QUANTIFYING EVAPOTRANSPIRATION IN POPLAR TREES WITH IMPLICATIONS FOR PHYTOREMEDIATION

Phytoremediation can be an effective technology for containing impacted groundwater. However, evaluating the effectiveness of these systems requires a careful quantification of how much water is being removed from the subsurface. A series of research studies was conducted for a phytobarrier of poplar trees installed in a western New York site. The purpose of the current project was to develop a regression model to predict the volume of water being transpired by each poplar tree in a single day as a function of: daily weather fluctuations, the local depth to groundwater, and the cross sectional area of each tree. The modeled relationship between observed factors was then scaled up to predict the evapotranspiration of the fully mature phytobarrier system. The model expressed the predictive variables by quantifying: (1) the relationship between daily weather fluctuations via the Penman-Monteith equation (ET₀ ), (2) the rooting distance to groundwater via a categorical location variable (LOC), and (3) the tree trunk cross sectional area (CSA). These variables were related to sap flow measurements of evapotranspiration (ET_sap) using various regression models similar to the following formula:

ET_sap=β₀+β₁LOC+β₂CSA+β₃ET₀

Variations on the above model, including nonlinear variable transformations and interactions, were compared using statistical goodness-of-fit measures. The development of a robust statistical model will eventually support phytobarrier design calculations to identify the tree density needed to provide plume containment.