Numerous studies on methane emission from rice paddies have been conducted since 1980s, but most of them employed chamber techniques. To measure spatially averaged methane fluxes under non-disturbed conditions, micrometeorological techniques were applied in several pioneering studies at rice paddies, but the periods of those studies were restricted to short terms. In this paper, we present long-term measurements of methane emission from a customarily cultivated Japanese paddy using micrometeorological techniques. The measurements cover almost entire rice cropping seasons from 1993 to 1995. The soil type of the paddy was mainly of peat dressed with lowland gray soil, and mineral fertilizers were amended to the paddy before transplanting. The methane fluxes were determined from the vertical gradient of methane concentration over rice canopy multiplied by the eddy diffusivity. The methane concentration was measured on line with a non-dispersive infrared gas analyzer equipped with specially designed preconditioning part for removing the disturbance of water vapor and non-methane hydrocarbons. The analyzer attained detection limit of 5 ppbv on 30-minute mean basis under field conditions. The eddy diffusivity was evaluated by an improved aerodynamic method which used friction velocity and atmospheric stability determined by the eddy correlation method. The CO2 fluxes over the rice paddy by the aerodynamic method were in good agreement with those by the eddy correlation method (the difference was 0.02+/-0.13 mgCO2 m-2 s-1) except for the conditions with the friction velocity less than 0.1 m s-1. The result demonstrates the aerodynamic method is suitable for the gas flux measurement over the rice paddy except for extremely week wind conditions.
The methane flux showed a distinct diurnal variation with a sharp peak in the early afternoon and a wide trough from midnight to early morning. The daily mean methane flux, ranging from 30 to 370 mg CH4 m-2 day-1, showed seasonal variation which was related to the changes in soil temperature and flooding conditions. The methane flux increased with days after flooding until the midsummer drainage (MD) followed by a temporal decline during MD. The flux recovered after MD, and reached a high level around the heading period of rice plants under high soil temperatures. The flux showed a short-term remarkable increase right after a drainage, then decreased. The daily mean methane fluxes during continuously flooded periods showed an exponential increase with the daily mean soil temperature at a depth of 5 cm. The values of Q10 in the three years ranged between 3.3 and 4.7, which were close to the values found in the previous chamber-employed studies. The seasonal amounts of methane emission ranged from 8.7 (in 1993, cool summer) to 17.7 gCH4 m-2 (in 1995, hot summer), and the inter-annual differences in the seasonal amounts indicated the significant influences of flooding depth and the drainage efficiency on the methane emission from the paddy. The results presented in this paper thus show the micrometeorological method employed in this study is quite applicable to the measurement of spatially averaged methane fluxes from intense methane sources such as rice paddies and natural wetlands.