Anthropogenic interventions of the global biogeochemical nitrogen (N) cycle have substantially raised the levels of reactive N [Nr, in general, including organic N, ammonium (NH 4-N in water and NH 3 in air), nitrogen oxides (NO x), nitrous oxide (N 2O), and nitrite & nitrate] in circulation, and caused severe damage to the environment at local, regional and global scales 1 2. Such dramatic increases can aggravate the impacts on the balance of climate forces in many ways, either through changes to the atmospheric constituents, or through feedbacks within the terrestrial ecosystem 3. The atmospheric N 2O emission directly contributes to global warming and ozone depletion; NO x compounds indirectly produce the short-lived climate-forcing agents (e.g. ozone, aerosol particles) and remove the long-lived climate-forcing agents (e.g. methane (CH 4)); NH 3 also increases the formation of aerosols. In addition, the deposited and applied N into the terrestrial ecosystem alters the carbon (C) exchange and greenhouse gas (GHG) fluxes between land and air 3. There are many cooling and warming interactions between N and the climate; the dominant effect of N on climate change is likely to vary across regions 4. For an equivalent increase in the human population, the emission mass of CO 2-C into the environment is about 40 times that of Nr 3. Although the absolute climate impact of Nr is dwarfed by the impact of CO 2 emissions, the variety of Nr species and sources, and the complexity of climate forcers, indicates a need to better assess the integrated effects of Nr and to identify the key forcers at different scales, so as to take targeted measures for N management. A few studies have attempted to assess the combined impacts of Nr at global and regional scales 5 6. The lack of systemic study of Nr on climate at the national level has also been noted by Pinder et al. 7, taking the United States as a case study. China is becoming the largest producer of Nr from both agricultural and industrial activities 8. The climate effects of Nr emissions in China are global rather than regional. The magnitude and acceleration of Nr creation in China warrants an integrated assessment of the significance of anthropogenic Nr on climate change.
N 2O, as one of the GHGs, has gained worldwide attention, and has also been thoroughly assessed in China 9 10. It is well known that reducing N 2O emissions offers the combined benefits of mitigating climate change and protecting the ozone layer 10. The increasing demands for food and energy in China have led to ever-increasing emissions of NO x and NH 3 11. These short-lived Nr emissions have multiple and profound effects on climate, the environment and human health 12. However, their relevance to climate in China is still unclear in previous studies. In addition, ecosystem N availability strongly affects the terrestrial net C balance 13, as well as many other ecosystem services associated with environmental quality 14. The terrestrial ecosystem C balance in China was investigated to under the influence of multiple global change factors 15 16. The effects of N deposition and N fertilizer use have been identified in a few studies 17 18.
In this context, a comprehensive assessment of Nr emissions and use on climate impacts in China is needed to fully understand the consequences of Chinese agricultural, industrial and urban development. To address this need, we seek to answer three questions: (i) What are the climate change impacts of Nr emissions in China, (ii) what are the relative contributions from agriculture and industry sources, and (iii) what are the future climate impacts of Chinese Nr based on different development scenarios? A key challenge is to establish a calculation database for the different climate forcers related to anthropogenic Nr in China. We review and reconcile the relative data and parameters from a variety of literatures related to the Nr emissions in China and their radiative forcing intensities and atmospheric lifetimes. Global temperature potential (GTP) is the common metric to quantify and intercomparing the relative effects of climate forcers over varying timescales. GTP evaluates not only the long-term climate effect caused by emissions of long-lived species like N 2O, but also the near-term climate fluctuations caused by emissions of the short-lived species NO x and NH 3 7 19. Accordingly, we use GTP to convert the climate change impacts of each of the processes to common units of equivalent Tg CO 2 (CO 2e).
In this analysis, we combine the set of Representative Concentration Pathway (RCP) scenarios, based on consistent radiative forcing levels 20, and the set of Special Report on Emission Scenarios (SRES) scenarios, based on the best estimates of the current emissions and further mitigation potential 21. The comparison between RCPs and SRES is useful to determine the mitigation potentials or pressures of China's development and make the reasonable N management measures.
