摘要 : The large temporal variation in nitrous oxide (N2O). methane (CH4) and carbon dioxide (CO2) flux rates is a major source of error when estimating cumulative fluxes of these radiative active trace gases. We developed an automated s... 展开 The large temporal variation in nitrous oxide (N2O). methane (CH4) and carbon dioxide (CO2) flux rates is a major source of error when estimating cumulative fluxes of these radiative active trace gases. We developed an automated system for near-continuous, long-term measurements of N2O, CH4 and CO2 fluxes from cropland soils and used it to study the temporal variation of N2O and CH4 fluxes from potato (Solanum tuberosum L.) fields during the crop periods of 1997 and 1998, and also to determine the effects of management practices and weather. Additionally, we evaluated the error of other common methods, namely, weekly or monthly measurements, used for estimating cumulative fluxes. ne fluxes were quantified separately for the ridges, uncompacted interrows and tractor-compacted interrows. Total N2O-N emission from the potato field during the growing period (end of May to September) was 1.6 kg ha(-1) in 1997 and 2.0 kg ha(-1) in 1998; emissions were highest for the tractor-compacted soil. Periods of increased N2O losses were induced by heavy precipitation (in particular in compacted soil) and by the killing of potato tops (on the ridges) by herbicide application. The total CH4-C uptake in the potato field during the growing period was 295 g ha(-1) in 1997 and 317 g ha(-1) in 1998. The major fraction of the total CH, uptake (approximate to 86%) occurred on the ridges. Weekly measurements of N2O fluxes complemented by additional event-related flux determinations provided accurate estimates of total emissions. The monthly flux determination was not adequate for determining the temporal variation of the N2O emission rates. Weekly measurements were sufficient to provide reliable estimates of the cumulative CH4 uptake. 收起