relatively high fertilizer inputs such as in South Asia, an efficient use of the extra fertilizer must be warranted. Nevertheless, the net CO₂ benefit through implementation of this measure in the developing world is insignificant compared to the worldwide CO₂ output by human activity. Thus, reforestation is only one mitigating measure and not the solution to unconstrained fossil fuel CO₂ emissions. Carbon emissions should, therefore, first of all be reduced by the avoidance of deforestation in the developing world and moreover by higher energy efficiency and the use of alternative energy sources.     

Evaluation of carbon stock variation in Northern Italian soils over the last 70 years

Carbon (C) sequestration in soils is gaining increasing acceptance as a means of reducing net carbon dioxide (CO₂) emissions to the atmosphere. Numerous studies on the global carbon budget suggest that terrestrial ecosystems in the mid-latitudes of the Northern Hemisphere act as a large carbon sink of atmospheric CO₂. However, most of the soils of North America, Australia, New Zealand, South Africa and Eastern Europe lost a great part of their organic carbon pool on conversion from natural to agricultural ecosystems during the explosion of pioneer agriculture, and in Western Europe the adoption of modern agriculture after the Second World War led to a drastic reduction in soil organic carbon content. The depletion of organic matter is often indicated as one of the main effects on soil, and the storage of organic carbon in the soil is a means of improve the quality of soils and mitigating the effects of greenhouse gas emission. The soil organic carbon in an area of Northern Italy over the last 70 years has been assessed In this study. The variation of top soil organic carbon (SOC) ranged from −60.3 to +6.7%; the average reduction of SOC, caused by agriculture intensification, was 39.3%. This process was not uniform, but related to trends in land use and agriculture change. For the area studied (1,394 km²) there was an estimated release of 5 Tg CO₂-C to the atmosphere from the upper 30 cm of soil in the period 1935–1990.     

A laboratory study on CO2 emission from asphalt binder and its reduction with the use of warm mix asphalt

Oxidation of hydrocarbon in asphalt binder leads to the production of carbon dioxide (CO₂) during the production of hot mix asphalt. The objective of this laboratory study was to investigate the effects of the asphalt additive Sasobit^®, asphalt content and mixing/placement temperature on CO₂ emissions from binder with laboratory measurements. The isolated effects of Sasobit on asphalt absorption into the aggregate were also looked at. Temperature was found to be the only statistically significant factor on emissions. This would suggest that warm mix asphalt technology, which employs the use of Sasobit in asphalt mixtures, is a very effective way of lowering the industry's CO₂ emission impact, both directly and by the use of less energy for heating. This work predicts that greater than 30% reduction of CO₂ emissions is possible with typically used levels of Sasobit.     

Soil gas (²²²Rn, CO₂, ⁴He) behaviour over a natural CO₂ accumulation, Montmiral area (Drôme, France): geographical, geological and temporal relationships

The south east basin of France shelters deep CO₂ reservoirs often studied with the aim of better constraining geological CO₂ storage operations. Here we present new soil gas data, completing an existing dataset (CO₂, ²²²Rn, ⁴He), together with mineralogical and physical characterisations of soil columns, in an attempt to better understand the spatial distribution of gas concentrations in the soils and to rule on the sealed character of the CO₂ reservoir at present time.Anomalous gas concentrations were found but did not appear to be clearly related to geological structures that may drain deep gases up to the surface, implying a dominant influence of near surface processes as indicated by carbon isotope ratios. Coarse grained, quartz-rich soils favoured the existence of high CO₂ concentrations. Fine grained clayey soils preferentially favoured the existence of ²²²Rn but not CO₂. Soil formations did not act as barriers preventing gas migrations in soils, either due to water content or due to mineralogical composition. No abundant leakage from the Montmiral reservoir can be highlighted by the measurements, even near the exploitation well. As good correlation between CO₂ and ²²²Rn concentrations still exist, it is suggested that ²²²Rn migration is also CO₂ dependent in non-leaking areas - diffusion dominated systems.     