The different gaseous Nr species show the different increase rates during 2000–2005 in China, with 11.4%, 44.0% and 8.3% for N 2O, NO x and NH 3, respectively ( Fig. 1). The emission sectors of Nr also show large variations. N 2O is the only form of Nr that has unambiguous global implications through the strong warming effect on climate (E1). Total anthropogenic N 2O emissions in China for 2005 were estimated at 1.04–1.22 Tg N. The primary source of N 2O is agriculture. N 2O emissions from agricultural soils (including emissions from applied fertilizer, manure and crop residues) amounted to a total of 0.59–0.87 Tg N, it can be further categorized as either direct (0.40–0.59 Tg) or indirect (0.19–0.28 Tg) emissions. Direct emissions are caused by N applied to the soils, while indirect emissions are related to subsequent processes after Nr emissions and re-deposition, or after leaching and runoff. Animal manure, including emissions from manure management and grazing animals, contributed 0.17–0.18 Tg N to the atmosphere. The emission factors for direct and indirect N 2O emissions from agriculture in China were 1.87% and 0.66%, a little higher than the global mean values of 1.48% and 0.42%, respectively 22. In addition, N 2O from fuel combustion, industrial production, waste disposal and biomass burning delivered only 0.14–0.30 Tg N to the atmosphere. As a result, N 2O exerts a strong warming effect at 345 ± 19 Tg CO 2e on a GTP 20 basis, then transforms to 311 ± 17 Tg CO 2e on a GTP 100 basis ( Fig. 2, Table S7). The close climate impacts on both timescales imply lasting effects on the climate system resulting from N 2O due to its long atmospheric lifetime of around 120 years.
NO x and NH 3 contribute to climate change indirectly through altering the atmospheric constituents and corresponding climate forcers. NO x is a major byproduct of combustion. Fuel combustion in China resulted in a total of 4.04–6.08 Tg N of NO x emissions in 2005, mainly from the combustion sectors of power plants (1.15–2.08 Tg N), transportation (1.26–1.66 Tg N), industry (1.35–1.79 Tg N) and household use (0.27–0.54 Tg N). Agricultural soils contribute a minor part with only 0.17 Tg NO x-N, mainly through microbial activity from agricultural soils. In sum, 4.34–6.24 Tg N is released to the atmosphere in the form of NO x. In comparison, the total emissions of NH 3 were 7.69–12.3 Tg N in 2005, becoming the largest source of atmospheric Nr in China. Agricultural soils (including volatilization and applied fertilizer and manure) and animal and human manure management are the main contributors, at 2.30–5.20 and 4.37–6.02 Tg N, respectively; other sources are negligible for the country as a whole. As a result, the combined climate impact of NO x on ozone and CH 4 (i.e. ozone-CH 4 effect, E2) in China is −350 ± 166 and −13.2 ± 7.2 Tg CO 2e on a 20-y and 100-y basis, respectively ( Fig. 2, Table S7). Our results show that the net cooling effect of NO x on ozone and CH 4 is notable only at a shorter time horizon, and gradually wears off with increasing time horizon. In addition, exposure to tropospheric ozone is known to inhibit photosynthesis, damage plants, and reduce plant primary productivity and land C sequestration capacity (E6) 23. The reduction in C sequestration due to the ozone effect for China has been simulated by the terrestrial ecosystem model and estimated at 111 ± 100 Tg CO 2e under controlled simulations for different scenarios 24. Therefore, the negative effects on the climate system and crop yields due to the indirect radiative forcing of ozone in China cannot be ignored. Both NO x and NH 3 can enhance the formation of light-scattering aerosols (E3). Compared with the NO x effect on the ozone-CH 4 effect (−350 ± 166 and −13.2 ± 7.2 Tg CO 2e on a GTP 20 and a GTP 100 basis, respectively), the direct aerosol effect of NO x is much less, with only −97 ± 39 and −0.006 ± 0.004 Tg CO 2e on a GTP 20 and a GTP 100 basis, respectively; while the corresponding NH 3 effects on aerosol are −59 ± 23 and −0.11 ± 0.13 Tg CO 2e, respectively. As the time horizon is extended, the direct aerosol effects of NO x and NH 3 both become smaller, and can be effectively ignored on a GTP 100 basis ( Fig. 2, Table S7).
