DNDC模型模拟不同施肥水平下旱作麦田土壤N2O排放及其敏感性分析

Simulation and sensitivity analysis of N2O emissions in dryland wheat fields at different fertilization levels using the DNDC model

【目的】研究陇中黄土高原不同施肥水平下旱作麦田土壤N_(2)O的排放特征,验证DNDC模型对N_(2)O排放模拟的适应性。【方法】以甘肃省定西市旱作麦田为对象,设置不施氮肥(0 kg/hm^(2))、低量氮肥(55 kg/hm^(2))、中量氮肥(110 kg/hm^(2))、高量氮肥(220 kg/hm^(2))4种施肥梯度,采用静态箱-气相色谱法对土壤N_(2)O排放通量排放进行田间原位观测。结合试验点连续监测得到的N_(2)O排放通量等实测数据,运用DNDC模型探究多因子对模拟N_(2)O排放的敏感程度。【结果】不同施肥水平下,旱作麦田土壤N_(2)O排放趋势基本一致,且排放通量的大小与施氮量成正相关关系;DNDC模型对4种施肥水平下N_(2)O排放的模拟效果较好,具体表现为不施氮肥(R^(2)=0.88、EF=0.84、P=0.002),低量氮肥(R^(2)=0.92、EF=0.82、P=0.001),中量氮肥(R^(2)=0.86、EF=0.81、P=0.003),高量氮肥(R^(2)=0.84、EF=0.74、P=0.004),表明DNDC模型能够用于模拟不同施肥水平下旱作麦田土壤N_(2)O排放;敏感性分析则显示,相比其他因子,土壤pH值是影响模型模拟N_(2)O排放最为敏感的参数。【结论】为陇中黄土高原旱作麦田N_(2)O排放模型模拟的参数本地化提供基础依据,为今后农田管理措施的优化提供模型基础技术支撑。

【Objective】To study the emission characteristics of soil N_(2)O and verify the adaptability of the DNDC model to N_(2)O emission under different fertilization levels on the Loess Plateau of Longzhong.【Method】Wheat fields in Dingxi City,Gansu Province,were set up with no nitrogen fertilizer(0 kg/hm^(2)),low nitrogen fertilizer(55 kg/hm^(2)),medium nitrogen fertilizer(110 kg/hm^(2))and high nitrogen fertilizer(220 kg/hm^(2)).Soil N_(2)O emission flux was measured in situ by static box gas chromatography.Combined with the measured data of N_(2)O emission flux continuously monitored at the test site,the DNDC model was used to investigate the sensitivity of several factors to simulated N_(2)O emission.【Result】Under different fertilization levels,the trend of soil N_(2)O emission was basically the same,and the emission flux was positively correlated with the nitrogen application rate.The DNDC model has a good simulation effect on N_(2)O emission under four fertilization levels,which is manifested as no nitrogen fertilizer(R^(2)=0.88,EF=0.84,P=0.002),low nitrogen fertilizer(R^(2)=0.92,EF=0.82,P=0.001),and low nitrogen fertilizer(R^(2)=0.92,EF=0.82,P=0.001).Medium nitrogen(R^(2)=0.86,EF=0.81,P=0.003)and high nitrogen(R^(2)=0.84,EF=0.74,P=0.004)showed that the DNDC model could be used to simulate soil N_(2)O emission under different fertilization levels.Sensitivity analysis showed that compared with other factors,soil pH value was the most sensitive parameter to affect N_(2)O emission from farmland in the study area.【Conclusion】This study can provide the theoretical basis and technical support for local application of the regional N_(2)O emission model and optimization of farmland management measures.