Global monitoring of carbon dioxide in the atmosphere

The historic development of the scientific interest in monitoring CO₂ in the atmosphere is the subject of this article. Particular emphasis is placed upon activities initiated by the USA in the 1950s which led to the establishment of the Mauna Loa Observatory and further developed into the existing world-wide monitoring system for air constituents and air pollution operated under the auspices of the World Meteorological Organization, a major contributer to the Global Environmental Monitoring System of the UN Environment Programme. Recent studies on the feasibility of monitoring the background level of CO₂ at stations throughout the world have indicated considerable difficulties resulting from the influences of the biosphere. These problems have led WMO to adopt new criteria for CO₂ monitoring station locations which are presented in a discussion of future needs and plans for global monitoring of CO₂ in the atmosphere.     

Carbon cycle in spruce ecosystems of the northern Taiga subzone

The main components of the carbon balance have been determined in old spruce-bilberry forests of the northern taiga subzone. Annual carbon deposition in live phytomass and necromass has been determined by the weight method. Photosynthetic carbon binding has been calculated using the chlorophyll index, and the daily carbon balance has been estimated on the basis of direct measurements of CO₂ exchange. The results have shown that photosynthetic carbon binding by the phytocenosis amounts to 3.5–4 t/ha per year. Taking into consideration the litter yearly deposition decreased up to 1 t C/ha per year. With more than 70% of carbon accumulated in the organic mass being oxidized within the phytocenosis and returned to the atmosphere in the form of CO₂. Spruce ecosystems serve as a sink for 0.2–0.3 t C/ha per year.     

Land use and energy scenarios affecting the global carbon cycle

Past increase of atmospheric CO₂ involves significant contributions from both fossil and biospheric sources. The latter are controversial, partly because these CO₂ releases may be balanced by accelerated regrowth following clearing of some forests, while others were being converted to agricultural or other nonforest land. A simulation model was used to reconstruct changes since 1860 and project four hypothetical future scenarios of CO₂ injection to 2460. Nineteen compartments and their exchanges of carbon were considered. Areal extents of tropical forests, other wooded ecosystems, and nonforests were incorporated into the model. Rapidly and slowly exchanging pools of carbon per unit area were projected by integrating income-loss differential equations numerically. Estimated cumulative releases of CO₂ from fossil fuels (plus cement) near 120 Pg of carbon (1 Pg = 10¹⁵ g) from 1860 to 1970 were assumed to equal the prompt plus delayed releases due to forest clearing. Limits of exploitable forest area and biomass were evaluated and found to contribute much less future CO₂ than usable fossil fuels. Ultimate release from the latter (7500±2500 Pg C) could increase atmospheric CO₂ manyfold: doubling the assumed 1860 levels as early as (1) the year 2032 for assumed early fossil-use scenarios and (2) the year 2045 for late fossil-use scenarios. Depending on the poorly known parametes that were programmed to constrain the organic production rates, cumulative storage, and the response of plants and soils to enhanced atmospheric CO₂, biospheric storage might reach higher levels for all scenarios than the estimates given here. However, maximizing such storage in real life would require much closer understanding and wiser management of ecosystems than history has shown.     

Climate, entropy and environment

The IPCC Working Groups I–III 2007 publications does not consider the question of the influence of the entropy increase in the atmosphere on current climate development. An investigation into this question, both in general terms as well as by two quantitative approaches, reveals we must consider the entropy produced by man in connection with climate development, especially with regard to the temperature increase of the atmosphere. The IPCC report also fails to mention the production of CO₂ by humans and livestock, but calculations show we must also consider such greenhouse gas CO₂ production. For solving the mitigating processes, we therefore have to take into account both the human induced entropy production and the direct human and livestock CO₂ output. In consideration of these findings, it seems necessary to introduce an “entropy identity” to people who wish to be able to continue to live on the planet. The introduction of an entropy tax might also help in solving the most urgent fundamental problem humanity has ever had to face. Readers should send their comments on this paper to: BhaskarNath@aol.com within 3 months of publication of this issue.     

