Global modeling of precipitation partitioning by vegetation and their applications

The partitioning of precipitation by vegetation canopies into throughfall and evaporation can have a large effect on water availability for both ecosystems and human consumption. Canopies are often the first point of contact between precipitation and the land surface, yet the complexity and inaccessibility of this interface make measurements extremely difficult, and as a result our attempts to understand and model the relevant processes are only weakly constrained. Here, we describe common approaches to modeling canopy interception processes, particularly in global models. These models use a wide variety of parameterization and parameters internally, suggesting that we do not have a good understanding of how to model canopy interception on a global scale. To begin to quantify how big a problem this may be, we present a few ideal model experiments exploring a range of modeling approaches and assumptions to document what effect these choices have on the projection of changes in inputs to the eco-hydrologic system. Perhaps unsurprisingly, these effects vary with vegetation type and density, as well as precipitation type, intensity, and changes in precipitation. In many cases, these effects are likely dwarfed by the uncertainties in predicted changes in regional climate, but not accounting for our lack of certainty in canopy processes can lead to an overconfident, and in some cases likely incorrect, projection of future changes in water availability. This should serve as a call to action for those studying canopy interception processes to better document and consider how theories can be put into a numerical framework.