China has experienced rapid agricultural development over recent decades, accompanied by increased fertilizer consumption in croplands; yet, the trend and drivers of the associated nitrous oxide (N2O) emissions remain uncertain. The primary sources of this uncertainty are the coarse spatial variation of activity data and the incomplete model representation of N2O emissions in response to agricultural management. Here, we provide new data‐driven estimates of cropland‐N2O emissions across China in 1990–2014, compiled using a global cropland‐N2O flux observation dataset, nationwide survey‐based reconstruction of N‐fertilization and irrigation, and an updated nonlinear model. In addition, we have evaluated the drivers behind changing cropland‐N2O patterns using an index decomposition analysis approach. We find that China's annual cropland‐N2O emissions increased on average by 11.2 Gg N/year2 (p < .001) from 1990 to 2003, after which emissions plateaued until 2014 (2.8 Gg N/year2, p = .02), consistent with the output from an ensemble of process‐based terrestrial biosphere models. The slowdown of the increase in cropland‐N2O emissions after 2003 was pervasive across two thirds of China's sowing areas. This change was mainly driven by the nationwide reduction in N‐fertilizer applied per area, partially due to the prevalence of nationwide technological adoptions. This reduction has almost offset the N2O emissions induced by policy‐driven expansion of sowing areas, particularly in the Northeast Plain and the lower Yangtze River Basin. Our results underline the importance of high‐resolution activity data and adoption of nonlinear model of N2O emission for capturing cropland‐N2O emission changes. Improving the representation of policy interventions is also recommended for future projections.

No abstract

No abstract

Drained peatlands are hotspots for greenhouse gas (GHG) emissions, which could be mitigated by rewetting and land use change. We performed an ecological/economic analysis of rewetting drained fertile peatlands in a hemiboreal climate using different land use strategies over 80 years. Vegetation, soil processes, and total GHG emissions were modeled using the CoupModel for four scenarios: (1) business as usual-Norway spruce with average soil water table of -40 cm; (2) willow with groundwater at -20 cm; (3) reed canary grass with groundwater at -10 cm; and (4) a fully rewetted peatland. The predictions were based on previous model calibrations with several high-resolution datasets consisting of water, heat, carbon, and nitrogen cycling. Spruce growth was calibrated by tree-ring data that extended the time period covered. The GHG balance of four scenarios, including vegetation and soil, were 4.7, 7.1, 9.1, and 6.2 Mg CO eq ha year , respectively. The total soil emissions (including litter and peat respiration CO + N O + CH ) were 33.1, 19.3, 15.3, and 11.0 Mg CO eq ha year , respectively, of which the peat loss contributed 35%, 24%, and 7% of the soil emissions for the three drained scenarios, respectively. No peat was lost for the wet peatland. It was also found that draining increases vegetation growth, but not as drastically as peat respiration does. The cost-benefit analysis (CBA) is sensitive to time frame, discount rate, and carbon price. Our results indicate that the net benefit was greater with a somewhat higher soil water table and when the peatland was vegetated with willow and reed canary grass (Scenarios 2 and 3). We conclude that saving peat and avoiding methane release using fairly wet conditions can significantly reduce GHG emissions, and that this strategy should be considered for land use planning and policy-making.

Abstract. The CoupModel was used to simulate a Norway spruce forest on fertile drained peat over 60 years, from planting in 1951 until 2011, describing abiotic, biotic and greenhouse gas (GHG) emissions (CO2 and N2O). By calibrating the model against tree ring data a “vegetation fitted” model was obtained by which we were able to describe the fluxes and controlling factors over the 60 years. We discuss some conceptual issues relevant to improving the model in order to better understand peat soil simulations. However, the present model was able to describe the most important ecosystem dynamics such as the plant biomass development and GHG emissions. The GHG fluxes are composed of two important quantities, the spruce forest carbon (C) uptake, 413 g C m−2 yr−1 and the decomposition of peat soil, 399 g C m−2 yr−1. N2O emissions contribute to the GHG emissions by up to 0.7 g N m−2 yr−1, corresponding to 76 g C m−2 yr−1. The 60-year old spruce forest has an accumulated biomass of 16.0 kg C m−2 (corresponding to 60 kg CO2 m−2). However, over this period, 26.4 kg C m−2 (97 kg CO2eq m−2) has been added to the atmosphere, as both CO2 and N2O originating from the peat soil and, indirectly, from forest thinning products, which we assume have a short lifetime. We conclude that after harvest at an age of 80 years, most of the stored biomass carbon is liable to be released, the system having captured C only temporarily and with a cost of disappeared peat, adding CO2 to the atmosphere.