PubAg

Main content area

Comparison of soil erosion models used to study the Chinese Loess Plateau

Author:
Li, Pengfei, Mu, Xingmin, Holden, Joseph, Wu, Yiping, Irvine, Brian, Wang, Fei, Gao, Peng, Zhao, Guangju, Sun, Wenyi
Source:
Earth-science reviews 2016
ISSN:
0012-8252
Subject:
bedload, biodiversity, land productivity, pollution load, prediction, rural poverty, silicon, soil erosion, soil erosion models, stream channels, suspended sediment, sustainable land management, China, Yellow River
Abstract:
The Loess Plateau suffers from severe soil erosion that leads to a series of ecological and economic problems such as reduced land productivity, exacerbated rural poverty, decreased biodiversity and sedimentation of the riverbed in the lower reaches of the Yellow River. Soil erosion models are commonly used on the Loess Plateau to help target sustainable land management strategies to control soil erosion. In this study, we compared eleven soil erosion models that were previously used on the Loess Plateau. We studied their prediction accuracy, process representation, data and calibration requirements, and potential application in scenario studies. The selected models consisted of a broad range of model types, structures and scales. The comparison showed that process-based and empirical models did not necessarily yield more accurate results over one another for the Loess Plateau. Among the process-based models, Si’ model, WEPP and MMF had the highest prediction accuracy. However, some of the selected models were tested with total sediment load while others were tested with suspended sediment load (i.e. bedload is not included), which is subject to several drawbacks. Research questions that each of the models can address on the Loess Plateau were suggested. Further improvement of soil erosion models for the Loess Plateau should concentrate on enhancing the quality of data for model implementation and testing, incorporating key processes into process-based models according to their aims and scales, comparing models that address the same research questions, and implementing internal and spatial model testing.
Agid:
5690025