Modeling canopy interception under drought conditions: The relevance of evaporation and extra sources of energy

Modeling canopy interception is fundamental for understanding the forests' role in local and regional hydrology. In this study, canopy interception (CI), throughfall (TF), and stemflow (SF) were evaluated for a semi-deciduous Atlantic Forest (AFR) from 2013 to 2019, where a prolonged dry period occurred. The Gash and Liu models were analyzed in detail to determine the most appropriate for modeling CI throughout drought conditions. Thus, the climatic parameters were retrieved annually by a modified TF-based method (E_I%), whereas the structural parameters represented the entire period. The contribution of the energy stored in the forest (i.e. air and biomass; Q) to CI was also assessed in the AFR stand. Both models performed well when using E_I%, as the Gash model overestimated CI by 71 mm (4.6%), whereas the Liu model underestimated it by only 13 mm (0.85%). This performance is due to an increased Q and turbulent mechanisms (such as advection and strong updrafts) that occur in drought conditions and are indirectly accounted for in E_I%. However, the Liu model stood out for modeling CI under a prolonged dry period, as the exponential wetting approach better represents the complex canopies of the semi-deciduous forests. Thus, we recommend the Liu model and additional energy sources when dealing with prolonged droughts, as in the case of climate change scenarios projected to the studied region.