A New High-Resolution N2O Emission Inventory for China in 2008

Article pubs.acs.org/est

A New High-Resolution N2O Emission Inventory for China in 2008 Feng Zhou,*,†,‡ Ziyin Shang,† Philippe Ciais,‡,§ Shu Tao,† Shilong Piao,†,‡ Peter Raymond,∥ Canfei He,† Bengang Li,† Rong Wang,§ Xuhui Wang,†,§ Shushi Peng,§ Zhenzhong Zeng,† Han Chen,† Na Ying,⊥,† Xikang Hou,⊥,† and Peng Xu⊥,† †

Laboratory for Earth Surface Processes, College of Urban and Environmental Sciences, Peking University (PKU), Beijing, 100871, P.R. China ‡ Sino-France Institute of Earth Systems Science, Peking University, Beijing, 100871, P.R. China § Laboratoire des Sciences du Climat et de l’Environnement, CEA CNRS UVSQ, 91191 Gif-sur-Yvette, France ∥ School of Forestry and Environmental Studies, Yale University, New Haven, Connecticut 06511, United States ⊥ College of Territorial Resources and Tourism, College of Environmental Sciences and Engineering, Anhui Normal University (ANU), Wuhu, 241002, P.R. China S Supporting Information *

ABSTRACT: The amount and geographic distribution of N2O emissions over China remain largely uncertain. In this study, county-level and 0.1° × 0.1° gridded anthropogenic N2O emission inventories for China (PKU-N2O) in 2008 are developed based on high-resolution activity data and regional emission factors (EFs) and parameters. These new estimates are compared with previous inventories, and with two sensitivity tests: one that uses high-resolution activity data but the default IPCC methodology (S1) and the other that uses regional EFs and parameters but starts from coarser-resolution activity data. The total N2O emissions are 2150 GgN2O/yr (interquartile range from 1174 to 2787 GgN2O/yr). Agriculture contributes 64% of the total, followed by energy (17%), indirect emissions (12%), wastes (5%), industry (2.8%), and wildfires (0.2%). Our national emission total is 17% greater than that of the EDGAR v4.2 global product sampled over China and is also greater than the GAINS-China, NDRC, and S1 estimates by 10%, 50%, and 17%, respectively. We also found that using uniform EFs and parameters or starting from national/provincial data causes systematic spatial biases compared to PKU-N2O. Spatial analysis shows nonlinear relationships between N2O emission intensities and urbanization. Per-capita and per-GDP N2O emissions increase gradually with an increase in the urban population fraction from 0.3 to 0.9 among 2884 counties, and N2O emission density increases with urban expansion.

■

INTRODUCTION Driven mainly by the increasing spread of man-made reactive nitrogen into the environment,1 atmospheric concentrations of N2O have increased from 270 ppb before the industrial revolution to 325 ppb.2 China is estimated to have accounted for ∼15% of global anthropogenic N2O emissions since 2001.3 A prerequisite for regulating N2O emissions in China and defining realistic policy-relevant reduction targets is obtaining accurate and spatially explicit emission inventories, which is the major goal of this paper. Anthropogenic N2O emissions depend on a combination of natural processes and human drivers.4−7 This complexity makes N2O emissions difficult to model using process-based models. Local N2O flux data are also notoriously difficult to scale up into emissions maps due to their high variability.8−10 Typically, national N2O emissions are calculated using emission factors (EFs) and activity data,3,11 such as EDGAR v4.2.3 Using this approach, annual N2O emissions from China for the past decade have been estimated to range from 1080 to 1992 Gg.3,12,13 The large range of reported EF values, the lack of data © XXXX American Chemical Society

about other parameters that control emissions (EPs), and the scarcity of high-resolution activity data are the main sources of uncertainty associated with N2O emission inventories.11,14,15 Despite these difficulties, the EFs and other EPs used by most inventories are from IPCC guidelines.3,12,13,16 Recent direct observations indicate that the IPCC default values are often erroneous over China17−22 as well as in other countries.8,23 Additionally, national or provincial statistical data on fertilizers, agricultural production and energy consumption are used for estimating N2O emissions, and the emissions are usually downscaled from the scale of large administrative units to higher spatial resolution using proxies (e.g., landuse, population density), such as in EDGAR v4.23 and GAINS-China.12 A limitation of this approach is that it assumes perfect consistency between national data from international organizations and the Received: April 10, 2014 Revised: June 23, 2014 Accepted: June 25, 2014

A

dx.doi.org/10.1021/es5018027 | Environ. Sci. Technol. XXXX, XXX, XXX−XXX

Environmental Science & Technology

Article

resolution N2O emission inventory presented here will be useful for the Earth system modeling.

sum of regional/subnational data from local governments. A second limitation is that it assumes a linear relationship between activity and proxy data, whereas per-area and percapita activity data actually vary widely in space. Thus, the mean value and trends of inventories based on coarse-scale data may be affected by a series of biases of unknown sign and magnitude. Improvements in the spatial allocation of N2O emissions can be made by using local data. However, such studies have been limited to specific sectors of N2O emissions. For instance, Klemedtsson et al.24 developed a model using C:N ratios of the topsoil to estimate N2O emissions from forested Histosols (as a soil consisting primarily of organic materials) at Finish and German sites. Lily et al.25 developed a spatially explicit model to derive emission rates from managed soils in Scotland based on a combination of local soil, climate, land use and fertilizer data. Huang et al.26 applied a Bayesian model to predict the spatial patterns of N2O emissions of pastures in Australia based on soil temperature, moisture and NO3−. The spatial dependency considered in their models improved the upscaling of field observations to regional estimates, and the effectiveness of the approach was shown to depend on the sample density and scale of observations. These three studies produced highresolution emission inventories but were limited to natural sources and specific sectors. Recently, Zhang et al.27 and Wang et al.28 developed regression models for disaggregated province-level data to produce high-resolution emissions maps of PAHs,27 combustion CO229 and black carbon28 in China. Their emissions maps are derived from province-level data of fuel consumption that were further downscaled into 0.1° × 0.1° maps using regression models with county-level predictive variables. Although this spatial disaggregation approach may be applicable for countylevel estimates of N2O emissions and may improve the spatial accuracy relative to the disaggregation of national/provincial data, potential biases can arise, particularly if the activity data are dramatically unbalanced within the area of the provinces. The results of the studies described above indicate that the uncertainty of the spatial distribution of N2O emissions can be substantially decreased by using data with the highest possible spatial resolution. In this study, a high-resolution N2O emission inventory for China (PKU-N2O) for 2008 was produced on both countylevel and 0.1° × 0.1° scales. We compiled and used point, county-level and gridded activity data with regional EFs and other EPs for most emission sources. The detailed raw data from each county in China allowed us to construct a model of the geographic distribution of emission estimates at a high resolution for six emission sectors (Materials and Methods section). We then analyzed the emission totals, source profile, spatial patterns and predictive uncertainty (Results section) and compared our results with previous studies6,30,31 (Discussion section). To understand the effect of using county-level data on the spatial distribution of N2O emissions, local EFs and EPs, two sensitivity tests that use coarser-resolution data are also presented in the Discussion section. Because our county-level estimates of N2O emissions do not rely on spatial proxies, we were able to analyze the relationships between the N2O emission intensities and urbanization variables, such as population density, urban population percent, and urban area percent, in the context of agricultural management and urbanization policies (last section of the paper). The high-