Carbon Emission from the Surface of Coarse Woody Debris in Korean Pine Forests of Southern Primorye

Carbon dioxide fluxes from the surface of coarse woody debris (CWD) have been measured in Korean pine forests of the southern Sikhote-Alin mountain range. The seasonal dynamics of oxidative conversion of CWD carbon have been evaluated, and average values of the CO₂ emission rate have been determined for CWD fragments of three tree species at different stages of decomposition. The degree of decomposition is an important factor of spatial variation in CO₂ emission rate, and temporal variation in this parameter is adequately described by an exponential function of both CWD temperature and air temperature (R² = 0.65–0.75).     

Greenhouse gas emissions from a semi-arid tropical reservoir in northeastern Brazil

We estimated carbon dioxide (CO₂) and methane (CH₄) emissions by diffusion, ebullition, and degassing in turbines from a semi-arid hydropower reservoir in northeastern Brazil. Sampling sites were allocated within the littoral and deeper waters of one embayment, the main-stream, and at turbines. Annual carbon emissions were estimated at 2.3?×?10⁵?±?7.45?×?10⁴ t C year^?1, or in CO₂-equivalents (CO₂-eq) at 1.33?×?10⁶?±?4.5?×?10⁵ t CO₂-eq year^?1. Diffusion across the water surface was the main pathway accounting for 96% of total carbon emissions. Ebullition was limited to littoral areas. A slight accumulation of CO₂, but not of CH₄, in bottom waters close to the turbines inlet led to degassing emissions about 8?×?10³ t C year^?1. Emissions in littoral areas were higher than in main-stream and contribute to 40% of the total carbon. Carbon (C) emissions per electricity generated, at 60% of installed capacity, is 0.05 t C-CO₂-eq MWh^?1. The ratio increases to 0.09 t C-CO₂ MWh^?1, equating 80% of the emissions from natural gas and 40% of diesel or coal power plants. Retention time and benthic metabolism were identified as main drivers for carbon emissions in littoral areas, while water column mixing and rapid water flow are important factors preventing CH₄ accumulation and loss by degassing. Our results indicate that Itaparica Reservoir, located in the semi-arid region of Northeastern Brazil, acts as a source of GHGs. Management measurements are needed to prevent emissions to raise in the future.     

Observations and modelling of thoron and its progeny in the soil-atmosphere-plant system

Samples of pasture vegetation, mainly Trifolium pratensis, were collected at the Botanic Garden of the University of Bologna during the period 1998-2000 and measured by gamma-spectrometry for determining thoron progeny. Concentrations of ²¹²Pb were between 1.5 and 20 Bq m⁻², with individual peaks up to 70 Bq m⁻². Soil samples were collected at the same location and physically characterised. Their chemical composition (particularly Th and U) was determined by X-ray fluorescence spectroscopy. Lead-212 on plants mainly originates from dry and wet deposition of this isotope generated in the lower atmosphere by the decay of its short-lived precursor ²²⁰Rn, which is produced in the upper soil layers as a member of the natural thorium decay chain and exhales into the atmosphere. Concentrations of ²²⁰Rn in the atmosphere depend on (1) the amount of Th present in soil, (2) the radon fraction which escapes from the soil minerals into the soil pore space, (3) its transport into the atmosphere, and (4) its redistribution within the atmosphere. The mobility of radon in soil pore space can vary by orders of magnitude depending on the soil water content, thus being the main factor for varying concentrations of ²²⁰Rn and ²¹²Pb in the atmosphere. We present a simple model to predict concentrations of thoron in air and its progeny deposited from the atmosphere, which takes into account varying soil moisture contents calculated by the OPUS code. Results of this model show close agreement with our observations.     

