Surface Renewal Measurements of Sensible Heat Exchange from Tree, Crop, and Soil in a Coffee Agroforestry System

Abstract:

Agroforestry is promoted as the solution to declining crop productivity due to changing rainfall and temperature regimes. However, little is known about the water and energy exchange of the various strata in these systems, not least because of the complexity and costs involved in measuring these fluxes. This study evaluated the potential of the surface renewal (SR) method to quantify sensible heat exchange (H) from the tree, crop, and soil layers in a shaded coffee plantation. The analysis included data from above the tree canopy and data from above the coffee and ground layers in an intensively and an organically managed plot. The SR model of Chen et al. (1997, https://doi.org/10.1023/A:1000342918158) was evaluated. Key inputs to this model are the maximum of the third-order temperature structure function and friction velocity (u_*). The model has three empirical parameters: α, β, and γ. The latter two were obtained from the temperature and u_* data, while α was obtained by comparison with H measured by eddy covariance. For the above-canopy and the above-coffee levels the roughness sublayer formulation of the model was used (requiring u_* from above the canopy), while for the above-ground measurements the inertial sublayer formulation was used (requiring u_* at ground level). After calibration, SR derived H was in good agreement with that measured by eddy covariance at all three levels. Even though H measured above the coffee and at ground level was < 10% of that observed above the tree canopy, the model captured well the hourly and day-to-day variability of H at these levels. Also after correcting for the effect of slow response time, the empirical coefficients were found to depend on thermocouple diameter (0.07-0.12 mm), except for the smallest diameter (0.01 mm). However, using the sonic anemometer temperature corrected for the effect of humidity yielded remarkably similar values of the combined empirical coefficient (αβ^2/3γ) for the above-canopy (0.22) and above-coffee levels (0.23 and 0.24). Greater variability of this coefficient was found at ground level (0.18 to 0.21 in the organic plot and 0.46 in the intensive plot). The present results indicate that, in combination with information on radiation, the SR method can be used to make indirect estimates of evaporation from the various strata in agroforestry systems.