Analysis of Scale-dependent Spatial Correlations of Actual Evapotranspiration Measured by Lysimeters

Actual evapotranspiration (ETa) is difficult to measure and limited long-term information is available about ETa. With eddy covariance systems ETa can be measured at the field scale, but the method is associated with energy balance closure issues. For measuring ETa, weighing lysimeters are considered to be the most accurate and reliable method. However, weighing lysimeters have some disadvantages like elevated costs of installation and maintenance, and a small footprint (e.g., about 1 m2). A main question is therefore whether the precise ETa-measurements by lysimeters are representative for a larger area like a field, a meso-scale catchment, or even a larger region. Our hypothesis was that a lysimeter provides information about ETa that represents a larger area than its underlying measurement area. To this end, we examined here the daily ETa measurements from lysimeter at four study sites across Germany (separation distances 10 - 500 km) for the years 2015 to 2020. The Pearson correlations of the standardized anomalies (SA) of daily ETa between different lysimeters were calculated and compared with SA of daily ETa obtained from the corresponding eddy covariance tower. The correlations were further analyzed and related to spatial correlations of SA of environmental controls like precipitation, potential evapotranspiration (ET0), and soil moisture. We found that SA of daily ETa shows high spatial correlations (>0.5) for considerable separation distances between sites of up to 50km, with similar correlations for lysimeters and eddy covariance systems. ET0 is the dominant factor for the spatial correlation of ETa, as SA of ET0 shows stronger spatial correlations than SA of ETa.