■

MATERIALS AND METHODS Emission Sources. Twenty-two types of N2O sources in six sectors are considered: agriculture [A], energy [E], industry [I], waste (W), wildfires (L), and indirect emissions (ID) (Table 1). Emissions of N2O from product uses, domestic navigation, fugitive emissions, waste incineration, and composting/storage of wastes were not re-estimated because they make up a small part (0.4%) of the total emissions estimated by EDGAR v4.2.3 For these sources, we used the published estimates of EDGAR v4.2.3 In the PKU-N2O emission model, the emission ES,T(y) of county T from emission source S on an annual basis y is calculated by 44 ES , T (y) = ·∑ ∑ [ADS , T , R (y) ·EFS , T , C · 28 R C f (EPi , S , T , R , C(y), ai)]

(1)

where ADS,T,R(y) is the county-specific (county-average, rastertype, point source location) activity data for each source S and each predictive variable R in year y, EFS,T,C is the region-specific EF for each source S under condition C (e.g., management systems, cultivation types), EPi,S,T,R,C(y) is a region-specific emission controlling parameter i, ai is a coefficient, and f(·) is a function whose shape depends on source type. The term 44/28 is a conversion coefficient of N2O−N to N2O emissions. Detailed descriptions of the equations used for each source (Section S1), activity data, EFs, and EPs (Section S2) are available in the Supporting Information (SI) and are briefly summarized below. Activity Data. County-level data of the annual amounts of synthetic fertilizers, crop types (15), livestock types (8), and urban and rural populations for 2008 were obtained for 2884 counties from 334 municipal statistical registers in Mainland China, Taiwan,30 Hong Kong,31 and Macau.32 However, fertilizer, crop, and livestock data for 332 counties in Mainland China were not accessible for 2008. Because the county-level data from other years were available, we used temporal interpolation to 2008 for these counties. Complementary gridded activity data include fuel consumption and wildfire burned area (0.1°), ammonia emissions (1 km), soil carbon changes (0.5°), organic soils (1 km), and NOx emissions (0.1°). These activity maps come from PKU-FUEL-2007,29 the Huang et al.’s NH3 emission inventory,33 simulations by global ecosystem models (ORCHIDEE,34 LPJ,35 and LPJ-GUESS;36 SI Figure S2), the Harmonized World Soil Database v1.2,37 and EDGAR v4.2,3 respectively. Finally, point data of Adipic acid industrial production (OAP) from 14 large-scale chemical companies (representing >92% of the total Chinese industrial production in 200838) were precisely allocated to their individual grid cells. Nitric acid production (NAP) was taken from the global point source data set of SRI Consulting39 but was corrected by actual provincial-level production in 2008 from the China Industrial Economic Statistical Yearbook.40 A detailed description of activity data is given in SI Table S1. EFs. Eight types of EFS,T,C were assessed using published field measurements from China after elimination of the experiments without control treatment. Information, including their means, standard deviations [SD], sample sizes, and data sources for all EFs are given in SI Table S4. In contrast to previous studies,28,29 we use the arithmetic mean of different B



Article

Table 1. Total N2O Emissions (GgN2O·yr−1), Source Profile in 2008 and their Comparisons with Previous Works PKUN2O

EDGAR v4.2f

GAINSCHINAg

other estimates

Agriculture (A)a A1 manure management A2 manure in pasture/ range/paddock A3 synthetic fertilizers A4 organic fertilizers A5 crop residues A6 N-mineralization in soil A7 histosols management A8 agricultural waste burning

1375.2 550.8

985.0 124.8

1697.9 260.9

−

57.7

198.9

5.1

Energy (E) E1 public electricity and heat production E2 manufacturing industries and construction E3 residential and other sectors E4 road transportation E5 rail transportation E6 other energy sourcesb

code

name

Industrial Processes and Product Use (I) I1 nitric acid production I2 adipic acid production I3 product uses Waste (W)c W1 wastewater handling W2 other waste handling Wildfires (L)d Indirect emissions (ID) ID1 indirect emissions from atmospheric deposition ID2 indirect emissions from N leaching/ runoffe total

546g 1339.9

390.9h 70.6h 29.2h

69.3

96.2

137−181i

4.1

0.8

2.4h−2.8i

0.4

−

−

355.5

176.1

123.9

−

168.8

85.9

44.0

156.0j

110.0

21.9

16.5

−

36.4

44.8

36.6

−

27.0 6.2 7.1f

8.5 7.9 7.1

20.5 0.3 6.0

− − −

58.1

195.7

42.5

− 0.6−1.4k

5.8

14.0 1.9 26.6

108.6 107.6 1.0

68.3 67.3 1.0

67.9 67.9 −

− 102.5l −

4.8

2.4

−

−

248.3

354.2

−

−

−

−

−

−

1932

1080m

484.4 166.5 37.2

9.3 43.0 5.8f

661.0

189.9

224.3

for different animal types. Both slurry and drylot systems are widespread across China according to the results of Livestock Manure Sector in the National Pollution Source Survey [NPSS] Database41−43). The EFs of different sector-fuel combinations for energy consumptions were compiled from reported measurements in China for coal use (public electricity and heating production,44,45 manufacturing industries and construction,45 and the residential sector45), oil used in road transportation,46 and biomass used in the residential sector.46 EPs. The values of specific EPs are based on local statistics and data from the literatures or NPSS database.41−43 The EP η(T,S), which defines the fraction of manure managed in each management system for each livestock type, differs greatly by province, animal type, and farm size in China.33,47 η(T,S) was calculated as the product of the percent of intensive (or extensive) livestock production system and the proportion of manure management systems (SI Table S5); the former was taken from China’s Animal Industry Yearbook,47 and the latter was provided by Huang et al.33 For the N excretion rate Nex(T), different IPCC default values from the NPSS database42 are used for the livestock species/category in each province (SI Table S6). The ratio of nitrogen (N) lost through leaching and runoff Leach(T,R) was obtained from long-term observations (362 samples for N leaching at 55 experimental locations and 477 samples for N runoff at 79 experimental locations; SI Figure S5 and Table S7) regrouped for each cultivation type and Agro-Climate Zone (SI Table S1). The values of the other EPs are listed in SI Table S8. Uncertainty Analysis. A MC ensemble simulation was performed to estimate the uncertainty of the emissions in our model. The emission model was run 1 000 000 times by randomly varying all of the input data given a priori uncertainty distributions given by the coefficients of variation (CVs), where uniform distribution was applied for activity data. The CV of each activity data is assumed to be equal to the absolute value of the average difference between a given data set from China used in PKU-N2O and a default global data set. CV values of 0.2 were set for the following activities in the absence of global data sets: productions of peanut, sugar cane, other oilplants, highland barley and alfalfa, loss of soil carbon, NAP, OAP, area of organic soils, NH3 and NOx emissions, and dry matters burned of vegetation 67. The CV of energy consumption is set to 0.05.48 The prescribed a priori uncertainty of the EFN2O and EP values for the MC simulations were based on their means and SDs (see SI Tables S4−S8) and were calculated from observed data sets or set to a default CV = 0.5 when no local observations were available, and normal distribution was applied. The differences between local and global data sets in 2008 and the detailed values of CVs are listed in SI Tables S2 and S3, respectively. The PKU-N2O map of total emissions at the resolution of county was generated by summing 22 maps corresponding to different sources. We further mapped the emission on a 0.1° × 0.1° grid to compare our results with other studies, where the detailed description of spatial allocation to 0.1 degree grid is shown in SI Text S2.

