Net carbon sink of China was overestimated by more than 35%


 As a carbon source/sink of atmospheric carbon dioxide, the net regional carbon budget (NRCB) of terrestrial ecosystems is very important to effect global warming, especially China with the largest emissions at present. However, the carbon consumption is difficult to measure accurately, which is caused by the emissions of CH4 and CO, the utilization of agriculture, forestry and grass, and the emissions from rivers and other physical processes, such as forest fires. Therefore, the spatial patterns and driving factors of NRCB are not clear. Here, we used multi-source data to estimate the NRCB of 31 provincial administrative divisions of China and to develop NRCB datasets from 2000 to 2018. We found that the average of NRCB was 669 TgC yr−1, and it significantly decreased at a rate of 2.56 TgC yr−1. The relative contribution rates of fluxes of emissions from anthropogenic (FEAD), reactive carbon and creature ingestion (FERCCI), autotrophic respiration (Ra), heterotrophic respiration (Rh) and natural disturbances (FEND) were 35.17%, 26.09%, 19.68%, 17.38% and 1.68% respectively. In addition, NRCB datasets of the different administrative regions of China were mapped. These datasets will provide support for China's carbon neutrality and the study of the global carbon cycle.

the NRCB of 31 provincial administrative divisions of China and to develop NRCB datasets from 23 2000 to 2018. We found that the average of NRCB was 669 TgC yr −1 , and it significantly 24 decreased at a rate of 2.56 TgC yr −1 . The relative contribution rates of fluxes of emissions from 25 anthropogenic (FEAD), reactive carbon and creature ingestion (FERCCI), autotrophic respiration 26 (Ra), heterotrophic respiration (Rh) and natural disturbances (FEND) were 35.17%, 26.09%, 27 19.68%, 17.38% and 1.68% respectively. In addition, NRCB datasets of the different 28 administrative regions of China were mapped. These datasets will provide support for China's 29 carbon neutrality and the study of the global carbon cycle. 30

31
The main sources of carbon in terrestrial ecosystems are aboveground and underground 32 biomass, soil, and dead organic matter 1 grasp the dynamics of the NRCB, based on the relationship between the gross primary 49 productivity (GPP) and the various types of carbon consumption, such as natural and 50 anthropogenic, a theoretical framework was constructed to quantify regional ecosystem 51 productivity, the composition of the carbon budget, and the spatiotemporal pattern of the carbon 52 sources/sinks 10 . Promoting the rapid development of net ecosystem productivity (NEP) and NRCB 53 measurement and evaluation. For example, in this study, it was found that China's NEP decreases 54 from south to north and from southwest to northeast 11 , and the average annual NEP is 1.91 PgC 55 yr −1 12 . 56 China's NRCB has been estimated using methods such as atmospheric retrieval, resource 57 inventory, and vortex covariance 13,14 . However, the carbon consumption caused by CH4 and CO 58 emissions, carbon consumption caused by the utilization of agriculture, forestry and grass (such as 59 food, fuel wood, grazing, livestock raising, etc.), and carbon consumption from rivers and other 60 physical processes such as forest fires are difficult to measure accurately, few studies have 61 quantitatively analyzed the reserves of the NRCB in the various administrative regions in China 62 and their driving factors. In addition, the uncertainties of the results of various studies are still high. 63 For example, China's NRCB is 310-970 TgC yr −1 (Supplementary Table.

1). A systematic and 64
accurate dataset is the key to improving the accuracy of NRCB calculations on regional and global 65 scales. 66 Integrated multi-source data for the ecosystem and eddy covariance measurements were used 67 to quantitatively evaluate the response mechanism of China's NRCB to natural and anthropogenic 68 processes from 2000 to 2018. The purpose of this study was to develop an NRCB dataset for 69 China. To analyze the spatial and temporal evolution of the NRCB. And to determine the driving 70 factors of the spatiotemporal changes in the NRCB. 71