The inputs of Nr into the terrestrial ecosystem are in the forms of N deposition and N fertilization. N addition into ecosystems can alter the physiology of soil microbes and vegetation, and concomitantly alter the balance of biogenic fluxes of GHGs 25. Our literature analysis results provide an explicit spatial map of N deposition over China ( Fig. 3a). N deposition sums to 15.1 Tg N/yr across China, with 3.56, 5.28, 5.68 and 0.57 Tg N deposited into cropland, forest, grassland and wetland, respectively ( Fig. 3b–e). The estimated N deposition is closed to the published inventories of NO x and NH 3 emissions used in our study. Without considering the response of N 2O emissions mentioned above, N addition can stimulate terrestrial CO 2 assimilation in both natural vegetation and agricultural land; meanwhile, CH 4 emissions caused by N simulation to some extent offset the benefit of N-induced CO 2 uptake in terms of climate change 25. These opposed impacts of Nr enrichment differ markedly depending on the land cover type 25 26. The assessed N-induced CO 2 and CH 4 emission/uptake factors for different ecosystem types of China are summarized in Table S6. Our results show that in China forest land has the largest potential for CO 2 uptake (−27 to −5 kg CO 2-C/kg N), while cropland has the largest potential for CH 4 emissions (0.01 to 0.03 kg CH 4-C/kg N). Applying the flux factors of CO 2 uptake and CH 4 emission by ecosystem types in China from Table S6 to estimate the net C exchange (NCE) due to N deposition, the overall result indicates that the uptake of CO 2-C (27 to 176 Tg C) overwhelmingly surpasses the CH 4-C emissions (0.11 to 0.21 Tg C), resulting in an NCE of 27 to 176 Tg in terrestrial ecosystems during 2000–2005 ( Table S6). Most regions in China's terrestrial system function as a net C sink; only a minor area in western China is an N-induced C source ( Fig. 3f). Our results further show that, during 2000–2005, forest land has the largest C sequestration potential, with 31.9 to 136 Tg C as a result of N deposition, followed by grassland (4.5 to 45 Tg N) and cropland (1.8 to 5.6 Tg N). For fertilized cropland, in addition to N deposition, applied N, including chemical and organic fertilizers, was evaluated at 35.7 Tg in 2005 8, resulting in a net CO 2 uptake of 14 to 58 Tg C and CH 4 emissions of 0.40 to 1.06 Tg C. The NCE of croplands due to N fertilization amounts to 13.6 to 57.0 Tg C.
On the whole, N addition has a positive effect on terrestrial C sequestration over the past several years. The C sequestration simulated by both N deposition and fertilization exerts a cooling effect, −50 ± 86 and −86 ± 51 Tg CO 2e on a GTP 20 and a GTP 100 basis, respectively ( Table S7). The climate cooling effects of N-induced C sequestration, however, offset only 18% and 35% of the warming effects of N 2O emissions from agriculture in terms of GTP 20 and GTP 100, respectively. Basically, taking the effects of these three GHG fluxes between air and land into account, Nr enrichment of the terrestrial ecosystem can be recognized as a strong warming agency.
On the national scale, anthropogenic Nr in China's ecosystem has a cooling effect on a GTP 20 basis in theory (−100 ± 414 Tg CO 2e) and a warming effect (322 ± 163 Tg CO 2e) on a GTP 100 basis ( Fig. 2, Table S7). However, the uncertainty of GTP 20 is so large that it is hard to conclude whether Nr causes cooling or warming in practice. The uncertainty is largely caused by the climate effects of atmospheric Nr. The collective climate-change impact of gaseous Nr emissions is cooling (in theory) at −50±328 Tg CO 2e on a GTP 20 basis but warming at 408±112 Tg CO 2e on a GTP 100 basis, respectively, while the climate-change impact of Nr addition into the terrestrial ecosystem is cooling at −50 ± 86 and −86 ± 51 Tg CO 2e, respectively. The climate forcers of these two components offset each other through the Nr flows and exchanges between the atmosphere and the land.
