A GIS-based method for modeling methane emissions from paddy fields by fusing multiple sources of data

Abstract

Quantification of regional methane (CH₄) gas emission in the paddy fields is critical under climate warming. Mechanism models generally require numerous parameters while empirical models are too coarse. Based on the mechanism and structure of the widely used model CH4MOD, a GIS-based Regional CH₄ Emission Calculation (GRMC) method was put forward by introducing multiple sources of remote sensing images, including MOD09A1, MOD11A2, MOD15A2H as well as local water management standards. The stress of soil moisture condition (f(water)) on CH₄ emissions was quantified by calculating the redox potential (Eh) from days after flooding or falling dry. The f(water)-t curve was calculated under different exogenous organic matter addition. Combining the f(water)-t curve with local water management standards, the seasonal variation of f(water) was obtained. It was proven that f(water) was effective in reflecting the regulation role of soil moisture condition. The GRMC was tested at four Eddy Covariance (EC) sites: Nanchang (NC) in China, Twitchell (TWT) in the USA, Castellaro (CAS) in Italy and Cheorwon (CRK) in Korea and has been proven to well track the seasonal dynamics of CH₄ emissions with R² ranges of 0.738-0.848, RMSE ranges of 31.94-149.22 mg C/m²d and MBE ranges of -66.42- -14.79 mg C/m²d. The parameters obtained in Nanchang (NC) site in China were then applied to the Ganfu Plain Irrigation System (GFPIS), a typical rice planting area of China, to analyse the spatial-temporal variations of CH₄ emissions. The total CH₄ emissions of late rice in the GFPIS from 2001 to 2013 was in the range of 14.47-20.48 (10³ t CH₄-C). Ts caused spatial variation of CH₄ production capacity, resulting in the spatial variability of CH4 emissions. Overall, the GRMC is effective in obtaining CH₄ emissions from rice fields on a regional scale.

Keywords: CH(4) flux; Paddy fields; Remote sensing; Semi-empirical model.