Nighttime Turbulence Statistics and Trace Gas Exchange in a Mixed Deciduous Forest

Turbulence statistics were measured at eight levels in and above a mixed northern hardwood forest in northern lower Michigan (USA) during a 15 day intensive period in July and August 1999. Three dimensional sonic anemometers were operated at the mean canopy height (h=22 m), and at 1.6 h and 2.1 h. At the upper two levels, continuous eddy-covariance measurements of CO2 and vapor fluxes are made as part of the AmeriFlux sampling agenda. Vertical velocity fluctuations were measured using 1-D sonic anemometers at five levels in the canopy layer. In addition, soil and bole CO2 emission were measured with portable flux chambers, on a number of clear nights during that period. CO2 storage in the canopy layer is evaluated from an eight-level concentration profile. These measurements are used to address the concern that, at night, turbulence intensity is often small and turbulent fluxes as measured by eddy covariance become ill defined. Moreover, advection or drainage flows due to horizontal heterogeneity or topography at various scales may export material released at the forest floor from the ecosystem horizontally, without being accounted for above the canopy. Our data show that vertical velocity statistics follow established similarity relations above canopy, and in the canopy layer the attenuation of turbulence is closely related to the profile of vegetation area density, but with distinct differences between day and night. We examine the role of turbulence in the nocturnal exchange of trace gases from the forest floor through the canopy by evaluating the one dimensional CO2 conservation equation in the canopy-air layer. Advective transport out of a finite "footprint volume" may be examined from imbalances in the 1-D equation and their magnitude will be discussed in terms of the soil level emission fluxes and the turbulence activitity within and above the canopy.