Estimation of surface energy fluxes using surface renewal and flux variance techniques over an advective irrigated agricultural site ☆

Estimation of surface energy fluxes over irrigated agriculture is needed to monitor crop water use. Measurements are commonly done using well-established techniques such as eddy covariance (EC) and weighing lysimetry, but implementing these to collect spatially distributed observations is complex and costly. Two techniques that could simplify flux observations are the surface renewal (SR) and flux variance (FV) approaches. These methods infer sensible heat fluxes from high frequency observations of near surface air temperatures using low cost thermocouples. In combination with net radiation and soil heat flux observations, surface renewal and flux variance observations produce latent heat fluxes as a residual of the surface energy balance. The viability of these techniques was tested in a strongly advective irrigated agricultural setting as part of the Bushland Evapotranspiration and Agricultural Remote Sensing Experiment in 2008 (BEAREX08). Using 20 Hz air temperature data collected between 12 June and 13 August from two cotton field sites and one senescent/dormant grass site, sensible heat flux estimates were computed. Surface flux conditions ranged widely and include episodes of latent heat fluxes exceeding net radiation. Overall, flux estimates from SR and FV were similar to simultaneously obtained eddy covariance observations on most days. During strong advection neither approach closely agreed with EC data, although the surface renewal technique more reliably estimated the correct sign of sensible heat fluxes. Both techniques were found to offer flux estimates comparable to EC data, though with different advantages. SR is self-contained, requiring no additional instrumentation beyond air temperature equipment. SR correctly diagnosed the sign of sensible heat fluxes and produced better estimates at early morning and late afternoon times than FV, although these were achieved after lag time selection using EC data for calibration. FV, by contrast, required wind speed observations, as well as thermal infrared data to resolve heat flux directions. However, using nominal parameters and no local calibration, FV produced mid-day estimates equal to or better than SR. These outcomes indicate that flux data with accuracies approaching EC capabilities is feasible with the potential for reduced deployment complexity and cost.