Identification of sediment sources in agricultural catchment of the Pernambuco state, Brazil

Abstract

Poor land management is among the controlling factors causing an increase in erosion rates and sediment transport in river catchments. Sediment source fingerprinting is increasingly used to investigate key sources of elevated sediment loss. The objective of this study was to evaluate the ability of colour proprieties, Near Infrared Spectroscopy (NIR) and Mid Infrared Spectroscopy (MIR) to apportion land use sediment sources (sugarcane (SC), unpaved roads (UR), cropland (CL) and channel banks (CB)) at sub-catchments (1.338 km2 and 1.453 km2) and catchment (2.857 km2) scales using both suspended (SS) and bed (BS) sediment samples. The results for the colour properties, NIR and MIR showed that the largest sources that contribute to sediments on the catchment-wide scale are the CB that contribute mostly to the discharge of SS, followed by the source of SC with the highest discharge of BS, for the properties organic compound with NIR spectroscopy, soil components with MIR spectroscopy and the combination of colours, NIR and MIR spectroscopy (CNM). Using the spectra of organic compounds from the NIR as tracers proved to be robust and suitable for discrimination between sediment sources in both sediments and basin scale size, with errors (RMSE and MAE) below 8%. We conclude that it is urgent to use management practices, such as the revitalization of areas of permanent preservation that comprise the set of riparian forests, slopes and headwaters of rivers, together with conservationist practices that control erosion and torrents in the agricultural perimeter of the hydrographic basin to reduce the environmental impact. impacts caused by the production of sediments for rivers.