Assessment of ephemeral gully erosion using topographic and hydrologically based models in Central Kansas

Abstract

The global requirements for food and agricultural products have increased enormously in recent years mainly due to increase in global population. More land is brought under human development and cultivation including marginal lands that are susceptible to degradation processes of erosion, waterlogging, and depletion of organic matter. The resulting effects include; deprivation of the roles performed by the environment, high costs of water treatment, and sedimentation of water reservoirs. This study aims at assessment of ephemeral gully (EG) erosion using topographic and hydrologically based models in two paired watersheds in Central Kansas. The effects of best management practices (BMPs) implementation on EG formation, and erosion rates within the watershed are discussed. The topographic index (TI) models used include: slope area model (SA), compound topographic index model (CTI), wetness topographic index model (WTI), slope area power (SA2), kinematic wave model (nLS), and modified kinematic wave model (nLSCSS). EGs predicted by each model threshold were compared with observed EGs obtained through digitization and field reconnaissance. The agreement of thresholds obtained from location and length approaches were compared by means of drainage density concept. Statistical analysis was performed by error matrix for EG location analysis, and root mean square error (RMSE) and Nash–Sutcliffe efficiency (NSE) for EG length analysis. A TIN-based real-time integrated basin simulator (tRIBS) model, a physically-based, distributed hydrological model was coupled with an EG erosion component (Foster and Lane model) to estimate the erosion rates, and effect of installation of BMPs on reduction of EG erosion rates from agricultural fields. The results indicated that TI models could predict EG location with a maximum total accuracy of 70%. The effectiveness of TI models at prediction of EGs is affected by watershed features such as installed structural best management practices, roads, and culverts. The CTI model outperformed all the TI models at prediction of EGs with maximum Kappa and NSE values of 0.32 and 0.55 respectively, and a minimum RMSE value of 0.087 m. Structural BMPs are effective at controlling erosion from croplands, however, the effectiveness of structural BMPs at reduction of sediment loadings from EGs vary depending on surface cover, and BMP geometry.