Assessing Rates of Divide Migration in Reversed Drainages Based on Dating of Abandoned Terraces in the Negev Desert

Drainage divides define the boundaries between neighboring basins. In landscapes subjected to climatic, tectonic, or lithologic perturbations, gradients in erosion rates across divides may promote divide migration through time, which modifies the basins geometry. Despite the key role of divide migration processes in landscape evolution, field-based measurements of migration rates remain rare, most likely because potentially datable geomorphic markers are often erased during the migration processes. We present an exceptional field case from the Negev desert, Israel, where migration rates are inferred from in situ terrace dating. In this site, a divide migrated along an antecedent valley, promoting the development of a reversed channel that incises into the sedimentary fill of the antecedent valley, and generates terraces along the channel. The migration and reversal processes were inferred based on three observations: (1) A barbed tributary, whose orientation preserves the antecedent drainage geometry; (2) The local divide is a valley confined windgap; and (3) The terraces grade in the direction of the antecedent valley, opposite to the active flow direction. We hypothesize that the abandonment age of the terraces reflects the timing at which the windgap migrated past the terrace location, promoting the reversed channel incision that disconnected the terrace from the active drainage. According to this conceptual model, the age of the terrace is expected to increase with distance from the windgap along the reversed channel. We estimated the abandonment age of seven terraces, extending along 350m from the windgap, by relative and absolute dating. Relative dating was based on the characteristics of the terraces surface and of the chronosequence of reg soil that develops on the surface (i.e., dust content, salts depth and concentration). Absolute dating was determined by luminescence dating of dust that accumulated on the terraces after their abandonment. As the model predicts, both dating methods show that the age of the terraces generally increases with distance from the divide. Luminescence ages range between 27 to 260ky, suggesting an average divide migration rate of 0.66-1.54 m/ka. Our approach allows, for the first time, a high-resolution record of lateral divide migration based on dated field-markers.