Geosciences 2013, 3(1): 1-12 DOI: 10.5923/j.geo.20130301.01 Connectivity Aspects in Sediment Migration Modelling Using the Soil and Water Assessment Tool Jacobus J. Le Roux 1 , Paul D. Sumner 1,* , Simon A. Lorentz 2 , Talita Ge rmis huyse 3 1 Department of Geography, Geoinformatics and Meteorology, University of Pretoria, Pretoria, 0002, South Africa 2 School of Bioresources Engineering and Environmental Hydrology, University of KwaZulu-Natal, Scottsville, 3209, South Africa 3 Golder Associates Africa (Pty) Ltd, M idrand, 1685, South Africa Abstract Sediment migration modelling at the catchment scale is complicated by various connectivity aspects between sources and sinks, including the extent that sediment generated on hillslopes is connected to a channel and linkage within a channel network. The Soil and Water Assessment Tool (SWAT) is applied within the context of connectivity in a catchment (Mkabela near Wartburg, South Africa) with identified source (cabbage plot) and sink (farm dams and wetlands) zones. The study illustrates SWAT can be applied in scenario analysis to assess connectivity aspects in sediment migration modelling. Scenario analyses establish the extent that sediment outputs from the cabbage plot create input for downstream sub-catchments, as well as the impact of farm dams and wetlands on sediment yield at the catchment scale. SWAT effectively identifies the cabbage plot as an important source of sediment at sub-catchment scale, but the sediment is not spatially identified within the sub-catchment where it is located and all the sediment is modelled to reach the channel, whether connected or not. Despite this, no significant changes are simulated by SWAT at the catchment outlet since increased discharge and sediment load from the cabbage plot is counterbalanced by sinks at the catchment scale. The effect of sediment sinks becomes dominant over sediment sources with increasing spatial scale. The channel serves as an important sink zone due to its relatively rough surface conditions. The model also appears to be efficient in representing farm dams as a series of storages where connectivity is reduced at the catchment scale, but sediment deposited in farm dams mainly originates from surrounding sugarcane fields, not the cabbage plot. SWAT could not correctly identify wetlands as sink zones for cabbage sediment since, in contrary to farm dams, wetlands in SWAT are simulated off the main channel and water or sediment flowing into the wetlands must originate from the sub-catchment in which they are located. The suitability of SWAT for use in connectivity studies is discussed in the context of these findings. Keywords Sediment Connectivity, Source-Sink Zones, SWAT Model, Catchment Scale, South Africa 1. Introduction Water scarce countries such as South Africa are increasingly threatened by pollution and sedimentation of water bodies due to suspended sediment concentrations in streams which affects water use and ecosystem health (e.g.[1]-[5]). It is imperative to devise the means through which these problems can be controlled but prevention and remediation relies largely on the understanding of factors controlling the sediment dynamics in a catchment, including sediment generation, transport and deposition ([6],[7]). The term connectivity is used to describe the extent to which sediment generated on hillslopes is connected to a channel by overland and subsurface flow, as well as the linkage of streamflow and sediment within a channel network ([2],[8],[9]). Connectivity aspects from hillslopes to * Corresponding author: paul.sumner@up.ac.za (Paul D. Sumner) Published online at http://journal.sapub.org/geo Copyright © 2013 Scientific & Academic Publishing. All Rights Reserved channels, as well as channel connectivity downstream needs to be considered. Good vegetation cover usually reduces connectivity fro m hillslopes to channels[2], whereas different sinks reduce connectivity within channels ranging from partial retention in small wetlands[10] to full blocking in large reservoirs[9]. At the catchment-scale, connectivity aspects are driven by complex physical processes that involve interaction of a large number of spatial and temporal factors that cannot be monitored directly[4]. Spatial and temporal variability poses a severe limitation, not only for local-scale measures, but also for procedures with a lumped nature, such as sediment rating curves and sediment delivery ratios that do not take connectivity aspects into account ([11],[12]). Assessments are usually carried out by means of semi-distributed models such as the Soil and Water Assessment Tool (SWAT)[13]-[15]. Semi-distributed models such as SWAT, however, employ certain compromises or assumptions that disregard connectivity aspects[11]. In this context, the aim of the study is twofold: to apply the SWAT model in scenario analysis to assess sediment