Modeling rainfall interception components of forests: Extending drip equations

Abstract

Assessment of forest interception, I, and its components; the average evaporation rate during the storm, E¯, and canopy storage, S, are essential for simulating the contribution of forests to the water cycle and the climate system. The objectives of this study were to: (i) propose a new model to predict I, E¯, and S, as well as rainfall duration, RD and rainfall intensity, R¯ ; (ii) correlate E¯, RD, and R¯ assessments; and (iii) quantify the role plant surfaces play on the generation of interception from four forests in Mexico. Based on extended drip equations, the model was calibrated using field measurements from forty-five forest interception case studies (N = case studies, n = number of rains) in tropical dry, TDF (N = 21, n = 347), arid/semi-arid, A&SF (N = 15, n = 659), temperate, TF (N = 4, n = 258), and tropical montane cloud, TMCF, forests (N = 6, n = 658) and validated using field measurements from sprinkling experiments in ne Mexico. The model performed very good in predicting both individual and cumulative I values, with average errors, ME%, as a function of precipitation, P, smaller than 4% and Nash-Sutcliffe, NSE, values > 0.33 for three out of four forests. E¯ assessments accounted for between 65% and 93% of I in these forests. Higher E¯ and I figures were found in individual trees (3.78A mm h−1, 27%) in contrast to forest plots (2.24B mm h−1, 14%). E¯ assessments decreased as a function of RD but increased as a function of R¯ for all forests (p ≤ 0.05). Leaf area index, LAI, significantly explained part of the I variance in complex non-linear fashions (p ≤ 0.05). The novel independent assessments of I, E¯, S, RD, and R¯, the significant relationships between I components, and the complex role plant surfaces play on the generation of I fill an important scientific gap in this area of forest hydrology.