Observational study of turbulent exchange between the surface and canopy layer over several forest types (Micrometeorology)

Several field projects have been organized (e.g., AMERIFLUX, and EUROFLUX) to quantify the role of the terrestrial ecosystems as a sink of CO₂ and source of H₂O_(v) and other trace gases that can be important in global climate change. As forests are a porous boundary, the understanding of turbulent exchange from the canopy layer to the atmosphere is a step to understand this role.; Using information on canopy structure such as the vertical canopy density area (CAD), we seek similarity forms of universal functions to describe turbulent parameters. These general forms will allow a more general description of the exchange properties, such as variances and covariances, for a wide variety of forests within the canopy layer (CL). Introducing a dimensionless height, z_c′ = f(CAD), a normalized cumulative canopy area density, we find that turbulent parameters profiles are bounded by two curves when z_c ′ is used, which represent sparse (mainly conifers) and dense canopies (mostly deciduous). Using z_c′ to determine the displacement height (d) provides a better agreement of wind profiles above the canopy among several forest types.; Presence of the canopy alters the shape of the spectra and cospectra. An attenuation of the spectra in a frequency band (0.4 < f ′ < 2, where f′ is the dimensionless frequency) in w spectrum is observed when the canopy is foliated, but it is absent when the canopy is leafless. The observed spectra in the roughness sub layer are more peaked than those observed over flat surfaces. As the measurement height is higher, away from to the canopy top, the normalized heat flux cospectra start to be broader than the normalized momentum flux, going asymptotically to the cospectral shapes over flat surfaces. At high wind speed, the cospectra of the turbulent parameters shift to high frequencies, affecting the proper period of the Reynolds' averaging that has to be chosen. Therefore, during low wind speed days, longer the Reynods' averaging should be used to sample the slow moving eddy. The 4–30 minute oscillations can contribute about 17% of the noontime period fluxes. Accounting the contribution from these oscillations we are able to improve the energy budget closure.