The importance of initial sediment conditions on model calibration and validation is analysed. A sediment model was calibrated and validated under three different initial sediment conditions: (0) no sediment availability, (1) calibration of the initial sediment condition and (2) using a warm-up simulation. The model results were assessed in terms of the graphic of fine sediment transport, or sedigraphs, and the visual fit of the hysteresis on the sediment rating. All strategies provided adequate results. However, the loop rating curve analysis demonstrated that the choice of initial sediment conditions affected the simulation results. Without any initial sediment condition, the model results were typically inferior to the simulation results with calibration or warm-up. The calibration of initial conditions proved to be the most reliable technique to generate clockwise hysteresis loops, but failed in reproducing other loop types. Overall, the warm-up simulations showed encouraging results, providing satisfactory fine sedigraph simulation results. (C) 2014 Elsevier Ltd. All rights reserved.

This study was supported by the Spanish Ministry of Economy and Competitiveness through the research projects FLOOD-MED (CGL2 008-06474-C02-02/BTE), Consolider-SCARCE (CSD2009-00065) and ECOTETIS (CGL2011-28776-C02-01).