Modeling Rainfall‐Runoff and Soil Erosion Processes on Hillslopes With Complex Rill Network Planform

Abstract

The effect of rill network planforms on hillslope rainfall-runoff and soil erosion processes is usually neglected in modeling practices, although they can markedly alter the hydrologic and geomorphic processes. Based on the CeRIRM model and WEPP erosion theory, a simple approach is developed to account for these effects. In the framework, several characteristic parameters including the average rill width, rill orientation angle, the number of rills, and number of discontinuous rills and their variations along the hillslope are introduced to represent the planar characteristics of the rill network. The model is tested against experimental erosion data from a hillslope subjected to three successive rainfall events, resulting in continuous rill network evolution. The results show that rill network planforms alter the partitioning of interrill and rill flows, thereby modifying the hydraulics, erosion, and sedimentation in the rills. Rill characteristics are found to significantly affect the amount of rill erosion. The new approach is compared to WEPP, which ignores rill network features. The WEPP approach of simulating rill erosion on hillslopes using one set of model parameters leads to errors that are up to 30% larger than the new approach that is able to account for spatially and temporally varying rill characteristics. The differences in cumulative rill erosion amounts between the two models vary significantly with slope length, slope angle, rill orientation angle, number of rills, and discontinuous rills. The results are valuable for the development of general rainfall-runoff and soil erosion models on hillslopes with complex geomorphic features. Plain Language Summary Rill erosion is one of the main forms of soil erosion and is associatedwith fast flowing water in networks of small channels, that is, rills. Modeling rill erosion is an important yet challenging task informing soil conservation. Although actual rill networks are usually composed of irregular, tortuous, and even discontinuous rills, using parallel and straight rills to simplify the actual rill network is still the prevailing method in current models. This study demonstrates the impact of the irregular planform of rill network on soil erosion and proposes a simple and effective conceptualization to represent complex rill networks using only a few parameters, including rill orientation angle, number of rills, and number of discontinuous rills. The approach can better represent the spatially intricate partitioning of interrill and rill flow and improve the modeling of rill hydraulics, erosion, and sedimentation prediction. Owing to its improved physical basis, the proposed approach simulates both the total amount and the spatial distribution of erosion with improved accuracy. The results of this study are expected to help improve general rainfall-runoff and soil erosion models for hillslopes with complex geomorphic features, and benefit other scientific fields involving transport through complex rill networks.

Modeling Soil Erosion With Evolving Rills on Hillslopes

“…This is based on the consideration of soil surface sealing occurring during the first rainfall (Shen et al, 2015). The soil hydraulic properties of the two layers are quantified by Wu et al (2018). The soil saturated hydraulic conductivity of the top layer (1.0 cm in thickness), K 1s , decreases from 3.1 to 0.05 cm h −1 during the first rainfall and remained 0.05 cm h −1 for the second and third rainfalls, while the soil saturated hydraulic conductivity of the underlying layer soil, K 2s , remains at 3.1 cm h −1 .…”

“…The validation of simulated rill evolution includes the four key rill network characteristic parameters, that is, rill width, depth, orientation angle, and lateral density, as well as the runoff and erosion processes. Note that the approaches such as WEPP and Wu et al (2018) ignore the rill evolution processes, and thus, we only present the rill evolution results of the proposed model. Figures 3a-3c compare the measured lateral rill density to the simulated values, which are a function of the downslope distance, at the end of the three successive rainfalls.…”

“…The accurate prediction of rill width and depth also validates the performance of the present modeling approach. Figure 4 shows the measured cumulative runoff depth and erosion compared with the simulated results by the present model, the model of Wu et al (2018), and the WEPP approach. The runoff depths simulated by the present model are very close to the observed values.…”

“…At the hillslope scale, soil erosion by overland flow usually generates complex rill networks (e.g., Berger et al, 2010; Bennett et al, 2015; Brunton & Bryan, 2000; Favis‐Mortlock, 1998; Favis‐Mortlock et al, 2000; Hofer et al, 2012; Moreno‐de las Heras et al, 2010). The characteristics of rill network vary remarkably with space and time and have great impacts on flow and sediment routing and soil erosion processes (Brunton & Bryan, 2000; Gumiere et al, 2009; Mancilla et al, 2005; Robichaud et al, 2010), which in turn affect the geomorphology of rill network (Oz et al, 2017; Shen et al, 2015; Wu et al, 2018). This means that soil erosion and rill evolution create a feedback loop where each part of the system responds to changes of other parts.…”

Variable power‐law scaling of hillslope Hortonian rainfall–runoff processes

“…This discrepancy likely occurred because we did not include the impacts of soil heterogeneities, vegetation, and rills in the present study, although they appear on many natural hillslopes. The soil heterogeneities can directly affect the soil infiltration and thus modified the run-off generation (Langhans et al, 2010(Langhans et al, , 2011, the vegetation may further complicate the results by directly enhancing the local infiltration rates and increasing the overland flow resistance (Chen et al, 2013(Chen et al, , 2016Moreno-de las Heras et al, 2010;Sidle, Hirano, Gomi, & Terajima, 2007), and the rills change the hillslope microtopography, overland flow routing, and function as flow conveyors (Nearing et al, 1997;Wu, Chen, Wang, Yu, & Assouline, 2018). Both of these factors can alter the rainfallrun-off mechanism and affect the scaling relationship, which warrant further investigations.…”