Results 72
Spatial pattern of the NRCB 73 In this study, it was found that from 2000 to 2018, the total NRCB in China was 12.71 PgC, 74 and the average annual NRCB was about 669 TgC, with significant heterogeneity of its spatial 75 distribution (Fig.1). Among the 31 administrative units, the high-value areas were mainly 76 distributed in southwestern and southern China, particularly Yunnan (161.6 TgC yr −1 ) and 77 Guangxi (93.5 TgC yr −1 ), accounting for 24.28% and 13.98% of the NRCB, respectively. In 78 contrast, the low-value areas were mainly concentrated in northern, central, eastern, and 79 northwestern China, and Tibet, which is located in southwestern China, had the lowest NRCB per 80 unit area (−52.17 TgC yr −1 ). This is mainly due to the low latitudes, strong solar radiation, and 81 abundant sunshine in southwestern and southern China, which has a subtropical monsoon climate, 82 and the fact that the temperature, rainfall, and evapotranspiration are higher in the southeast and 83 lower in the northwest 15 . The total GPP due to photosynthesis is large, but owing to the high 84 altitude and low precipitation in Tibet, the total GPP is relatively small. In addition, land 85 management in southern China offsets 33% of the carbon equivalent of regional fossil fuel CO2 Temporally, the overall NRCB decreased at a rate of 2.56 TgC yr −1 , with the largest decreases 104 occurring in northeast, central, eastern, southwestern, and southern China (Fig.2). Among them, 105 the NRCB decreased the most in Heilongjiang, and its rate of decrease has reached 1.08 TgC yr −1 . 106 However, the NRCB has been increasing in the inland areas of northwestern China and parts of 107 northern and southwestern China. Tibet is the region with the largest increase in the NRCB, with 108 an increase rate of 1.76 TgC yr −1 . This is mainly due to the implementation of ecological projects, 109 the prevention of the degradation of grasslands and woodlands in northwestern inland areas such 110 as Xinjiang and Tibet and in northern China, and the increasing restoration of regional ecosystems, 111 which have reduced the carbon consumption of key natural and anthropogenic activities. In 112 contrast, in most areas, including northeastern, central, eastern, southwestern, and southern China, 113 due to the development of the local society and economy, the effect of human activities on the 114 surface cover has increased, leading to an increase in the carbon consumption of key natural and 115 anthropogenic activities. Based on our research, we found that the FERCCI has increased 116 significantly in parts of northern, central, and southwestern China, but the rate of increase was 117 very small ( Supplementary Fig.1). The total increase was 3 TgC in 19 years (Supplementary 118  Fig.1), and the regions with the largest increases were mainly located in 130 northeastern, central, eastern, southwestern, and parts of southern China (Supplementary Fig.2). 131

The increase in the FEAD was the main driving force behind the decrease in the NRCB 132
Since 2000, the significant increase of FEAD has contributed the most to the decrease of 133 NRCB (35.17%), especially in northeast China, North China and central China, with a relative 134 contribution rate of more than30% (Fig.3). This increase in the FEAD is attributed to the large-scale 135 agricultural production on the plains, which, while producing high yields, continuously fertilizes 136 the land with pesticides, changes the soil structure, and damages the integrity of the ecosystem, 137 thus increasing the carbon consumption of the land. We found that 81.58% of the FEAD (742 TgC 138 yr −1 ) was carbon consumption caused by the use of agricultural products. In addition, the 139 contribution of the FERCCI to the change in the NRCB is also very important, reaching 26.09%. In 140 terms of the FERCCI (76 TgC yr −1 ), 50.25% of the carbon consumption was from the FECO. This 141 phenomenon was mainly concentrated in northeastern, Central, southern, and southwestern China. 142 Of course, the NRCB is also affected by other factors. Among them, the relative contribution rate 143 of the Ra was 19.68%, the relative contribution rate of the Rh was 17.38%, and the relative 144 contribution rate of the FEND was only 1.68%. Although it is less than those of the first two, the 145 relative contribution rate of the FEND also has a very important impact on the NRCB. 146

147
At present, the methods of assessing the key components of the terrestrial ecosystem carbon 148 budget mainly include the resource inventory method, the process-based ecological model or 149 remote sensing model method, and the vortex covariance method 23,24 . Based on the vortex 150 covariance method, we evaluated the main components of the carbon budget of the terrestrial 151 ecosystem in China (Fig.4)