Many tree species rely on both precipitation and groundwater to thrive. But it remains challenging to predict how trees might respond to reduced precipitation, as most plant models ignore the role of groundwater. This question is particularly pressing when severe drought is anticipated in the future and water demand from human society keeps increasing. An integrated model is developed by coupling a plant hydraulics model (TREES) to a groundwater hydraulics model (ParFlow). This integrated model, ParFlow-TREES, captures the water movement within soils, and along the soil-plant-atmospheric continuum. We used the model to examine the physiological response of a riparian forest to contrasting rainfall amount during two consecutive years but with similar evapotranspiration (ET) in southwestern Alberta, Canada. Two model configurations were compared in both years: a 1D model that only considers the vertical interaction between trees and groundwater; and a 3D model that accounts for both the lateral and vertical movement of water within soil based on pressure gradients.
Results/Conclusions
Based on the 1D model, ET could only be sustained during the dry year when a shallow water table was in place, suggesting groundwater subsidy provided additional water supply and mediated water stress when precipitation was low. If the water table was too deep, trees were more likely to have hydraulic stress and suppressed ET. Although the 1D model reasonably captures the dynamics of ET and soil moisture during the dry year (with R2 of 0.8 and 0.75, respectively), it did not work well for the wet year. The 3D model captured not only the vertical interaction between plant and groundwater, but also the lateral movement of water from the stream to the floodplain, and provided better characterization of the high soil water content and ET during the wet year. These results suggested groundwater can be critical to better understand the hydraulic water stress of riparian forests under different climatic conditions. The promising performance of the coupled ParFlow-TREES model suggests it is a useful tool for providing realistic predictions of riparian forest health under the changing climate.