Evapotranspiration Measurement and Estimation: Weighing Lysimeter and Neutron Probe Based Methods Compared with Eddy Covariance

Abstract:

Evapotranspiration (ET) may be measured by mass balance methods and estimated by flux sensing methods. The mass balance methods are typically restricted in terms of the area that can be represented (e.g., surface area of weighing lysimeter (LYS) or equivalent representative area of neutron probe (NP) and soil core sampling techniques), and can be biased with respect to ET from the surrounding area. The area represented by flux sensing methods such as eddy covariance (EC) is typically estimated with a flux footprint/source area model. The dimension, position of, and relative contribution of upwind areas within the source area are mainly influenced by sensor height, wind speed, atmospheric stability and wind direction. Footprints for EC sensors positioned several meters above the canopy are often larger than can be economically covered by mass balance methods. Moreover, footprints move with atmospheric conditions and wind direction to cover different field areas over time while mass balance methods are static in space. Thus, EC systems typically sample a much greater field area over time compared with mass balance methods. Spatial variability of surface cover can thus complicate interpretation of flux estimates from EC systems. The most commonly used flux estimation method is EC; and EC estimates of latent heat energy (representing ET) and sensible heat fluxes combined are typically smaller than the available energy from net radiation and soil heat flux (commonly referred to as lack of energy balance closure). Reasons for this are the subject of ongoing research. We compare ET from LYS, NP and EC methods applied to field crops for three years at Bushland, Texas (35° 11’ N, 102° 06’ W, 1170 m elevation above MSL) to illustrate the potential problems with and comparative advantages of all three methods. In particular, we examine how networks of neutron probe access tubes can be representative of field areas large enough to be equivalent in size to EC footprints, and how the ET data from these methods can address bias and accuracy issues.