Structural control on drainage network orientation in the north East African rift: the case of northern main Ethiopian rift

This paper evaluates relationship between Cenozoic rift induced geological structures and stream network orientation at a regional scale in the northern Main Ethiopian rift (MER). The northern MER is located at the heart of the tectonically and volcanically active northern tip of the great East African rift system. The region is broadly fractured by two tectonically categorized structures such as marginal faults and Woniji Fault belts (WFB) which is manifested mainly along the rift axis. Geologically, the rift is bounded by an uplifted Oligocene-Miocene volcanic product. Rift margin to rift axis transitional zone is covered by Pliocene-Pleistocene basalts. The rift axis of this region is dominated by geochemically bimodal quaternary volcanics with minor sediments. The area does not have extensive unconsolidated sediment or deep weathered regolith except some volcanic ash coverage in minor areas. Because of this and the active morphological features of the faults, the geological setting is quite clear for the quantification of structural features and their relationship with the stream network orientations. The degree of correspondence between the orientation of stream network segments and structural features of the faults investigated in two study areas. The first order streams are the less controlled streams either by marginal faults or by axial faults in both study areas. The fourth order streams are strongly controlled by the orientation of marginal structures. The higher the stream network order, the weaker correlation with the structural orientations of the region. However, there is strong structural influence on the fourth order stream flow orientations mainly by marginal faults. Statistically significant correlation observed between the WFB and stream flow direction along the rift floor. The structural control of stream flow orientations in the northern main Ethiopian rift varies from 20% to 65% of the normalized channel length. The lower limit is considered to be the effect of stream incision across the slope gradient of fault scarps whereas the upper limit is considered to be the effect of larger displacement and longer marginal faults.