−

354.2 24.0 2150

1782

a

Exclude indirect N2O from leaching/runoff in agriculture and savanna fires. bInclude domestic aviation, fossil fuel fires, fugitive emissions from oil and gas, fugitive emissions from solid fuels, other energy industries, other transportation, inland navigation. cInclude waste incineration and other wastewater handling from EDGAR v4.2. d Include forest, deforestation, savanna, and peat fires. eInclude indirect N2O from leaching/runoff in agriculture. fFrom EDGAR v4.2; gResults in 2007 from Hu and Wang.49 hResults in 2007 from Gao et al.14 i Results in 2000 from Gu et al.66 jResults in 2007 from Wu et al.44 k Results in 1990 from Li and Lin.50 lResults in 2008 from Zhou et al.51 m Results in 2004 from NDRC.13

■

experiments, which is an input into Monte Carlo (MC) ensemble simulations. As an example, the EF of direct N2O emissions from managed soils were compiled individually for different cultivation types in six Agro-Climate Zones (defined in SI Figure S1) based on 620 individual values from 73 experimental locations (SI Figure S4). The EFs of manure management were determined for slurry and drylot systems and

RESULTS Total Emissions and Source Profile. The total N2O emissions (Etotal) from China for 2008 are estimated as 2150 GgN2O·yr−1. The mean per unit area N2O emissions (Earea) is 0.2 MgN2O·km−2·yr−1, and the per-capita and per-GDP average emission intensities (Ecap and Egdp) are 1.6 kgN2O·cap−1·yr−1 C



Article

Figure 1. Geographic distributions of total and per capita N2O emissions from all sources at county-level in 2008 from the PKU-N2O inventory developed in this study. (a) Total N2O emissions, (b) per capita N2O emissions; (c) relative contributions of six sectors (i.e., A, E, W, I, L, ID) defined in Table 1 for seven regions (SI Figure S1), where the area of each pie is proportional to the emission totals. Note: AH-Anhui, HB-Hebei, CQ-Chongqing, HN-Henan, IM-Inner Mongolia, JL-Jilin, JS-Jiangsu, TB-Tibet, XJ-Xinjiang, ZJ-Zhejiang, 1-Changchun City, 2-Nong’an County, 3Jiutai County, 4-Chengdu City.

and 0.5 gN2O·USD−1·yr−1, respectively. The interquartile range derived from the MC simulation is from 1174 to 2787 GgN2O· yr−1 (SI Figure S6). The main economic sectors that dominate the total N2O emissions are agriculture (1375 GgN2O, 64%), energy production (356 GgN2O, 16%), and indirect emissions (248 GgN2O, 12%). Waste (109 GgN2O, 5%), industry (58 GgN2O, 2.8%), and wildfires (5 GgN2O, 0.2%) have smaller contributions. Table 1 shows the emissions by source. Manure management, synthetic fertilizers, indirect emissions from nitrogen depositions, public electricity and heat production, and organic fertilizers contribute 74% in the country’s total emissions. Fuel consumption by manufacturing industries and construction (5%) and waste handling (5%) are of less importance but are non-negligible. Spatial Patterns. Figure 1 shows the geographic distributions of Etotal and Ecap in 2008; the relative contributions of six sectors in seven regions are shown in pie charts. Using countylevel data to create this N2O emissions map reveals a strong spatial association with the distributions of cropland and major city clusters. The average emission density over eastern China (including the Northeast Plain, North China Plain, and southern hills; defined in SI Figure S1) is 0.43 MgN2O·km−2· yr−1, which is more than five times greater than that over western China (0.08 MgN2O·km−2·yr−1). In addition, the highest emission densities are found in the North China Plain, Northeast Plain, Lianghu Plain, and the Sichuan Basin (defined in SI Figure S1), where most of the cereal and livestock productions and energy consumption in China are distributed. High N2O emission densities are also present over megacities and city clusters (especially along the eastern coast), where the population, power plants, and industries are concentrated, and large amounts of wastewater are generated. The values of Ecap in China vary dramatically from 0 to 64 kg N2O·cap−1·yr−1 across counties, but its spatial pattern is almost opposite to that of the N2O emission densities (Figure 1b). For example, Ecap in

Tibet is approximately 7 kg N2O·capita−1·yr−1 compared to only 0.5 kg N2O·cap−1·yr−1 in the industrialized Zhejiang province. In addition, the considerable differences between the relative contributions of the six sectors across the six AgroClimate Zones (SI Figure S1) primarily reflect the different distributions of industrial activities and land use. Eastern China (8.7% area of China) is the largest contributor of N2O emissions and accounts for nearly 25% of the total (Figure 1c). Agriculture is the most important emission sector in northeast, central and southwest China (>55% of the total). Agriculture is also the largest contributor in northern China (44%), but the relative contribution of energy consumption (29%) is higher than in other regions due to a large number of coal consumed.

■

DISCUSSION Differences with Other Studies. We compared our estimates for China to a global-scale data product (EDGAR v4.2),3 GAINS-China,12 a report from the National Development and Reform Commission of China (NDRC),13 and other estimates.14,44,49−51 Our estimate is 17% higher than EDGAR v4.2 (1782 GgN2O·yr−1; same source mix and year as this study; Table 1), 10% higher than GAINS-China (1932 GgN2O· yr−1 interpolated between 2005 and 2010; same source mix as this study, but wildfires and indirect emissions were ignored, 2008), and 50% higher than NDRC (1080 GgN2O·yr−1; source mix not clearly documented, 2004). These differences are explained by the differences in the use of local and highresolution activity data, regional EFs and EPs for agricultural soils, manure management, energy-related combustions, and indirect emissions. The source mix of PKU-N2O is significantly different from that used in previous estimates. Our agricultural and energyrelated emissions are factors of 1.4 and 2.0, respectively, larger than those of EDGAR v4.2 (Table 1). The results are different from EDGAR v4.2 (573 GgN2O·yr−1) because PKU-N2O uses D



Article

Figure 2. N2O emissions from six sectors in 2008. Note: Panels a−d show PKU-N2O at 0.1° grid scale, whereas Panels e−h illustrate EDGAR v4.2 estimates at 0.1° grid scale; (a, e) agriculture (MgN2O·km−2·yr−1), (b, f) energy (MgN2O·km−2·yr−1), (c, g) sum of waste (W), industry (I), and wildfires (L) (MgN2O·yr−1), (d, h) indirect emissions (MgN2O·km−2·yr−1).

