オープンパス型とクローズドパス型の渦相関法によるＣＯ２フラックスの系統的差異と密度変動補正の影響

Abstract

Since both open- and closed-path eddy covariance techniques have been used in ecosystem observation networks such as AsiaFlux, estimating the random and systematic differences in CO2 fluxes observed by the two techniques is important. In the present study, CO2 fluxes measured by open- and closed-path eddy covariance systems were compared using year-round data obtained in a single-cropping rice paddy and in a larch forest. The two sites were similar in eddy covariance instrumentation and data processing, but dissimilar in canopy height and structure, phenology, and climate. For both sites, differences in half-hourly CO2 fluxes measured by the open- and closed-path systems followed a Laplace distribution, but were significantly biased, resulting in the open-path system measuring lower values than the closed-path system and in a larger carbon sequestration by vegetation being estimated with the open-path system. The systematic difference between the two systems mainly arose from an inconsistency between the open-path CO2 flux and the WPL correction for the effect of air density fluctuations, and increased with the magnitude of the WPL correction. During the growing season, the systematic difference at the paddy site was consistently smaller than that at the larch site because of the much low Bowen ratio and the resulting small WPL correction term caused by standing water in the paddy field. The WPL correction term was large and thus, the systematic difference became apparent in late winter because of the high Bowen ratios at the paddy and larch site, each covered with dried soil and snowpack, respectively. The year-round, half-hourly data accumulated after quality control tests showed that, between the two systems, relative differences existed of 9.9% (N=6074) for the paddy site and 15.8% (N=4591) for the larch site.