Traces of high frequency temperature data are known to show ramp-like structures. Air temperature ramps result from turbulent coherent, which consist of ejection and sweep phases that occur under shear condition. These temperature ramps can be used to estimate sensible heat flux density (H) over vegetated surfaces using the surface renewal (SR) method. The method was tested with temperature recorded for different crop canopies and provided good estimate of H regardless of the flux direction and without the need of temperature profiles and wind speed data. Earlier reports showed that SR could provide estimates of H over short and dense canopies that are typically within 50 W m-2 of values measured with a sonic anemometer.
The objectives of this study were to evaluate (1) SR analysis over sparse (grapevine) canopies, (2) measurement height on accuracy, and (3) the need for a calibration factor to account for uneven heating of the mean air volume under a measurement level.
Three separate field experiments were conducted over grapevines in Napa Valley, California, in August 1995, and in Sardinia, Italy, in July 1996 and 1997. High frequency temperature data were collected at the canopy top and at 0.3 m intervals above the canopy top for the experiment in Napa Valley and at the canopy top and at 0.3 m intervals within and above the canopy height for both experiments in Sardinia. Structure functions using several sample time lags were used to determine the temperature ramp amplitude and the inverse ramp. The ramp characteristics were used to estimate H and were compared with H from a uni-axial sonic anemometer. Net radiation (Rn), soil heat flux density (G), and high frequency humidity flux from krypton hygrometer were also measured.
The accuracy of H estimate was evaluated using a regression of H from SR versus H from the sonic anemometer through the origin. The slope (alpha), the R2, and the RMSE were determined. The alpha values decreased with increasing of sampling height above the canopy. The decrease of alpha is consistent with earlier trials that show alpha depends on the measurements level in relation to the postulated height to which the renewal volume was heated or cooled and the vertical temperature profile during the ramp events. The best-calibrated H estimates were observed at the canopy top where alpha values were about 0.8 with R2 larger than 0.85. Temperature measurements were collected inside the canopy profile and the total H value was calculated as the sum of H values from canopy layers. This eliminates the need for the alpha weighting factor. When H from the SR method was used in the energy balance equation, good estimates for latent heat flux density (LE) were obtained.