Sensible heat flux estimates using two different methods based on surface renewal analysis. A study case over an orange orchard in Sicily

► We examine the reliability of two surface renewal methods to estimate sensible heat flux. Input requirements are contrasting. ► The apparent advantages (simplicity and affordability) attributed to one method were not clear. ► Significant differences were found in terms of robustness and reliability which are mandatory for method selection. An experiment was carried out over a mature orange orchard to evaluate the reliability of two methods based on surface renewal, SR, analysis, SR1 and SR2, to estimate sensible heat flux, H. After calibration, the method SR1 only requires as input the air temperature measured at high frequency. However, method SR2 requires air temperature measurements taken at high frequency, the mean and turbulent standard deviation of the horizontal wind speed, the leaf area index, the canopy height and the vertical extent (m) of the foliage. Methods SR1 and SR2 operated at the canopy top, z = 4 m. The H measured using the eddy covariance, EC, method operating at height slightly higher than twice the canopy height, z = 8 m, was taken as a reference, H EC_8 m . For completeness, because the method SR1 may also operate well above the canopy, its performance was also analyzed at z = 8 m, and another EC system was deployed close to the canopy top at z = 4 m, H EC_4 m . For calibration, three periods of 15 days were selected. For method SR1, regardless the height at which operated, it is shown that calibration was dependent on weather conditions, including daily and seasonal patterns. Therefore, in contrast to other experiments that recommended application of method SR1, this study questions its reliability. For method SR2, calibration was the same for each calibration period. Validation was made for three periods of three months each. In relation to H EC_8 m , regardless of the validation period SR2 was closer than SR1 and it was less biased than H EC_4 m . Because reliability is mandatory for method selection SR2 is recommended over SR1 and it could be considered to fill gaps of the EC method for samples affected by flow distortion.