Simulating Spatial Variability of Fluvial Sediment Fluxes Within the Magdalena Drainage Basin, Colombia.

This study explores the application of an empirical sediment flux model BQART, to simulate long-term sediment fluxes of major tributaries of a river system based on a limited number of input parameters. We validate model results against data of the 1612 km long Magdalena River, Colombia, South America, which is well monitored. The Magdalena River, draining a hinterland area of 257,438 km2, of which the majority lies in the Andes before reaching the Atlantic coast, is known for its high sediment yield, 560 t kg- 2 yr-1; higher than nearby South American rivers like the Amazon or the Orinoco River. Sediment fluxes of 32 tributary basins of the Magdalena River were simulated based on the following controlling factors: geomorphic influences (tributary-basin area and relief) derived from high-resolution Shuttle Radar Topography Mission data, tributary basin-integrated lithology based on GIS analysis of lithology data, 30year temperature data, and observed monthly mean discharge data records (varying in record length of 15 to 60 years). Preliminary results indicate that the simulated sediment flux of all 32 tributaries matches the observational record, given the observational error and the annual variability. These simulations did not take human influences into account yet, which often increases sediment fluxes by accelerating erosion, especially in steep mountainous area similar to the Magdalena. Simulations indicate that, with relatively few input parameters, mostly derived from remotely-sensed data or existing compiled GIS datasets, it is possible to predict: which tributaries in an arbitrary river drainage produce relatively high contributions to sediment yields, and where in the drainage basin you might expect conveyance loss.