Fluctuations in soil greenhouse gas (GHG) are an important part of the terrestrial ecosystem carbon-nitrogen cycle, but uncertainties remain about the dynamic change and budget assessment of soil GHG flux. Using high frequency and consecutive soil GHG fluxes measured with an automatic dynamic chamber system, we tested the applicability of the current Forest-DNDC model in simulating soil CH₄, CO₂ and N₂O fluxes in a temperate broad-leaved Korean pine forest at Changbai Mountain. The results showed that the Forest-DNDC model reproduced general patterns of environmental variables, however, simulated seasonal variation in soil temperature, snow melt processes and soil moisture partly deviated from measured variables, especially during the non-growing season. The modeled CH₄ flux was close to the field measurement and co-varied mainly with soil temperature and snowpack. The modeled soil CO₂ flux had the same seasonal trend to that of the observation along with variation in temperature, however, simulated CO₂ flux in the growing season was underestimated. The modeled N₂O flux attained a peak in summer due to the influence of temperature, which was apparently different from the observed peak of N₂O flux in the freeze-thaw period. Meanwhile, both modeled CO₂ flux and N₂O flux were dampened by rainfall events. Apart from consistent estimation of annual soil CH₄ flux, the annual accumulation of CO₂ and N₂O was underestimated. It is still necessary to further optimize model parameters and processes using long-term high-frequency observation data, especially transference of heat and water in soil and GHG producing mechanism. Continues work will improve modeling, ecosystem carbon-nitrogen budget assessment and estimation of soil GHGs flux from the site to the region.