Farm-level assessment of CO2 and N2O emissions in Lower Saxony and comparison of implementation potentials for mitigation measures in Germany and England

Greenhouse gases (GHG) emissions from agricultural farming practice contribute significantly to European GHG inventories. For example, CO₂ is emitted when grassland is converted to cropland or when peatlands are drained and cultivated. N₂O emissions result from fertilization. Enabling farmers to reduce their GHG emissions requires sufficient information about its pressure–impact relations as well as incentives, such as regulations and funding, that support climate-friendly agricultural management. This paper discusses potentials to improve the supply of information on: farm-specific climate services or impacts, present policy incentives in Germany and England that support climate-friendly farm management and related adaptation requirements. Tools which have been developed for a farm environmental management software (to be added after review because of potential identification) are presented. These tools assess CO₂ emissions from grassland conversion to cropland and peatland cultivation, as well as N₂O emissions from nitrogen fertilization. As input data, the CO₂ tool requires a classification of soil types according to soil organic carbon storage. The input data based on soil profile samples was compared with reference data from the literature. The N₂O tool relies on farm data concerning fertilization. These tools were tested on three farms in order to determine their viability with respect to the availability of required data and the differentiation of results, which determines how well site-specific conservation measures can be identified. Assessing CO₂ retention function of grassland conservation to cropland on the test farms leads to spatially differentiated results (~100 to ~900 potentially mitigated t CO₂ ha^?1). Assessed N₂O emissions varied from 0.41 to 1.1 t CO₂eq. ha^?1 a^?1. The proposed methods support policies that promote a more differentiated funding of climate conservation measures. Conservation measures and areas can be selected so that they will have the greatest mitigation effects. However, even though present policy instruments in Germany and England, such as Cross Compliance and agri-environmental measures, have the potential to reduce agricultural GHG, they do not appear to guide measures effectively or site-specifically. In order to close this gap, agri-environmental measures with the potential to support climate protection should be spatially optimized. Additionally, the wetland restoration measures which are most effective in reducing GHG emissions should be included in funding schemes.     

Soil Respiration and Carbon Storage of an Acrisol Under Forest and different Cultivations in rio de Janeiro State, Brazil

Soil respiration rates of a clay-loam textured Acrisol under different uses (Atlantic forest, manioc, horticulture and pasture) from Rio de Janeiro State were measured. The relationship between carbon dioxide (CO₂) emissions and soil physico-chemical properties were investigated. Rates of CO₂ emission of two sites (Atlantic forest and horticulture) were also evaluated in different seasons in 1997 and 1998. In the forest site, monthly means of measured respiration rates showed good correlation with soil temperature in the range 19.6–24.1°C (r² = 0.89). In the horticulture site, no change was observed with soil moisture alone, in the range 3.0–13.2 wt%. In the horticulture soil, even when the surface soil was very dry, respiration rates increased in the hot, wetter summer but remained higher than the mean flux from forest soil. The CO₂ emission flux of the Acrisol under different use showed good correlation with soil temperature (r² = 0.72) and moisture (r² = 0.61).     

Carbon dioxide removal from power plant exhausts

The various options for removing carbon dioxide (CO₂) from power plant exhaust gases are reviewed. An absorption-stripping system using amine solutions and dissolution of gaseous CO₂ into deep ocean waters is a feasible choice with a marginal cost of electricity below 15%.     

Radon, helium and CO2 measurements in soils overlying a former exploited oilfield, Pechelbronn district, Bas-Rhin, France