Figure 3. China’s N2O emission map and its comparison analysis. Note: Panel a shows PKU-N2O at 0.1°, Panel b shows EDGAR v4.2 global product sampled over China, Panel c shows the S1 estimate, Panel d shows the S2 estimate. Note: Emissions from Northeast Plain and Sichuan Basin are shown in the insets at the bottom-left of the maps.

capacities used in EDGAR v4.2 (∼0.22 million tons38). Our indirect emissions of N2O through leaching and runoff from managed soils are only 70% of the value from EDGAR primarily because the Leach(T,R) values used in this study (1.2%−8% for paddy and 0.45%−7.9% for upland; SI Table S7) are much smaller than the IPCC value (30% for paddy and upland).52 Although the difference in the country emission totals between PKU-N2O and GAINS-China is smaller than the

lower percentages of manure in pasture/range/paddock (27%− 39%), higher percentages of drylot manure (44%−75%), and two times greater values of EFs for rice paddy cultivation and coal and petroleum consumption than the IPCC defaults used in EDGAR v4.2.25 Conversely, the N2O emitted by industrial processes in PKU-N2O is only 30% of the EDGAR v4.2 estimate; this occurred because we used actual adipic acid production data for 2008 (0.15 million tons38) rather than the E



Article

national total and source mix; these values define very different baselines for each sector that can be used to support emission reduction policies. Effects of Input Data on Emission Spatial Patterns. To test the sensitivity of the N2O emission spatial patterns to the input data and methodology, we designed two sensitivity tests with the PKU-N2O emission model at a scale of 0.1°. The S1 test assumes uniform IPCC EF and EP values and uses the actual county-level activity data (Figure 3c), where the differences between S1 and EDGAR v4.2 are their source and spatial resolution of activity data. The symmetrical S2 test keeps the regional EF and EP distributions but creates maps from provincial disaggregation using regression models (Figure 3d), which are described in SI Text S3 and Figure S3. The S1 test gives total N2O emissions of 1781 GgN2O·yr−1, which is 17% less than the original PKU-N2O value. We also found differences between the three runs in the spatial allocation of the emissions (Figure 3). First, the S1 emission pattern is highly correlated with the PKU-N2O pattern (R = 0.97, p < 0.05) but systematically under-predicts values in eastern, northeastern, and southern China and the Sichuan Basin. For example, the high N2O emission densities in the PKU-N2O results are located primarily in the Anhui and Jiangsu provinces of eastern China (Figure 3a), whereas S1 allocates more N2O emissions to the Henan and Hebei provinces of northern and central China because S1 ignores regional differences in EFs (Figure 3c). Second, the spatial correlation between S2 and PKU-N2O is 0.74 (p < 0.05). This lower correlation suggests that provincial disaggregation27 using regression models is not capable of capturing the fine scale structure of the spatial distribution of fertilizers, crops and animal husbandry within provinces even it fits limited province-level data sets well. Consequently, extreme values of several counties within a province can be overestimated or underestimated by S2. For example, the regression equation for poultry (y=1.16 × 10−10x2+7.28 × 10−04x+4.02 × 1003, R2 = 0.89, p < 0.001, where x is agricultural GDP) approximates the provincial data well but under-predicts the emissions in Jiutai and Nong’an counties, which contribute ∼70% and ∼20% of the poultry production in Jilin province and China, respectively. Similarly, the amount of N fertilizers (mineral and organic) used in agriculture across China can be predicted accurately using agricultural GDP as the predictive variable (y = 0.1103x, R2 = 0.87, p < 0.001), but the specific N2O emissions from fertilizers around the city of Chengdu in the Sichuan Basin are overestimated by more than 40% in S2 (Figure 3d). These results suggest that in China and other countries, it is essential to compile and use activity data at the smallest spatial scale possible (e.g., counties) and to compile local EF and EP data if possible.

difference with EDGAR v4.2, GAINS-China has larger direct emissions from managed soils than this study (by 618 Gg N2O· yr−1; i.e., 75%) and lower emissions from E1, E2, rail transportation (E5), and OAP (I2) (Table 1). Local estimates of several emission sources have also been documented since 2000 (Table 1). For example, the emissions from Chinese croplands (sectors A3, A4, A5) estimated by Gao et al.14 are smaller than our estimates, likely because their EFs were calculated using a cube root transformation and because they compiled less original EF data (456 samples). Although the mean value after the transformation could be a better estimate of the general tendency if the distribution of EF samples is positively skewed, it reduced the contributions of the higher values of observations on the N2O emission estimates. Differences in the Spatial Distribution of Emissions. Systematic differences in the spatial patterns of emissions between this study and EDGAR v4.2 are shown in Figures 2 and 3. EDGAR v4.2 shows that emissions decline sharply outside the intensive cropland areas of the North China Plain and Sichuan Basin and the urbanized hot-spots of the Pearl River and Yangtze River Delta (Figures 2b). Our emissions maps have smoother spatial distributions from the denser cropland and urban regions to less dense forest/grassland and rural regions (Figure 2a). EDGAR v4.2 has lower local emission densities than our study for the agriculture, energy, and waste sectors (Figure 3e and f) but higher densities for the industrial and indirect emissions sectors (Figure 3g and h). For a more detailed comparison, the absolute (AD) and relative differences (RD) are defined as AD = E1 − E2 and RD = (E1 − E2)/E2, where E1 and E2 (>0) are the emissions of each 0.1° grid cell in EDGAR v4.2 and PKU-N2O, respectively. SI Figure S7 shows the spatial distributions of AD and RD. The distribution of AD has a mean (maximum or minimum) value of 0.14 (20.8) and −0.58 (−32.4) Mg N2O·km−2·yr−1 for the regions of positive and negative AD values, respectively. The correlation coefficient between EDGAR v4.2 and PKU-N2O is only 0.13 when compared at a resolution of 0.1° (p < 0.05). The RD values between PKU-N2O and EDGAR v4.2 range between 2.03 and −0.42. Approximately 45% of the grid cells have RD > 0.5. Larger RD values are located in low-emission regions (Tibet, Qinghai, Xinjiang, Inner Mongolia, non-agricultural lands in Southern China; defined in Figure 1). Interestingly, very high emission densities in some counties within a province are revealed by PKU-N2O. For example, the high emissions in Changchun City, where poultry and cattle production is very intensive (5 Mg N2O·km−2·yr−1), are not apparent in EDGAR v4.2, which assigns the average value of Jilin province to this county (0.3 Mg N2O·km−2·yr−1; see inset maps in Figure 2b). Similarly, the very high emissions in the North China Plain and Sichuan Basin are concentrated in a few counties in PKU-N2O but are smoothed over the entire region in EDGAR v4.2 (Figure 2b). These differences between PKU-N2O and previous inventories mainly reflect methodological differences. Both EDGAR v4.2 and GAINS-China used national activity data, spatially uniform EF data for each sector, and default values for EPs. Except for point sources, both EDGAR v4.2 and GAINS-China have emission patterns that parallel the spatial extents of cropland, population density and FAO animal densities. However, the spatial variations of fertilizer application rates, energy use efficiencies, and the EPs η(T,S), Nex(T) and Leach(T,R) were not accounted for. Therefore, the other methodologies tend to have systematic differences with PKU-N2O for both the

