Stream capture

Abstract: A simulation model of drainage network optimization is presented in which channels shift to minimize total stream power pgQS within the network. The simulation model starts from an arbitrary initial stream network developed on a square matrix, such as produced by random headward growth. Discrete stream capture then is simulated within the network, occurring wherever a new stream linkage would produce a steeper course than the original. Such capture produces a network with minimum power optimization but flow directions constrained to eight directions. Individual segment end points are then allowed to migrate by iterative relaxation with a direction and rapidity of motion governed by the gradient of stream power at the node. This valley migration is subject to the constraint that the sources and outlet remain fixed. The resulting networks are visually and morphometrically more similar to natural stream networks than the original networks produced by the random headward growth model.

Abstract: We use unique fluvial gravel deposits preserved atop a regional drainage divide to confirm the role of stream capture in driving ∼250 m of incision in the transient Roanoke River basin of the Appalachian Mountains (United States). Gravel provenance constrains the pre-capture position of the divide, indicating that ∼225 km 2 of basin area were abruptly connected to the base level of the capturing stream. The resulting wave of incision is currently manifest as major knickzones separating adjusting reaches from relict headwaters resembling streams of the New River basin, from which the Roanoke River was captured. The unusual preservation of the unconsolidated gravels on small relict surfaces adjacent to bedrock gorges indicates extreme spatial variability in erosion rates within the Roanoke basin, which is the first documented example of a transient passive margin basin connected to a capture event by stranded fluvial debris. Our results show the potential for stream capture across an asymmetric drainage divide to drive major transient incision independent of external forcings, such as climate change or tectonic uplift. A continuation of this process will lead to eventual capture of ∼7000 km 2 of the New River basin in the relatively near geologic future.

Abstract: [1] Geomorphic mapping of a ∼1 million square kilometer section of Terra Cimmeria on Mars (1:1M scale) indicates that prolonged, intense fluvial erosion occurred during the period of heavy bombardment. Crater counts date the termination of ubiquitous, intense erosion to the late Noachian, although some valleys may have continued downcutting into the early Hesperian. Stratigraphic and topographic relationships indicate that early erosional processes created large, integrated drainage basins, affected primarily by large impact basin structures and regional slopes. This terrain is not consistent with an origin purely by volcanic or impact processes. Cratering competed with drainage basin development, minimizing valley length, catchment area, and valley network integration. Drainage basin disruption resulted from impacts on valley thalwegs, or when the ejecta of large (usually >75 km) craters created low divides. Near-level upland intercrater plains are lightly dissected, in part because smaller craters could interrupt flow paths on gently sloping or flat terrain. Some closed drainage basins became integrated by continued erosion of drainage divides and stream capture, overflow of the divides, or headward growth of valleys fed by groundwater collected in closed basins. Drainage divide breaching was most effective on steeper (>0.5°) regional slopes, where observed drainage density is also highest. This is due to greater runoff volumes and velocity encouraging valley incision, as well as steep slopes allowing valleys to bypass or breach superimposed craters. Valley systems commonly extend nearly to the crests of sharp drainage divides, and spatially ubiquitous valley source points throughout the higher elevations suggest that runoff derived largely from precipitation.