The Pechelbronn oilfield (Rhine Graben, France), where mining activity ended in the 1960s, has been used for waste disposal for twenty years. Since the wastes are varied, work is underway to identify the discharged materials and their derivatives, as well as to locate and quantify potential discharge sites. Two major goals were assigned to the present work. The first was to identify or refine the location of hidden structures that could facilitate gas emanation up to the surface, by studying soil gas concentrations (mainly ²²²Rn, CO₂, CH₄ and helium) and carbon isotope ratios in the CO₂ phase. The second was devoted to examining, from a health and safety viewpoint, if the use of the oilfield as a waste disposal site might have led to enhanced or modified gas emanation throughout the area.It appeared that CO₂ and ²²²Rn evolution in the whole area were similar, except near some of the faults and fractures that are known through surface mapping and underground observations. These ²²²Rn and CO₂ anomalies made it possible to highlight more emissive zones that are either related to main faults or to secondary fractures acting as migration pathways. In that sense, the CO₂ phase can be used to evaluate ²²²Rn activities distant from tectonic structures but can lead to erroneous evaluations near to gas migration pathways. Dumping of wastes, as well as oil residues, did not appear to have a strong influence on soil gaseous species and emanation. Similarly, enhanced gas migration due to underground galleries and exploitation wells has not been established. Carbon isotope ratios suggested a balance of biological phenomena, despite the high CO₂ contents reached. Other monitored gaseous species (N₂, Ar, H₂ and alkanes), when detected, always showed amounts close to those found subsurface and/or in atmospheric gases.     

The Suess effect: Carbon-Carbon interrelations

The Suess Effect is a term which has come to signify the decrease in ¹⁴C in atmospheric CO₂ owing to admixture of CO₂ produced by the combustion of fossil fuels. This term is here extended, as a concept, to the shifts in isotopic ratio of both ¹³C and ¹⁴C in any reservoir of the carbon cycle owing to anthropogenic activities. To explain this generalized Suess Effect a four reservoir global model of the natural carbon cycle is developed in which isotopic fractionation and radioactive decay are fully taken into account. The model includes the cases in which the deep ocean is treated either as a single undifferentiated box model reservoir or is vertically differentiated with eddy diffusion governing the transport of carbon. Also, the governing equations are expressed with sufficient generality to apply simultaneously to both rare isotopes. In so far as possible, the model is expressed without approximation of the isotopic processes even though this leads to non-linear differential equations to describe the rates of change of rare isotopic carbon within carbon reservoirs. Linear approximations also developed and solved using the method of Laplace transforms. The sensitivity of the predicted Suess Effects to uncertainties in the assigned values of the model parameters is investigated in detail, including estimates of some of the effects of linearizing the governing equations.The approximation of Stuiver, in which the atmospheric Suess Effect is assumed to be 0.018 times the corresponding effect for ¹⁴C, is examined in detail and shown to arise when both isotopic fractionation and radioactive decay are left out of the model. This approximation, although correct as to order of magnitude, is found to be too imprecise to be recommended in modeling studies.As found in previous work, the predicted atmospheric Suess Effect for ¹³C for a given airborne fraction of industrial CO₂ is of similar magnitude whether the land biosphere has been a net source or sink of carbon during recent times. On the other hand, the corresponding effect for a surface ocean water is considerably smaller than otherwise if the land biosphere has been a source of CO₂ instead of a sink. The model is thus useful in indicating the need to consider isotopes in several reservoirs simultaneously.Although the emphasis is on formulating models rather than surveying and interpreting data, observational data are summarized and compared with model predictions. The oceanic data are seen to be too meager as yet to help settle the question of biospheric response to man's activities.     

Identifying the determinants and spatial nexus of provincial carbon intensity in China: a dynamic spatial panel approach

Is emission intensity of carbon dioxide (CO₂) spatially correlated? What determines the CO₂ intensity at a provincial level? More importantly, what climate and economic policy decisions should the China’s central and local governments make to reduce the CO₂ intensity and prevent the environmental pollution given that China has been the largest emitter of CO₂? We aim to address these questions in this study by applying a dynamic spatial system generalized method of moment technique. Our analysis suggests that provinces are influenced by their neighbours. In addition, CO₂ intensities are relatively higher in the western and middle areas, and that the spatial agglomeration effect of the provincial CO₂ intensity is obvious. Our analysis also shows that CO₂ intensity is nonlinearly related to gross domestic product, positively associated with secondary-sector share and foreign direct investment, and negatively associated with population size. Important policy implications are drawn on reducing carbon intensity.