Satellite remote sensing model for estimating canopy transpiration in cypress plantation using in situ sap flow observations and forest inventory

Abstract

Transpiration (Et) in forests accounts for a significant fraction of evapotranspiration. However, because Et is influenced by meteorological factors and the physiological response of vegetation, developing an accurate model for predicting Et presents a challenge. In this study, we developed a model to estimate Et specific to cypress plantation forests using satellite remote sensing and sap flow data in Japanese cypress plantations composed of a single forest species. The contribution of Et to meteorological (Etbase) and plant physiology (Etleaf) was determined using the decoupling index along with the surface temperature. Using the Forest Inventory (FI), in which the extent of growth of each species, including cypress, was surveyed and Land Remote Sensing Satellite imagery of the FI range was obtained, an index of activity estimation for cypress trees was developed. The sapwood area, which significantly influences Et, was calculated from the tree age obtained from FI or diameter at breast height and incorporated into the model. The developed model exhibited a high correlation of r = 0.74–0.88 with the measured Et. The antecedent precipitation index (API), which represents the wetness of the land surface, can be used to evaluate the confidence level and correction coefficient (if needed) of model estimates. When corrected by API, Et could be estimated with even higher accuracy (r = 0.76–0.89). The model developed in this study considers the physiological responses of vegetation and plant species, and the same approach can be used to develop models to estimate Et for other tree species. Additionally, our model uses reflectance at wavelengths close to visible light and surface temperature and can, therefore, be readily applied to other remote sensing methods, such as unmanned aerial vehicles and airborne methods. The model developed in this study can be used with FI data to estimate Et by tree species to provide detailed and accurate estimates of Et.