Rainfall interception by maize canopy: Development and application of a process-based model

Abstract

• Simple canopy generator reproduces observed canopy gap fraction. • Model predicts dripping locations, which can trigger localized soil erosion. • Model confirms splashing as considerable source of small throughfall droplets. • Predicted throughfall spatial distribution poses a challenge to correct sampling. The interaction between rain drops and crop canopies changes the microphysical characteristics of precipitation. Understanding the mechanisms driving these changes is a key step towards unveiling the mechanics of soil water recharge, soil erosion under crop canopies, and evaporation of intercepted rainfall. The latter caused by, for example, the reduced aerodynamic resistance to evaporation of splashed droplets. We propose a model that uses drop-size and velocity distributions as well as the three-dimensional geometry of the maize canopy to simulate the movement of raindrops on the surface of the leaves. The model accounts for the interception, redirection, retention, coalescence, and re-interception of drops to predict the location, size, and velocity of throughfall drops beneath the canopy. The throughfall results are presented as two-dimensional matrices, in which each term corresponds to the accumulated volume of drops that dripped in that location, which offers insight into its spatial distribution under the foliage. We examine the modification of the drop-size distribution by the maize canopy by recalculating the drop velocity based on their size and detachment height. They built the three-dimensional digital canopy that is employed in the simulation by creating modified copies of one photogrammetry-generated digital model of a plant located inside their experimental site. The canopy model accounts for the projected overlap of the single plant leaves. We evaluate the results against measurements obtained during 10 storms that occurred between 14 July 2009 and 28 August 2009 in Shueyville, Iowa. The comparison between observations and simulations corroborates the drop detachment threshold of 3.75 mm and indicates that splashing, as opposed to considering solely the rolling or bouncing of droplets that lack the kinetic energy to attach to the foliage, is the likely source of throughfall drops with diameters smaller than 1 mm. The predicted throughfall spatial distribution unveils localized dripping points, which pose a challenge to the correct sampling of throughfall.