The relative contributions of agriculture and industry sectors are illustrated in Fig. 4. The overall impact of industrial Nr is cooling (−287 ± 306 Tg CO 2e) on a GTP 20 basis but converts to warming (136 ± 107 Tg CO 2e) on a GTP 100 basis. The reason is that the cooling effects are largely from the short-lived agency, i.e., NO x emissions from fuel combustion. The effect of agriculture Nr is dominated by the long-lived effect of N 2O emissions. As a result, agriculture shows a net warming effect and a minor difference at two timescales, at 187 ± 108 Tg CO 2e on a GTP 20 basis and 186 ± 56 Tg CO 2e on a GTP 100 basis, respectively. The climate effects of Nr from agricultural and industrial sources cancel each other out on a GTP 20 basis, but together contribute to warming effect on a GTP 100 basis.
Under the SRES I-Business as Usual (BAU) scenario, we assume that Nr emissions and Nr addition will keep the current annual rates of increase. In 2020 and 2050, Nr in China will produce a cooling effect at −314 and −725 Tg CO 2e on a GTP 20 basis and a warming effect at 528 and 939 Tg CO 2e on a GTP 100 basis ( Fig. 5 and Fig. 6, Table S8). The results indicate that the increase of Nr compounds without any measures will mitigate the climate change on a 20-y basis however aggravate the warming effect on a 100-y basis. On a 100-y basis, the cooling effect of N-induced C sequestration declines significantly, from 86.0 Tg CO 2e in 2005 to 57.3 Tg CO 2e in 2020, then approaches zero in 2050. By contrast, the impact of Nr emissions on climate change becomes more important, increasing from 408 Tg CO 2e in 2005 to 585 Tg CO 2e in 2020 and 939 Tg CO 2e in 2050. The SRES II-improvement scenario is built on the assumption that the N use efficiency of cropland will increase by 20% in 2020 and another 20% in 2050. Compared to the BAU scenario, on a GTP 20 basis, the improvement scenario enhances the net cooling effect with 41% and 38% for the years of 2020 and 2050, largely due to the agricultural N 2O emissions reduction. The SRES III-abatement scenario is built on the assumption of total quantity control for NO x in 2020 and an improved NO x removal rate in 2050, respectively. On a GTP 20 basis, the abatement scenario reduces the net cooling effect by 95% and 139% compared to the BAU scenario, ultimately appearing as a net warming effect in 2050. The reduction in NO x emissions has a strong negative feedback on climate change in the short term, whereas the increase in N use efficiency represents a positive effect on the mitigation of climate warming. Improving agricultural N use efficiency (i.e. N 2O reduction) is the most timely and effective measure for the mitigation of the climate impact of Nr in China. On a GTP 100 basis, the improvement and abatement scenarios both reduce the warming effect by 28% and 14% in 2020, respectively; the reduction effects transfer to 35% and 31% in 2050. Hence, reductions in agricultural N 2O and industrial NO x emissions have a comparatively large mitigation potential in the long term.
The detailed comparisons between SRES and RCP scenarios are showed in Fig. 5 and Fig 6. All RCP scenarios on a GTP 20 basis show a net warming effect, in contrary to our improvement scenario. However, the advantage of our improvement scenario diminishes with time. On a GTP 100 basis, its warming effect in 2020 (379.4 Tg CO 2e) nearly approaches that of the lowest RCP scenario (RCP2.6, 356.4 Tg CO 2e), or even in 2050 (613.6 Tg CO 2e) surpasses that of the highest RCP scenario (RCP8.5, 499.6 Tg CO 2e). The warming effect of our abatement scenario on a GTP 20 basis is lower than that of the RCP8.5 in 2020 and 2050. However, on a GTP 100 basis, its warming effect in 2020 (452.3 Tg CO 2e) and in 2050 (646.1 Tg CO 2e) become higher than that of RCP8.5 (423.5 and 499.6 Tg CO 2e in 2020 and 2050, respectively). The comparison results between SRES and RCP indicate that, (i) our reported SRES scenarios reveal the obvious mitigation potential on climate change but only on a short timescale not on a long timescale. (ii) the objectives of our scenarios in 2020 (i.e. realizing the total quantity control for NO x and improving the N use efficiency with 20%) is well-targeted to the concentration levels of RCP. However, the objectives in 2050 are too cautious to meet the emission under the RCP scenarios. More aggressive target should be made for the year of 2050.