■

N2O EMISSIONS AND URBANIZATION A key research and policy question is how the urbanization process affects China’s N2O emissions. In general, N2O emissions show a strong positive correlation with population size and GDP (R2 = 0.81 and 0.76, respectively, using power law functions), but there is still a debate about the relationship between the spatial gradients of N2O emission intensities and urbanization. Our method relies on county-level data and thus is independent of population density or land use data sets (except for wildfires and parts of energy sector). It provides a chance to analyze the spatial relationships between Ecap, Egdp, and Earea and independent predictor variables such as F



Article

Figure 4. Trends in Ecap, Egdp, Earea along with urbanization gradients based on county-level data set. (a) Size (population density), (b) structure (Upopulation), (c) urban expansion (Uarea). Note: The lines refer to 100-ordered-county moving average for Ecap (red line, kgN2O·cap−1·yr−1), Egdp (blue line, gN2O·USD−1·yr−1), and Earea (black line, tN2O·km−2·yr−1); Ecap (green solid cycle) at national scale was also demonstrated for two regions (1-Africa and 2-EU27) and the largest nine countries by area (China, 3-Kazakhstan, 4-Russia, 5-Brazil, 6-U.S., 7-Australia, 8-Argentina, 9-Canada, 10India), the source of N2O emissions are UNFCCC for Annex I Parties (https://unfccc.int/ghg_data/) and EDGAR v4.23 for non-Annex I Parties, respectively; Population and GDP data for these countries are from UNSD (https://unstats.un.org/) and World Bank (http://data.worldbank.org/), respectively.

relationship between N2O emission intensities and Upopulation, except over sparsely populated regions, casts doubt on the hypothesis that urbanization is an analogous process to the growth of ecosystems.55 The relationships shown in Figure 4b suggest that the increasing Upopulation in China’s counties may increase N2O emission intensities in the future, possibly because urban food demand, with a greater per capita calorific intake and an increased proportion of animal consumption, would stimulate agricultural activities both within urban counties and in surrounding counties.56 Effects of Urban Land Cover Fraction. Figure 4c shows the relationships between N2O emission intensities and Uarea. Interestingly, the shapes of Ecap and Egdp are similar to Laplace distributions; in other words, both Ecap and Egdp increase rapidly with Uarea and reach maximums of 4 kgN2O·cap−1·yr−1 and 3 gN2O·USD−1·yr−1, respectively, until Uarea > 0.025. However, Ecap and Egdp then decrease abruptly with increasing Uarea (>0.025). Earea behaves differently from Ecap and Egdp and steadily increases with Uarea. This result suggests that N2O emission intensities are lower in megacities (higher population densities and Uarea) than in counties with higher Upopulation but smaller urban areas; however, their N2O emission densities (Earea) are much higher than other regions in China. Therefore, it would be efficient for megacities and city clusters to reduce N2O emissions by updating their energy structure and shifting human dietary choices toward a lower per capita calorific intake in the future.56−58 This decrease of demand would eventually reduce some of the food and energy supplies in these regions. Implications. The results presented above show the utility of N2O emissions maps created from native high-resolution activity data and spatially explicit EFN2O and EP values. Compared with previous inventories, PKU-N2O provides highresolution maps of China’s N2O emission totals, source mix, and spatial patterns in 2008. These results emphasize the advantage of using emissions inventories at a high spatial resolution. Activity data at coarse spatial scales (national or province-level) cannot be used to accurately model spatial variations within provinces by scaling inference of provincial disaggregated regression approach. However, by using highresolution data including activity data, EFs, and EPs, we are able to identify the hot-spots of N2O emissions and their spatial

population density, urban population fraction (Upopulation), and percentage of urban area (Uarea). We focused on the spatial gradients between N2O emission intensities and urbanization using a simple moving average approach. Population and GDP data for 2884 counties were obtained from local statistics (SI Table S2), and a 1 km land use map was provided by the Institute of Geographic Sciences and Natural Resources Research53 (SI Table S2). We chose the moving average values at a fixed subset size of 100 because the spatial gradients were relatively insensitive to subset size from 30 to 200 (SI Figure S8). Effects of Population Density. The shapes of the curves in Figure 4a reflect the effects of increasing population density on the N2O emission intensities. Ecap and Egdp sharply decrease with increasing population density below 400 cap·km−2 (Figure 4a) then slowly decline and become stable at approximately 1 kgN2O·cap−1·yr−1 (moving average value) and 0.3 gN2O· USD−1·yr−1. It could be explained by the fact that Ecap and Egdp due to allochthonous emission sources (e.g., indirect emissions from nitrogen depositions) in low-population-density or lowGDP counties is greatly larger than that in high-populationdensity or high-GDP counties, though Ecap or Egdp due to autochthonous sources are approximately close to each other among different counties. Conversely, Earea is positively correlated with population density at a rate of approximately 1 kgN2O cap−1·yr−1 (Figure 4a) due to the increase of intensive agricultural activities or energy consumption with population growth. Effects of Percent of Urban Population. Ecap and Egdp are not negatively correlated with Upopulation. For example, Ecap decreases and remains stable at approximately 2 kgN2O·cap−1· yr−1 when Upopulation is less than 0.3 (Figure 4b), but Ecap increases gradually with Upopulation above this threshold. Interestingly, this county-level relationship between Ecap and Upopulation is consistent with that of national data except for high and low population density nations (e.g., India and Canada; Figure 4b). Egdp decreases dramatically with increasing Upopulation until Upopulation < 0.4 but increases for Upopulation > 0.7 (up to values 2.4 gN2O·USD−1·yr−1, Figure 4b). The result that the urban population generally has more efficient emission rates of greenhouse gases is not intuitive.54 The positive G



