Using Sensitivity Analysis and Fine-Scale Field Measurements to Understand How Canopy Interception Models Function

Abstract:

The capacity of the forest canopy to intercept precipitation and partition the remaining water into throughfall and stemflow largely influences the surface water budget in forested ecosystems. These processes are controlled by species-specific traits, canopy seasonality, and meteorological conditions. The complexity of these interacting factors at varying temporal and spatial scales can lead to errors in estimating canopy interception and reduce accuracy of derivative watershed hydrologic modeling efforts. To improve interception estimates, model calibration and validation must be assessed using long-term, fine-scale field measurements that capture the variability of all interacting factors. As such, field measurements of subcanopy hydrologic fluxes and meteorological conditions during discrete storm events were taken from 2007 to 2012 in a deciduous forest dominated by Fagus grandifolia and Liriodendron tulipifera in Fair Hill, Maryland, USA.

Preliminary results suggest that many of the current interception models (e.g., Gash and Rutter-types) are driven primarily by evaporation terms. However, field measurements indicate that a large degree of variability in both throughfall and stemflow partitioning is derived from biophysical characteristics. For example, even within the small 12-hectare research catchment, differences in species composition induced by slight changes in elevation, coupled with slope orientation, resulted in sufficient canopy variability whereby throughfall fluxes were definitively different across small distances. Additionally, smaller trees were more efficient in generating stemflow, while species with smoother bark generated large quantities of stemflow under a variety of storm conditions—a mechanism that may further confound modeling efforts. To improve canopy interception estimates, model sensitivity analysis was used to determine the influence of current model parameters and how biophysical canopy characteristics may be further integrated into such models.