China is mobilizing the largest anthropogenic Nr in the world due to agricultural, industrial and urban development. The climate effects of different Nr compounds are experiencing either the aggravation or the transformation processes. Based on the understanding of the dynamic of Nr on climate system, the investigation of possible changes and future trends is more meaningful than current status. Therefore, in this section, we analyze the dynamics and development trajectory of each climate forcer related to Chinese Nr, figuring out the key different points of China from other regions. Here, the differences between the SRES and BAU scenarios are defined as mitigation potential; while the differences between the SRES and RCP scenarios (here using RCP6.0) are defined as mitigation pressure. We further discuss the mitigation potential and mitigation pressure of individual N-related climate forcers ( Fig. 5).
The strong GTP and long lifetime of N 2O makes it the largest and most lasting contributor to global warming in China. Agriculture is responsible for about 80% of the climate warming effect of N 2O. The high N fertilizer input and low N use efficiency in China's agricultural fields are responsible for the increasing N 2O emissions 27. Our scenario analysis shows that a larger mitigation potential for climate change could be realized through improving N use efficiency to reduce agricultural N 2O emissions. Many approaches can be taken to enhance N use efficiency, for example, systematic crop rotation, wise use of fertilizer, and expanded use of nitrification inhibitors 28. Although the agricultural activities are the largest contributor, emission of N 2O from chemical industries also shows a large increase potential; and the control options are the most-effective in this sector 10. Hence, the all-around measures to reduce N 2O emission from agriculture and industry could lead to the co-benefits of climate mitigation and positive environmental effect through protecting ozone layer.
The realistic trajectory of NO x reflects a compilation of environmental policies and economic activities 29. NO x emissions in many industrialized countries are expected to decrease 30. For example, the NO x reductions in Europe in 2010 can be partly attributed to the global economic recession, and partly to the European NO x emission controls 29. Due to rapid economic development and the coal-dominated energy consumption structure, however, large increases in NO x from China have persisted over the past several decades 31. According to our results, the ozone-CH 4 effect of NO x emissions in China is the largest cooling agency at a short time horizon, but is nearly negligible at a longer time horizon. The ozone-CH 4 effect varies greatly across different regions 32 33 34. Investigations have shown that the ozone-CH 4 effect is more sensitive to ozone precursor changes from tropical regions (e.g. East Asia) than from mid-latitude and high-latitude regions (e.g. Europe and North America) 35. Therefore, control strategies that reduce emissions of ozone precursors in China have a significant impact on the net climate forces from ozone and CH 4. However, if only considering NO x alone, NO x emissions exert a negative force from decreased CH 4, which far exceeds the smaller positive force from increased ozone, resulting in a net cooling effect. The scenario of NO x emissions reduction alone lead to a positive force and a net warming effect from overwhelming increased CH 4, and as a result fails to produce mitigation potential for climate warming. Combined reductions in NO x and other ozone precursor emissions (such as CO) are beneficial, because they can produce a net cooling effect through the stronger negative force from decreased ozone 34. Hence, in order to obtain co-benefits from reducing radiative forces and mitigating air pollution, simultaneous emission reductions of various ozone precursors at the regional scale are feasible. In addition, the indirect radiative force on climate change by ozone effects on plant production has proved to contribute more to global warming than the direct radiative force of elevated ozone levels 36. The effect of NO x on plants is the secondary contributor to climate warming at both timescales. In China, the combined climate effects of current NO x emissions on ozone-CH 4, aerosol and the ozone-induced C sink were estimated at −336 ± 290 and 98 ± 97 Tg CO 2e on a GTP 20 and a GTP 100 basis, respectively. From the results of our scenario analysis, the NO x reduction has a secondary mitigation potential. Therefore, reduction of NO x in China can provide the co-benefits of climate-change mitigation and air pollution reduction in the long run. In especial, the NO x emission reduction still exist a large mitigation pressure when comparing with the RCP scenarios. China should confirm the implementation of total NO x quantity control required by taking policy measures. Meanwhile, continued push to improve the denitration level is an essential measure for mitigating NO x pollution.
