为了探索土壤剖面CO2浓度以及不同土壤层(腐殖质H层、A层、B层、C层)土壤呼吸的变化规律,应用非分散红外(NDIR)技术的新方法,持续不间断的测量土壤剖面二氧化碳浓度.实验所用的主要仪器为硅基非分散红外测量仪,能在高湿、高粉尘、污垢及其他恶劣环境中进行光谱数据采集.通过2013年全年光谱测定值的采集,并应用梯度法模型计算不同深度土壤碳通量,同时利用LI-8100碳通量自动监测系统持续监测的土壤碳通量值进行回归分析.结果显示:土壤剖面CO2浓度呈现明显的梯度变化,即随着土壤深度的增加,土壤CO2浓度增大;梯度法模型得出的不同土壤层的土壤呼吸模拟值与实测土壤呼吸值之间具有较好的线性相关,H,A,B,C层的模型预测的决定系数(R2)分别为0.906　9,0.718　5,0.838　2,0.903　0,均方根误差(RMSE)分别为0.206　7,0.104　1,0.015　6,0.009　6.均达到了较好的预测结果,表明该方法对定量分析不同土壤层碳通量是可行的.该方法具有清晰揭示土壤CO2在不同土壤层之间的传输规律,以及有助于分析不同土壤层土壤呼吸特性的优点,能为全球土壤剖面碳通量计算提供基础数据,是一种具有发展前途的传感器。

In order to explore the variation of CO2 concentration and soil respiration in soil profile,the nondispersive infrared (NDIR) spectroscopy technique was applied to continually estimate the soil CO2 concentration in different soil layers (the humus horizon,A-,B-,C-horizon) in situ.The main instrument used in this experiment was silicon-based nondispersive infrared sensor,which could work in severe environment.We collected the measurement value by NDIR spectroscopy technique throughout 2013.The values of soil carbon flux in different soil layers were calculated based on the model of gradient method and calibrated by measuring with an automated soil CO2 efflux system (LI-8100).The results showed that:a vertical gradient for the carbon dioxide concentration in soil profile was found,and the concentration was highest in the deepest soil horizon.Moreover,A linear correlation between the soil CO2 effluxes was calculated based on model and measurement,and the model prediction correlation coefficient was 0.906 9,0.718 5,0.838 2,and 0.903 0 in the H-,A-,B-,and C-horizon,respectively.The roots of mean square error (RMSE) were 0.206 7,0.104 1,0.015 6,and 0.009 6 in the H-,A-,B-,and C-horizon,respectively.These results suggest that the gradient method based on the NDIR spectroscopy technique can be successfully used to measure soil CO2 efflux in different soil layers,which reveal that diffusion and convection transport CO2 between the soil layers.It is a promising sensor for detecting CO2 concentration in soil profile,providing the basic data for calculating the global carbon in soil profile.