Since turbulent mixing often becomes discontinuous and intermittent in the stable stratified boundary layer, so do nighttime CO2 fluxes observed over an ecosystem. The friction velocity correction (u*-correction) is applied to nighttime CO2 fluxes measured with the eddy covariance technique for evaluating the ecosystem respiration rate. However, the applicability of this procedure is uncertain in intermittently turbulent conditions as the friction velocity oscillates between extremely large and small values.
We used a simple numerical model to simulate intermittent mixing in the stable boundary layer and analysed the characteristics of intermittent nighttime CO2 fluxes. We then examined the applicability of the u*-correction in intermittently turbulent conditions. The simulation results showed alternations between highly turbulent periods with large positive CO2 fluxes and quiescent periods with small positive and zero CO2 fluxes. With increasing the geostrophic wind in the range of 5.0 m s-1 to 8.0 m s-1, the magnitude of upward intermittent CO2 fluxes as well as the time intervals between intermittent turbulent periods became smaller. A non-linear dependency of the upward CO2 fluxes on the surface temperature was not clear during intermittently turbulent periods. The fluxes observed in conjunction with high u* values showed a better relationship with the surface temperature although use of such fluxes in the u*-correction procedure overestimated the ecosystem respiration rate: the u*-corrected estimates of the respiration rate were twice as high as the ecosystem respiration rate. During flux events, CO2 storage fluxes became large and negative. In such cases, the CO2 storage flux must be added to the CO2 flux for evaluating the ecosystem respiration rate using the u*-correction procedure.