China is a large agricultural country with an enormous animal population and tremendous N fertilizer consumption. The characteristics of NH 3 emissions reflect the intensity of synthetic fertilizer use and unique patterns of animal densities and types in China 37. As the human population continues to demonstrate considerable growth, food production and associated NH 3 emissions are likely to increase as well 38. Increased NO x and NH 3 produce radiative effects resulting from sulfate-nitrate aerosol and single-nitrate aerosol, respectively. The cooling effect of NO x through the direct aerosol effect is larger than NH 3 on a 20-y basis, but its depletion rate with time is faster than the rate of NH 3. Global model studies demonstrated that the increasing nitrate aerosol force would counteract the expected decline in the sulfate force by 2010 39. The future nitrate aerosol loads are projected to depend strongly on the emission strength of NH 3 40. It can be deduced that a larger increase in nitrate aerosols in China will be accompanied by larger increases of NH 3, as opposed to the smaller increases in Europe, the U.S. and Australia 40. Although the cooling effect of nitrate aerosols is negligible in the long term, the future air pollution from nitrate in the atmosphere can not be ignored.
Overall, the net cooling effects of NO x and NH 3 are short-lived and NO x even plays a warming role on the long timescale. Nonetheless, the multiple environmental problems resulting from NO x and NH 3, such as air pollution, eutrophication, etc., are more remarkable than the climate impact in China. In this regard, it is important to assess the trade-offs between the environmental and climate effects of the Nr compounds in the future studies.
China—especially the east region—is one of the global heavy N-deposition areas 41. Under the action of N deposition, our results show forest land in China is the single largest C sink, followed by grassland and cropland. The N status of different ecosystems in China can be reflected and further analyzed from the flux factors of CO 2 and CH 4 in Table S6. For example, N is usually found to be a limiting factor in plant growth in grassland ecosystems in China, and it is mostly unintended fertilization that has stimulated plant growth there 42; while almost all croplands can be regarded as N-saturated ecosystems due to continual application of chemical N fertilizers 43, as a results, the N-induced flux factor of CO 2 is only 0.45–1.7 kg CO 2/kg N. Additionally, there have been very different response patterns of forest ecosystems in China depending on the forest type, the N status of the soil, and the level of N deposition. In N-saturated mature tropical forests, high N addition can reduce soil respiration and CO 2 emissions 44, and decrease CH 4 uptake rate 45. However, in N-limited pine forests, elevated N deposition has no significant effect on soil respiration or CH 4 uptake 45 46. Overall, the N-induced flux factor of CO 2 in forest lands in China is 5–27 kg CO 2/kg N, much lower than that of the United States, with 24–65 kg CO 2/kg N 7. Throughout the whole of China, the terrestrial system acts as a net C sink due to N deposition. However, the potential for sustaining more terrestrial C storage as a result of atmospheric N deposition is limited 47. As the N deposition increases, the net primary productivity induced by per-unit N deposition has leveled off or declined since the 1980s, indicating that part of the deposited N may not be invested to stimulate plant growth, but instead leave the ecosystem by various pathways 18 48. In addition, globally N fertilizer application-induced CO 2 uptake began to level off in the 1980s whereas N 2O emission continued to grow 49. As a result, the climate benefits of the enhanced C sequestration would likely be eliminated due to increased N use 50. Finally, fertilized N will show an increasing warming effect in the future. This trend is more obvious in China. Excessive N-fertilization limits C sequestration. The declining rate of manure recycling to fields causes more N discharge into water or emitted into the air 51 52 53 54.
