How River Avulsions Build Floodplains

Abstract:

The repositioning of a river channel on its floodplain, or avulsion, typically occurs in two general styles that are linked to different floodplain morphologies and facies assemblages. Avulsions that divert flow directly into pre-existing channels with little to no floodplain sedimentation are termed annexational, whereas avulsions that create advancing distributary networks are termed progradational. While most avulsion events are documented after they occur, recent developments in remote sensing allow us to isolate and measure individual processes that create avulsions and build floodplains. Using a tasseled-cap transformation of Landsat data, we track annual land-cover changes in 63 avulsions in the Andean, Himalayan, and New Guinean foreland basins. We find that in many cases, a single avulsion event is a succession of both annexational and progradational processes. For instance, on the Ramu River, Papua New Guinea an avulsion that began as a prograding distributary network became annexational when it encountered an accommodating relict channel. Conversely, on the Moleto River, Bolivia one event began with annexation of a small adjacent channel. Over time the proximal portion of the annexed channel adjusted to higher discharge by widening and increasing meander amplitude and wavelength. Further downstream, channel adjustment ceases, and is replaced with dramatic dechannelization which causes overbank flooding. Remarkably, this dechannelization propagates upstream—an avulsion process we term retrogradation—until flow is directed out of the annexed channel through a crevasse splay that creates a prograding distributary network. Finally, this event ends as the distributary network collapses to a new primary channel, which promptly annexes a nearby channel. We find that of 63 mapped avulsions, 21 exhibit some combination of annexation, progradation, and/or retrogradation. Our analysis reveals that floodplain avulsion deposits can be spatially disconnected when progradational reaches are separated by annexational reaches. These results have important implications for the interpretation of fluvial stratigraphy and suggest that avulsion deposits may not always conform to end-member styles, but rather contain a variable expression of different avulsion processes.