Analysis of methane emissions from paddy rice using Bayesian assimilation

In this work, we have explored a new method which based on Bayesian assimilation for monitoring methane (CH 4 ) emissions at different rice phenological stages. Specifically, we investigate two algorithms, one based on the ground-based radar scatterometer (GBRS) with full polarization (HH, HV, VH, and VV) and the other based on the mechanistic process of agricultural model concluded the parameters of climate, soil and management. Experiments are conducted on conventional static box and results are compared to the outputs of Bayesian assimilation, microwave model, and Denitrification-Decomposition (DNDC) model, respectively. The result shows that the fusion has combined the tendency of mechanistic model and accuracy of microwave model. For the truth and authenticity of data and result, three criteria indicate the result with high accuracy not only agree well with the sample value but also provide high reliability on the real methane emissions process over time and space.