Simulation Study of Grain Yield and N₂O Emissions under Farmland Management Measures and Climate Change Based on DNDC Model: Case Study of Hilly Regions in Sichuan Province

It is crucial to propose suitable farmland management measures for addressing food security and ecological challenges caused by climate change. In this study, a typical farmland located in Jiangyou City of Sichuan Province was selected as the study site, which is characterized by hilly geographical feature. Based on field investigation and sample analysis, the DNDC model was firstly used to simulate the grain yield and N₂O emissions of the Wheat-Maize planting system under four different farmland management measures related to nitrogen fertilizer application rate, fertilization depth, tillage depth, and application of inorganic-organic fertilizers. Subsequently, the optimal farmland management measure was determined, and then integrated to future climate change scenarios (RCP4.5 and RCP8.5) to reveal their impacts on grain yield and N₂O emission. The detailed results show that: (1) During the baseline scenario, the R² values between the simulated and observed yields of maize and wheat were 0.59 and 0.62, respectively With the MRE to be -1.63% and -3.19%, and NRMSE to be 7.32% and 3.60%, respectively. These indexes indicate that the simulated results of DNDC model were credible and can be applied to next scenario simulation. (2) The optimal scenarios for various farmland management measures include: the conventional fertilizer application rate of 70% (F3), a fertilization depth of 12cm (D3), a tillage depth of 20cm (T3), and the application of inorganic-organic fertilizers of 25% (M1). The M1 is the best single farmland management measure scenario, while the best combined farmland management measure scenario is D3M1, with yields of 9 802.63 kg·hm^-2·a^-1 and 9 804.83 kg·hm^-2·a^-1, respectively. (3) Under future climate change scenarios, N₂O emissions in the D3M1 combined farmland management measure scenario decreased by 4.87% to 17.51% and 33.62% to 36.76% compared with the M1 and BS single farmland management measure scenario, respectively, while the yield did not show significant differences. Therefore, the D3M1 is determined as the optimal farmland management measure. In the long-term of D3M1-RCP4.5 and D3M1-RCP8.5 climate change scenarios, the maize yields respectively decreased by 0.03% and 0.07% compared with historical climate period, this is primarily attributed to increasing temperatures during the growing season and higher precipitation before seed production. The increased precipitation during the wheat growing season resulted in yield reductions of 1.31% and 2.72%, with rising temperature potentially accelerated maturity. Meanwhile, the increased precipitation under future climate change scenario led to a consistent rise in N₂O emissions. This research indicates that farmland management measures can be adopted to guarantee food security and mitigate the effects of climate change in the future, for example, using the combined farmland management measures (a fertilization depth of 12cm and the application of inorganic-organic fertilizers of 25%, D3M1), reducing the fertilizer application rate, adjusting crop seeding and fertilization timing appropriately, and so on.