漂浮通量箱法和扩散模型法测定内陆水体CH_4和N_2O排放通量的初步比较研究

Preliminary Comparison of the Static Floating Chamber and the Diffusion Model Methods for Measuring Water-Atmosphere Exchanges of Methane and Nitrous Oxide from Inland Water Bodies

采用漂浮通量箱法和扩散模型法同步地观测了模拟内陆水体在不同条件下的 CH4和 N2O 的水-气交换通量，旨在比较两类方法取得结果的异同。结果显示：这两类方法所测得的绝大多数CH4排放通量都与水中溶解氧呈显著线性负相关（显著性系数P＜0.001）。同时N2O排放通量与表层水温及水中铵态氮、硝态氮、溶解碳和溶解氧的关系可用包含所有上述水环境因素的 Arrhenius 动力学方程来表达，这些因素可以共同解释86%～90%的N2O通量变化（P＜0.0001），且不同方法测定的N2O通量的表观活化能和对表层水温的敏感系数分别介于47～59 kJ mol^-1和1.92～2.27之间；扩散模型法所获得的CH4和N2O通量分别是箱法测定值的13%～175%和15%～240％，差异程度因模型而异；不同模型取得通量间相差20%～1200％，平均相差2.3倍。上述结果表明：仅用一种模型方法来取得CH4或N2O排放通量易形成较大偏差；不同扩散模型法和箱法测定的通量在反映CH4和N2O排放的内在规律方面具有一致性，但它们对真实气体通量的测量是否都存在不同程度的系统误差，尚需进一步研究。

The static floating chamber technique and the diffusion model approach are the two methods most commonly used for quantifying methane （CH4） and nitrous oxide （N2O） fluxes from water bodies. However, the comparability of these two methods for measuring gas fluxes remains unclear. In this study, the water-atmosphere exchange fluxes of CH4 and N2O were simultaneously determined using both methods under variable conditions to investigate their comparability.The results showed that most of the CH4 fluxes determined by either method negatively and linearly correlation with the dissolved oxygen content in the water （significance coefficient P〈0.001）. Meanwhile, the N2O flux results from either method could be expressed as a compound function of water temperature, ammonia, nitrate, dissolved carbon, and oxygen following Arrhenius kinetics, by which 86%-90%of the variances in the N2O fluxes can be jointly explained by these five factors （P〈0.0001）. The fitting functions provided an apparent activation energy and a temperature sensitivity coefficient （Q10） of 47-59 kJ mol^-1 and 1.92-2.27, respectively. The CH4 and N2O fluxes measured using the diffusion model method went from 13%to 1.75 times and from 15%to 2.4 times the results from the chamber method, respectively. The magnitudes of the differences varied between the adopted methods. The differences in the gas fluxes between the two methods ranged from 20%to a 12-fold increase, with a mean of a 2.3-fold increase. Our results suggest that applying a single method exclusively to estimate the CH4 or N2O fluxes may inevitably yield huge errors, although the flux results from both methods can consistently reflect the processes of gas production kinetics. Our results also suggest that further studies are still needed to verify and quantify the systematic errors which are likely being caused by either the chamber technique or the diffusion model approach.