Article

(3) Emission Database for Global Atmospheric Research (EDGAR): Global Emissions EDGAR v4.2. http://edgar.jrc.ec.europa.eu/ overview.php?v=42 (accessed November 11, 2013),. (4) Bouwman, A. F.; Boumans, L. J. M.; Batjes, N. H. Modeling global annual N2O and NO emissions from fertilized fields. Global Biogeochem. Cycles 2002, 16 (4), 1080 DOI: 10.1029/2001GB001812. (5) Li, X. L.; Yuan, W. P.; Xu, H.; Cai, Z. C.; Yagi, K. Effect of timing and duration of midseason aeration on CH4 and N2O emissions from irrigated lowland rice paddies in China. Nutr. Cycling Agroecosyst. 2011, 91 (3), 293−305. (6) Liu, C. Y.; Wang, K.; Meng, S. X.; Zheng, X. H.; Zhou, Z. X.; Han, S. H.; Chen, D. L.; Yang, Z. P. Effects of irrigation, fertilization and crop straw management on nitrous oxide and nitric oxide emissions from a wheat-maize rotation field in northern China. Agric., Ecosyst. Environ. 2011, 140 (1−2), 226−233. (7) Stehfest, E.; Bouwman, L. N2O and NO emission from agricultural fields and soils under natural vegetation: Summarizing available measurement data and modeling of global annual emissions. Nutr. Cycling Agroecosyst. 2006, 74 (3), 207−228. (8) Griffis, T. J.; Lee, X.; Baker, J. M.; Russelle, M. P.; Zhang, X.; Venterea, R.; Millet, D. B. Reconciling the differences between topdown and bottom-up estimates of nitrous oxide emissions for the U.S. Corn Belt. Global Biogeochem. Cycles 2013, 27, 746−754. (9) Huang, J.; Golombek, A.; Prinn, R.; Weiss, R.; Fraser, P.; Simmonds, P.; Dlugokencky, E. J.; Hall, B.; Elkins, J.; Steele, P.; Langenfelds, R.; Krummel, P.; Dutton, G.; Porter, L. Estimation of regional emissions of nitrous oxide from 1997 to 2005 using multinetwork measurements, a chemical transport model, and an inverse method. J. Geophys. Res.: Atmos. 2008, 113 (D17), D17313 DOI: 10.1029/2007JD009381. (10) Thompson, R. L.; Chevallier, F.; Crotwell, A. M.; Dutton, G.; Langenfelds, R. L.; Prinn, R. G.; Weiss, R. F.; Tohjima, Y.; Nakazawa, T.; Krummel, P. B.; Steele, L. P.; Fraser, P.; Ishijima, K.; Aoki, S. Nitrous oxide emissions 1999−2009 from a global atmospheric inversion. Atmos. Chem. Phys. 2014, 14, 1801−1817. (11) Ogle, S. M.; Buendia, L.; Butterbach-Bahl, K.; Breidt, F. J.; Hartman, M.; Yagi, K.; Nayamuth, R.; Spencer, S.; Wirth, T.; Smith, P., Advancing national greenhouse gas inventories for agriculture in developing countries: Improving activity data, emission factors and software technology. Environ. Res. Lett. 2013, 8(1), DOI: 10.1088/ 1748-9326/8/1/015030. (12) International Institute for Applied Systems Analysis (IIASA): Greenhouse Gas-Air Pollution Interactions and Synergie (GAINSCHINA). http://gains.iiasa.ac.at/gains/EAN/index.login?logout=1 (accessed November 12, 2013). (13) National Development and Reform Commission (NDRC) China’s National Climate Change Program. http://www.china-un.org/ eng/gyzg/t626117.htm (accessed November 12, 2013). (14) Gao, B.; Ju, X. T.; Zhang, Q.; Christie, P.; Zhang, F. S. New estimates of direct N2O emissions from Chinese croplands from 1980 to 2007 using localized emission factors. Biogeosciences 2011, 8 (10), 3011−3024. (15) Lu, Y. Y.; Huang, Y.; Zou, J. W.; Zheng, X. H. An inventory of N2O emissions from agriculture in China using precipitation-rectified emission factor and background emission. Chemosphere 2006, 65 (11), 1915−1924. (16) Zhou, J. B.; Jiang, M. M.; Chen, G. Q. Estimation of methane and nitrous oxide emission from livestock and poultry in China during 1949−2003. Energy Policy 2007, 35 (7), 3759−3767. (17) Song, C. C.; Xu, X. F.; Tian, H. Q.; Wang, Y. Y. Ecosystematmosphere exchange of CH4 and N2O and ecosystem respiration in wetlands in the Sanjiang Plain, Northeastern China. Global Change Biol. 2009, 15 (3), 692−705. (18) Tang, X. L.; Liu, S. G.; Zhou, G. Y.; Zhang, D. Q.; Zhou, C. Y. Soil-atmospheric exchange of CO2, CH4, and N2O in three subtropical forest ecosystems in southern China. Global Change Biol. 2006, 12 (3), 546−560. (19) Tian, S. Z.; Ning, T. Y.; Zhao, H. X.; Wang, B. W.; Li, N.; Han, H. F.; Li, Z. J.; Chi, S. Y. Response of CH4 and N2O emissions and

variations. More importantly, by capturing the local variations and by compiling emission estimates that do not rely on a proxy for spatial allocation, we can describe the spatial trends of the emissions intensities along with the urbanization gradients. Additionally, the spatial patterns of N2O fluxes inferred using atmospheric N2O measurements and inverse models remain uncertain,59,60 partly due to the lack of a high-resolution emission inventory. Thompson et al.60 noted that the accuracy and spatial resolution of China’s N 2 O emissions are considerably more important than global prior fluxes for atmospheric inversions. Therefore, the high spatial resolution emissions inventory developed in this study would provide a better prior for N2O inversion and reduce the uncertainty of simulations of terrestrial N2O fluxes.61 Finally, our findings of nonlinear relationships between the emissions intensities and urbanization suggest that the efficiency of fertilizer use should be improved in counties with high Upopulation values but small urban areas and that the energy structure and human dietary choices in megacities or city clusters should be more environmentally friendly as urbanization increases in the future. However, there are several uncertainties in PKU-N2O, especially in the emissions from agricultural soils and manure management (SI Figure S6). Therefore, additional experiments and the use of a reliable data-driven approach4,62 or processbased models19,63−65 could significantly improve the spatial resolution and reduce the uncertainties of the emissions inventory.

■

ASSOCIATED CONTENT

S Supporting Information *

Section S1 (Methodology, including Texts S1−S3, Figures S1− S3), Section S2 (Data on activity data, EFs, and EPs, including Tables S1−S8, Figures S4−S5), Section S3 (Supplement results, including Figures S6−S8), Section S4 (Supporting references). This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +86 10 62758845; fax: +86 10 62756560; e-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was jointly financed by the National Natural Science Foundation of China (grant no. 41201077), the Research Fund for the Doctoral Program of Higher Education of China (grant no. 20120001120129), 111 Project (grant no. B14001) and China Scholars hip Council Program (grant no. 201308110277). We appreciated EDGAR team for providing N2O emission data set, and Wei Gao, Qiong Chen in PKU, Chuan Wang, Shutan Ma, Changming Wang in ANU for collecting and compiling huge data sets.

