Apportioning contributions of individual rill erosion processes and their interactions on loessial hillslopes

Authors: QIN, CHAO - Northwest Agricultural & Forestry University; ZHENG, FENLI - Northwest Agricultural & Forestry University; Wilson, Glenn; Zhang, Xunchang; XU, XIMENG - Institute Of Geographic Sciences And Natural Resources

Submitted to: Catena
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 5/31/2019
Publication Date: 6/13/2019
Citation: Qin, C., Zheng, F., Wilson, G.V., Zhang, X.J., Xu, X. 2019. Apportioning contributions of individual rill erosion processes and their interactions on loessial hillslopes. Catena. 181 (2019) 104099. https://doi.org/10.1016/j.catena.2019.104099.
DOI: https://doi.org/10.1016/j.catena.2019.104099

Interpretive Summary: Channels include a wide range of sizes from rills to ephemeral gullies, classic gullies and rivers or streams and these have a wide range in flow rates. These channels tend to dominate soil losses, land degradation and sediment deposition in the environment. Rill erosion is the initial stage of channel formation erosion and it accounts for > 80% of total eroded sediment on hillslopes in the Loess Plateau of China. However, current methods to predict soil losses do not include the contributions of some processes involved in rill erosion such as rill headcut advance, erosion of the bottom of the channel (bed incision), or erosion of the sides of the rills (sidewall expansion). The interactions of these individual processes are also not included in predicting the over-all rill erosion. To investigate the impacts of individual processes involved and their interactions, experiments were conducted with four inflow rates (1.0, 2.0, 3.0 and 4.0 L min-1) and two slope gradients (15° and 20°). Photographs were used to measure changes in the geometry of the channels along with manual sampling to measure sediment losses. The results show that headcut advance, bed incision and sidewall expansion cause rill erosion independent of each and also interact with each other to cause rill erosion depending upon the location within the rill and the stage of rill development. Rill depth and width are determined by the initial headcut size and shape during the initial stage of rill development. Headcut advance interacts with bed incision and sidewall expansion before rill headcuts reach a critical slope length. Bed incision before the rill bottom reaches a non-erodible layer and afterwards bed incision combined with sidewall expansion dominates rill erosion. Headcut advance contributes the largest amount of rill erosion (44%-68%), followed by bed incision (27%-44%) and sidewall expansion (3.8%-12%). Headcut advance contributes the largest amount (63%-83%) to increases in rill width while bed incision contributes the largest amount (51%-65%) to increases in rill depth. Equations for predicting length, width and depth of rills on a loessial hillslope and equations for the individual rill erosion processes were determined to be valid. This work defining the contributions of individual rill erosion processes provides the necessary scientific basis for the development of improved soil erosion models that will allow the evaluation of practices for prevention of future land degradation.

Technical Abstract: Channels (rill, ephemeral gully, gully and river channel) exhibit a continuum of sizes and flow magnitudes. These sources tend to dominate soil losses, land degradation and sediment deposition sources in streams. Rill erosion, as the initial stage of channel erosion, accounts for > 80% of total eroded sediments on upland slopes in the Loess Plateau. However, current prediction technologies do not address the contributions of individual processes including rill headcut advance, bed incision, sidewall expansion and their interactions to the over-all rill erosion. Thus, simulated upslope inflow experiments were designed to investigate the impacts of individual processes involved and their interactions in rill erosion under four inflow rates (1.0, 2.0, 3.0 and 4.0 L min-1) and two slope gradients (15° and 20°). Photogrammetry and manual sampling were used to detect hillslope morphology variation and sediment delivery. The results show that headcut advance, bed incision and sidewall expansion interact with each other and exhibit both independent and dependent features across spatial and temporal scales. Headcut advance interacts with bed incision and sidewall expansion before rill head advancing to a critical slope length. Initial rill depth and width are determined by initial headcut morphology. Bed incision and sidewall expansion dominate rill erosion before and after the non-erodible layer is exposed to concentrated flow, respectively. Headcut advance contributed the largest amount of rill erosion (44%-68%), followed by bed incision (27%-44%) and sidewall expansion (3.8%-12%). Headcut advance contributed largest amount (63%-83%) to total rill width increment while bed incision contributed the largest amount (51%-65%) to total rill depth increment. Prediction equations for length, width and depth of a single rill on a loessial hillslope and empirical equations of rill erosion for the individual erosion processes were fitted and validated. Quantification and understanding of the contributions of individual rill erosion processes provides the necessary scientific basis for the development of process-based soil erosion models, and evaluation of practices for prevention of future land degradation.