On the magnitude and dynamics of eddy covariance system residual energy (energy balance closure error) in subsurface drip-irrigated maize field during growing and non-growing (dormant) seasons

We investigated the magnitude and dynamics of the eddy covariance system (ECS) residual energy (energy balance closure error) for a subsurface drip-irrigated maize (Zea mays L.) field in 2005 and 2006 growing and non-growing (dormant) seasons. The corrections for coordinate rotation, oxygen, frequency, and Webb–Pearman–Leuning corrections improved the slope of the total convective energy (latent heat + sensible heat) with respect to the net available energy (from 0.68 to 0.84), but the data filtering (for horizontal and frictional wind speeds higher than 2 m s^?1 and lower than 0.2 m s^?1) had little effect on the slope. Also, the number of data points available for the analyses was reduced by 53 % after filtering. Overall, the daytime residual energy varied between ?100 and 200 W m^?2 during the dormant seasons and between ?500 and 600 W m^?2 during the growing seasons. Most of the nighttime residual energy ranged within ±40 W m^?2 during the calendar year in 2005 and within ?60 and 20 W m^?2 in 2006. During nighttime, the total convective energy is vertically distributed with respect to (R _n ? G), indicating that the total convective energy is independent of the variations in (R _n ? G). Secondly, it was observed that nighttime residual energy did not show any seasonal variation patterns throughout the two consecutive years and confined mostly within a narrow range of ±40 W m^?2, showing no dependency on seasonal changes in surface conditions. The maximum variation in residual energy was usually around frictional wind speed of 0.3–0.5 m s^?1 (varying between ?150 and 300 W m^?2) and then decreasing to a range of ±100 W m^?2 at higher frictional wind speeds. On average, the residual energy decreased by about 33 W m^?2 (after the intercept) for every 1.0 m s^?1 increase in frictional wind speed, whereas the residual energy decreased by about 4 W m^?2 (after the intercept) for every 1.0 m s^?1 increase in horizontal wind speed. Similar diurnal residual energy distribution patterns, with different magnitudes, were observed during growing and dormant seasons. Even though a slight decrease in residual energy was observed with increase in leaf area index (LAI) in both growing seasons, LAI did not have considerable influence on the seasonal variation in the residual energy. The residual energy was also evaluated by separating the data into morning and afternoon hours. We observed that the root-mean-squared difference value is slightly greater for the morning data than the afternoon, indicating greater residual energy in the morning hours due to weaker turbulent mixing than the afternoon. Overall, significant reduction in the available evapotranspiration data after applying a series of corrections possess challenges in terms of utilization of ECS for in-season irrigation management and crop water requirement determinations that needs to be further researched and addressed.