A Site-Level Comparison of Lysimeter and Eddy-Covariance Flux Measurements of Evapotranspiration

Abstract:

Accurate measurements of evapotranspiration are required for many meteorological, climatological, ecological, and hydrological research applications and developments. Here we examine and compare two widely used methods to measure evapotranspiration at the site level: lysimeter-based measurements (E_L) and eddy-covariance (EC) flux measurements (E_EC). The analyses are based on parallel measurements at the research catchment Rietholzbach in northeastern Switzerland, and focus on the period June 2009 to December 2013. The measurements are compared on the yearly, monthly, daily, and hourly time scales, and with respect to an over 35-year lysimeter-based evapotranspiration time series. Overall, the two measurement techniques agree well, especially on the annual time scale. They are also within the range of catchment water-balance estimates of evapotranspiration. The good agreement of these methods emphasizes the representativeness of the lysimeter and EC measurements for the entire catchment despite their comparatively small source areas and the heterogeneous land use and topography within the catchment. This finding is supported by the fact that the in-situ lysimeter estimates also agree well with remote sensing-based estimates of evapotranspiration on a 25 km scale.

We also assess different methods to close the EC-based energy balance. Based on the comparison of E_L and E_EC, the commonly used method of closing the energy balance according to the Bowen ratio is found to be reasonable and supported by our results. Furthermore, we identify that lack of EC observations during and following rainfall events (due to shortcomings of the measurement technique under these conditions) significantly contributes to an underestimation of E_EC and to the overall energy balance gap at the site.