Modeling Environmental Controls on Tree Water Use at Different Temporal scales

Abstract:

Vegetation covers 70% of land surface, significantly influencing water and carbon exchange between land surface and the atmosphere. Vegetation transpiration (E_t) contributes 80% of the global terrestrial evapotranspiration, making an adequate illustration of how important vegetation is to any hydrological or climatological applications. Transpiration can be estimated through upscaling from sap flow measurements on selected trees. Alternatively, transpiration (or tree water use for forests) can be correlated with environmental variables or estimated in land surface simulations in which a canopy conductance (g_c) model is often used. Transpiration and canopy conductance are constrained by supply and demand control factors. Some previous studies estimated E_t and g_c considering the stresses from both the supply (soil water condition) and demand (e.g. temperature, vapor pressure deficit, solar radiation) factors, while some only considered the demand controls. In this study, we examined the performance of two types of models at daily and half-hourly scales for transpiration and canopy conductance modelling based on a native species in South Australia. The results show that the significance of soil water condition for E_t and g_c modelling varies with time scales. The model parameter values also vary across time scales. This result calls for attention in choosing models and parameter values for soil-plant-atmosphere continuum and land surface modeling.