Analysis of experimental data gathered within and above several types of plant canopies shows that repeated, organized features in the turbulence (coherent structures) are responsible for a majority of scalar exchanges between the canopies and atmosphere. These coherent structures are related to temperature traces.. Under unstable conditions, the ramps exhibit a slow rise in temperature followed by a sudden drop. Under stable conditions, when there is warm advection and at night, the temperature ramps show a slow decrease in air temperature followed by a sudden rise. In both stability conditions, the temperature ramps can be used to estimate sensible heat flux density (H) using Surface Renewal (SR) method. Mean values for ramp amplitude (a) and inverse ramp frequency (l + s) during half-hour intervals are used to estimate H. A structure function and statistical moments are used to determine ‘a’ and ‘l+s’. The size of the temperature sensors used in the temperature measurements may affect results. Data from smaller sensors exhibits more variability, however this variability necessarily important to characterized ramps. On the other hand, as the sensor size increases, the possibility for bigger errors due to radiation loading could be a problem.
The objective of this study was to evaluate the effect of transient and radiation errors determined by wire size and configuration of different thermocouples in terms of accuracy of H estimate from SR analysis. To check the effect of different thermocouples on temperature traces, a field experiment was conducted over bare soil at the University of Sassari at Oristano research station, Sardinia, Italy, in September 1997. Data were collected for several days immediately after wetting the soil to obtain a wide range of H conditions. High frequency temperature data were recorded using exposed 0.013, 0.025, and 0.076 mm diameter single and double wired type E thermocouples. Data were sampled at 8 Hz by a data-logger. Half-hour data set were created to correspond to the H estimate from a sonic anemometer.
Temperature traces and statistical moments from different thermocouples were compared. Regression of H values from SR versus H values from sonic anemometer were computed. Results showed significant differences in both temperature traces and statistical moments obtained using single and double wired thermocouples for different thermocouple sizes.