While transpiration is a well studied process at the leaf and plant scale, integrating plant water fluxes across ecosystems remains a challenge. This is due in part to the wide diversity of plants that can be contained within a single community, each with specific niches and adaptations. The eddy covariance technique, which has shown success in estimating ecosystem carbon fluxes, has seen a number of transpiration estimation methods emerge in the last few years.
Results/Conclusions
We compare three transpiration methods which use existing eddy covariance datasets such as FLUXNET. The methods show a high correlation (R² between 0.80 and 0.87) but a divergence in magnitude, with the fraction of transpiration to evapotranspiration ranging between 45% to 77%. However, all three methods show patterns both consistent with independent measurement from sap flow and with expected physiology such as the response of water use efficiency to vapor pressure deficit. Finally we demonstrate how these transpiration estimates can be scaled to a high spatial and temporal global product using the FLUXCOM methodology, thus providing insights into ecosystem water fluxes around the world.