■

REFERENCES

(1) Galloway, J. N.; Cowling, E. B. Reactive nitrogen and the world: 200 years of change. AMBIO 2002, 31 (2), 64−71. (2) World-Meteorological-Organization. The State of Greenhouse Gases in the Atmosphere Based on Global Observations through 2012; Atmospheric Environment Research Division: Geneva, 2013. H



Article

wheat yields to tillage method changes in the North China plain. PLoS One 2012, 7 (12), e51206 DOI: 10.1371/journal.pone.0051206. (20) Zhu, J.; Mulder, J.; Wu, L. P.; Meng, X. X.; Wang, Y. H.; Dorsch, P. Spatial and temporal variability of N2O emissions in a subtropical forest catchment in China. Biogeosciences 2013, 10 (3), 1309−1321. (21) Zou, J. W.; Huang, Y.; Jiang, J. Y.; Zheng, X. H.; Sass, R. L. A 3year field measurement of methane and nitrous oxide emissions from rice paddies in China: Effects of water regime, crop residue, and fertilizer application. Global Biogeochem. Cycles 2005, 19, GB2021 DOI: 10.1029/2004GB002401. (22) Zheng, X. H.; Han, S. H.; Huang, Y.; Wang, Y. S.; Wang, M. X. Re-quantifying the emission factors based on field measurements and estimating the direct N2O emission from Chinese croplands. Global Biogeochem. Cycles 2004, 18, GB2018 DOI: 10.1029/ 2003GB002167. (23) Corazza, M.; Bergamaschi, P.; Vermeulen, A. T.; Aalto, T.; Haszpra, L.; Meinhardt, F.; O’Doherty, S.; Thompson, R.; Moncrieff, J.; Popa, E.; Steinbacher, M.; Jordan, A.; Dlugokencky, E.; Bruhl, C.; Krol, M.; Dentener, F. Inverse modelling of European N2O emissions: Assimilating observations from different networks. Atmos. Chem. Phys. 2011, 11 (5), 2381−2398. (24) Klemedtsson, L.; von Arnold, K.; Weslien, P.; Gundersen, P. Soil C:N ratio as a scalar parameter to predict nitrous oxide emissions. Global Change Biol. 2005, 11 (7), 1142−1147. (25) Lilly, A.; Ball, B. C.; McTaggart, I. P.; DeGroote, J. Spatial modelling of nitrous oxide emissions at the national scale using soil, climate and land use information. Global Change Biol. 2009, 15 (9), 2321−2332. (26) Huang, X. D.; Grace, P.; Hu, W. B.; Rowlings, D.; Mengersen, K. Spatial prediction of N2O emissions in pasture: A Bayesian model averaging analysis. PLoS One 2013, 8 (6), e65039 DOI: 10.1371/ journal.pone.0065039. (27) Zhang, Y. X.; Tao, S.; Cao, J.; Coveney, R. M. Emission of polycyclic aromatic hydrocarbons in China by county. Environ. Sci. Technol. 2007, 41 (3), 683−687. (28) Wang, R.; Tao, S.; Wang, W. T.; Liu, J. F.; Shen, H. Z.; Shen, G. F.; Wang, B.; Liu, X. P.; Li, W.; Huang, Y.; Zhang, Y. Y.; Lu, Y.; Chen, H.; Chen, Y. C.; Wang, C.; Zhu, D.; Wang, X. L.; Li, B. G.; Liu, W. X.; Ma, J. M. Black carbon emissions in China from 1949 to 2050. Environ. Sci. Technol. 2012, 46 (14), 7595−7603. (29) Wang, R.; Tao, S.; Ciais, P.; Shen, H. Z.; Huang, Y.; Chen, H.; Shen, G. F.; Wang, B.; Li, W.; Zhang, Y. Y.; Lu, Y.; Zhu, D.; Chen, Y. C.; Liu, X. P.; Wang, W. T.; Wang, X. L.; Liu, W. X.; Li, B. G.; Piao, S. L. High-resolution mapping of combustion processes and implications for CO2 emissions. Atmos. Chem. Phys. 2013, 13 (10), 5189−5203. (30) Council of Agriculture. Yearly Report of Taiwan’s Agriculture; Department of Agriculture and Forestry of Taiwan Provincial Government: Taipei, Taiwan, 2008. (31) Hong Kong Census and Statistics Department. Hong Kong Annual Digest of Statistics; Hong Kong, 2008. (32) Macau Statistics and Census Service. Macau Demographic Statistics; Macau, 2008. (33) Huang, X.; Song, Y.; Li, M. M.; Li, J. F.; Huo, Q.; Cai, X. H.; Zhu, T.; Hu, M.; Zhang, H. S. A high-resolution ammonia emission inventory in China. Global Biogeochem. Cycles 2012, 26, GB1030 DOI: 10.1029/2011GB004161. (34) Piao, S. L.; Fang, J. Y.; Ciais, P.; Peylin, P.; Huang, Y.; Sitch, S.; Wang, T. The carbon balance of terrestrial ecosystems in China. Nature 2009, 458 (7241), 1009−U82. (35) Sitch, S.; Smith, B.; Prentice, I. C.; Arneth, A.; Bondeau, A.; Cramer, W.; Kaplan, J. O.; Levis, S.; Lucht, W.; Sykes, M. T.; Thonicke, K.; Venevsky, S. Evaluation of ecosystem dynamics, plant geography and terrestrial carbon cycling in the LPJ dynamic global vegetation model. Global Change Biol. 2003, 9 (2), 161−185. (36) Smith, B.; Prentice, I. C.; Sykes, M. T. Representation of vegetation dynamics in the modelling of terrestrial ecosystems: Comparing two contrasting approaches within European climate space. Global Ecol. Biogeogr. 2001, 10 (6), 621−637.