N saturation in forest land and excessive N fertilization in cropland in China limit the C sequestration capacity. The effect of N enrichment on C sequestration, as a result, has appeared to diminish in recent years. Hence, through the results of our scenario analysis, there is a certain GHG mitigation pressure in C sequestration through improving N management or reduction gaseous Nr emission in China. Efficient forest development and improved field management have shown the large potential of C sequestration in China's forest and cropland 23 55 56. Rational N fertilizer application, no-tillage and preferential crop residues management are widely recognized measures for enhancing C sequestration 52 56. Climate and environmental quality trade-offs and co-benefits, especially if they are applied nationwide in the near term, provide strong additional motivation for reducing N fertilizer use without yield loss in China.
Overall, anthropogenic Nr in China's environmental system has an obvious warming effect on a long timescale. The reductions in agricultural N 2O and industrial NO x emissions have a comparatively great mitigation potential for climate change in the long term; while a certain mitigation pressure appears in N-induced C sequestration only if realizing the efficient forest development and improved field management. The mitigation pressure of climate warming of anthropogenic Nr is still under great pressure with time. In 2050, more aggressive targets and more powerful measures should be taken to maintain in a specific radiative forcing level.
The quantification of the significances of climate forcers E1–E6 were calculated normally through multiplying the effective Nr volumes by the corresponding metric GTP, so as to reconcile and inter-compare the diverse climate forcers. We used the Monte Carlo Sampling (MCS) and Latin Hypercube Sampling (LHS) methods to simulate and propagate the uncertainties through the calculations, respectively ( Table S7). These methods are based on the principle that the empirical process of actually drawing many random samples and observing the outcomes can assess the behavior of a statistic in random samples. The uncertainties of parameters, activity data, and formulas are propagated through the sequence of calculations. In the MCS approach, we assumed each value in the uncertainty range had equal likelihood and no correlation between the variables. A population of 10,000 estimates was built and ranked in order. The median and 90% confidence intervals were reported. The LHS approach was further used to calculate the uncertainties in our research. All variables were divided into 300 equal intervals. An ensemble of 300 sets of all parameters was selected randomly and ranked in order. The mean and standard deviation were reported. The results of the uncertainty analysis using MCS and LHS were similar. Hence, we only discuss the results of the LHS method here.
Based on the possible or expected changes of different driving forcers, we set up a new set of three scenarios to depict the future trend of climate change ( Table S8). Short-term projections (to 2020) were developed using a "baseline" path of development, which considered abatement options consistent with China's situation and policy. For long-term projections (to 2050), we applied a "storyline" approach similar to that used in the IPCC Special Report on Emission Scenarios (SRES) scenarios 21. The situation in 2005 was used as the reference and those in 2020 and 2050 as the target years. (i) SRES I–Business as Usual (BAU) scenario is based on the assumption that Nr emissions will maintain the annual increase rate of 2000–2005 with no change. (ii) SRES II–Improvement scenario, builds on the BAU scenario and assumes that the N use efficiency of cropland will increase with an aspirational target of 20% until 2020 57 and a further 20% until 2050, which approaches the level of developed countries 27. (iii) SRES III-Abatement scenario, is designed to cut 10% of NO x emissions in accordance with the reduction target of the "Twelfth Five-Year Plan" in 2020, and assumed that the nationwide NO x removal rate reaches 80% in 2050. Because of the limited C storage of terrestrial ecosystems, the N-induced C sequestration potential is projected to decline 50% and be negligible with the increased N use in 2020 and 2050. The values in 2010, 2030 and 2040 are calculated on the basis of the logistic growth model. In addition, we compared our SRES scenarios with the published Representative Concentration Pathway (RCP), with four scenarios based on different concentration levels (RCP2.6, RCP4.5, RCP6.0 and RCP8.5, with the numbers referring to different radiative forcing levels) rather than emissions 20.
This work was supported by the National Key Basic Research Program (2014CB953800). We thank the IIASA for the use of the GAINS Model database GAINS-CHINA; the JRC-IES for the use of the EDGRA v4.2 and the ECCAD-GEIA database to gain other emission inventories related to reactive nitrogen.