(37) FAO/IIASA/ISRIC/ISSCAS/JRC. Harmonized World Soil Database, (version 1.2); FAO, Rome, Italy and IIASA, Laxenburg, Austria, 2012. (38) Beijing Huajingzonghe Consultant Company. China Oxalic Acid Industry Research Report; Beijing, 2008. (39) SRI Consulting. Chemical Economics Handbook: Nitric Acid, 2006. (40) China industrial Newspaper Office. China Industrial Economic Statistical Yearbook; Beijing, 2009. (41) State Council of China. The First National Pollution Source Survey: Nutrient Loss Ratio of Agricultural Fertilizer (in Chinese). http://cpsc.mep.gov.cn/ (accessed November 12, 2013). (42) MEP/NSB/MOA. National Pollution Source Survey Database; Beijing, 2008. (43) Zhang, W. F.; Dou, Z. X.; He, P.; Ju, X. T.; Powlson, D.; Chadwick, D.; Norse, D.; Lu, Y. L.; Zhang, Y.; Wu, L.; Chen, X. P.; Cassman, K. G.; Zhang, F. S. New technologies reduce greenhouse gas emissions from nitrogenous fertilizer in China. Proc. Natl. Acad. Sci. U.S.A. 2013, 110 (21), 8375−8380. (44) Wu, X. W.; Zhu, F. H.; Yang, J. T.; Zhou, D. B.; L, Y.; Teng, N.; Yi, Y. P. Measurements of emission factors of greenhouse gas (CO2, N2O) from thermal power plants in China (in Chinese). Res. Environ. Sci. 2010, 23, 170−176. (45) Zhao, R. L.; Zhao, H. T.; Peng, M. S. The monitoring and estimated total volume of N2O from coal in China (in Chinese). Chin. J. Environ. Sci. 1996, 4, 11−13. (46) Yao, Z. L.; Wang, Z. D.; Wang, X. T.; Zhang, Y. Z.; Shen, X. B.; Yin, H.; He, K. B. An emission inventory of non-conventional pollutants from vehicles in typical cities (in Chinese). Environ. Pollut. Contr. 2011, 33, 96−101. (47) Ministry of Agriculture. China Animal Industry Yearbook; China Agricuture Press: Beijing, 2009. (48) Wang, R.; Tao, S.; Shen, H. Z.; Wang, X. L.; Li, B. G.; Shen, G. F.; Wang, B.; Li, W.; Liu, X. P.; Huang, Y.; Zhang, Y. Y.; Lu, Y.; Ouyang, H. L. Global emission of black carbon from motor vehicles from 1960 to 2006. Environ. Sci. Technol. 2012, 46 (2), 1278−1284. (49) Hu, X. D.; Wang, J. M. Estimation of livestock greenhouse gases discharge in China (in Chinese). Trans. Chin. Soc. Agric. Eng. 2010, 26 (10), 247−252. (50) Li, Y.; Lin, E. D. Emissions of N2O, NH3 and NOx from fuel combustion, industrial processes and the agricultural sectors in China. Nutr. Cycling Agroecosyst. 2000, 57 (1), 99−106. (51) Zhou, X.; Zheng, Y. F.; Wu, R. J.; Kang, N.; Zhou, W.; Yin, J. F. Greenhouse gas emissions from wastewater treatment in China during 2003−2009. Adv. Clim. Change Res. 2012, 8, 131−136. (52) IPCC. 2006 IPCC Guidelines for National Greenhouse Gas Inventories, National Greenhouse Gas Inventories Programme; IGES, Japan, 2006. (53) Liu, J. Y.; Zhang, Z. X.; Xu, X. L.; Kuang, W. H.; Zhou, W. C.; Zhang, S. W.; Li, R. D.; Yan, C. Z.; Yu, D. S.; Wu, S. X.; Nan, J. Spatial patterns and driving forces of land use change in China during the early 21st century. J. Geogr. Sci. 2010, 20 (4), 483−494. (54) Fragkias, M.; Lobo, J.; Strumsky, D.; Seto, K. C., Does size matter? Scaling of CO2 emissions and U.S. urban areas. PLoS One 2013, 8(6), DOI: 10.1371/journal.pone.0064727. (55) Golubiewski, N. Is there a metabolism of an urban ecosystem? An ecological critique. Ambio 2012, 41 (7), 751−764. (56) Reay, D. S.; Davidson, E. A.; Smith, K. A.; Smith, P.; Melillo, J. M.; Dentener, F.; Crutzen, P. J. Global agriculture and nitrous oxide emissions. Nat. Clim Change 2012, 2 (6), 410−416. (57) Stehfest, E.; Bouwman, L.; van Vuuren, D. P.; den Elzen, M. G. J.; Eickhout, B.; Kabat, P. Climate benefits of changing diet. Clim. Change 2009, 95 (1−2), 83−102. (58) Zhang, F.; Cui, Z.; Fan, M.; Zhang, W.; Chen, X.; Jiang, R. Integrated soil-crop system management: Reducing environmental risk while increasing crop productivity and improving nutrient use efficiency in China. J. Environ. Qual. 2011, 40, 1051−1057. (59) Saikawa, E.; Prinn, R. G.; Dlugokencky, E.; Ishijima, K.; Dutton, G. S.; Hall, B. D.; Langenfelds, R.; Tohjima, Y.; Machida, T.; Manizza, I



Article

M.; Rigby, M.; O’Doherty, S.; Patra, P. K.; Harth, C. M.; Weiss, R. F.; Krummel, P. B.; van der Schoot, M.; Fraser, P. B.; Steele, L. P.; Aoki, S.; Nakazawa, T.; Elkins, J. W. Global and regional emissions estimates for N2O. Atmos. Chem. Phys. Discuss. 2013, 13, 19471−19525. (60) Thompson, R. L.; Chevallier, F.; Crotwell, A. M.; Dutton, G.; Langenfelds, R. L.; Prinn, R. G.; Weiss, R. F.; Tohjima, Y.; Nakazawa, T.; Krummel, P. B.; Steele, L. P.; Fraser, P.; Ishijima, K.; Aoki, S. Nitrous oxide emissions 1999−2009 from a global atmospheric inversion. Atmos Chem. Phys. 2014, 14, 1801−1817. (61) Tian, H. Q.; Xu, X. F.; Lu, C. Q.; Liu, M. L.; Ren, W.; Chen, G. S.; Melillo, J.; Liu, J. Y. Net exchanges of CO2, CH4, and N2O between China’s terrestrial ecosystems and the atmosphere and their contributions to global climate warming. J. Geophys. Res.: Biogeosci. 2011, 116, G02011 DOI: 10.1029/2010JG001393. (62) Boisier, J. P.; de Noblet-Ducoudré, N.; Ciais, P. Historical landuse induced evapotranspiration changes estimated from present-day observations and reconstructed land-cover maps. Hydrol. Earth Syst. Sci. Discuss. 2014, 11, 2045−2089. (63) Li, C. S.; Frolking, S.; Xiao, X. M.; Moore, B.; Boles, S.; Qiu, J. J.; Huang, Y.; Salas, W.; Sass, R. Modeling impacts of farming management alternatives on CO2, CH4, and N2O emissions: A case study for water management of rice agriculture of China. Global Biogeochem. Cycles 2005, 19 (3), GB3010 DOI: 10.1029/ 2004GB002341. (64) Bouwman, A. F.; Beusen, A. H. W.; Griffioen, J.; Van Groenigen, J. W.; Hefting, M. M.; Oenema, O.; Van Puijenbroek, P. J. T. M.; Seitzinger, S.; Slomp, C. P.; Stehfest, E., Global trends and uncertainties in terrestrial denitrification and N2O emissions. Philos. Trans. R. Soc., B 2013, 368(1621), DOI: 10.1098/rstb.2013.0112. (65) Butterbach-Bahl, K.; Baggs, E. M.; Dannenmann, M.; Kiese, R.; Zechmeister-Boltenstern, S., Nitrous oxide emissions from soils: How well do we understand the processes and their controls? Philos. Trans. R. Soc., B 2013, 368(1621), DOI: 10.1098/rstb.2013.0122. (66) Gu, J. X.; Zheng, X. H.; Zhang, W. Background nitrous oxide emissions from croplands in China in the year 2000. Plant Soil 2009, 320 (1−2), 307−320.

J


A New High-Resolution N2O Emission Inventory for China in 2008

Recommend Documents