Fig. 1. SDR versus unit discharge (a) for experiments conducted on a 2% slope with three different inflow sediment classes and (b) for experiments conducted with a medium inflow sediment concentration (67–92 kg m⁻³) on a 1 and 2% slope.

Fig. 2. Outflow sediment concentration of each settling velocity class versus time for an experiment conducted on a 2% slope with a unit discharge of 0.00080 m² s⁻¹ and an inflow sediment concentration of 76 kg m⁻³. Settling velocity classes are obtained by dividing the inflow sediment in 10 classes of equal mass.

Fig. 3. SDR versus unit discharge for seven grain-size classes for experiments conducted on (a) a 1% slope and (b) a 2% slope with a medium inflow sediment concentration (67–92 kg m⁻³).

Fig. 4. Surface flow velocities and regression relationships for experiments conducted on (a) a 1% slope and (b) a 2% slope with a medium sediment concentration (67–92 kg m⁻³) in the presence and absence of rainfall (6: no raindrop impacted flow; 1: raindrop impacted flow).

Fig. 5. Relative increase in unit discharge at the flume outlet due to rainfall impact (rainfall intensity of 45 mm h⁻¹).

Fig. 6. SDR versus unit discharge for both rainfall impacted and unimpacted experiments conducted on (a) a 1% slope and (b) a 2% slope with medium inflow sediment concentrations (67–92 kg m⁻³).

Fig. 7. Predicted mean velocity (Manning's equation) versus estimated mean velocity (measured leading edge velocity multiplied by correction factors proposed by Li et al. (1996)) for all rain impacted experiments.

Fig. 8. Objective function of the residues (MSE) versus the re-detachability value.

Fig. 9. Predicted versus measured overall sediment concentration using the optimised re-detachability value to predict the sediment deposition data used for calibration.

Fig. 10. Predicted versus measured sediment concentration for each of the 10 settling velocity classes using the optimised re-detachability value to predict the sediment deposition data used for calibration.

Fig. 11. Predicted versus measured overall sediment concentration using the optimised re-detachability value to predict the sediment deposition data used for validation.

Fig. 12. Predicted versus measured sediment concentration for each of the 10 settling velocity class using the optimised re-detachability value to predict the sediment deposition data used for validation.

Fig. 13. Predicted relative effect of rainfall re-detachment on the SDR versus unit discharge on (a) a 1% and (b) a 2% slope (c_in=80 kgm⁻³, Sed. type B); Threshold for re-entrainment is at 0.002 m² s⁻¹ and at 0.00095 m² s⁻¹ for, respectively, a 1 and 2% slope.

Fig. 14. Predicted (model predictions in presence/absence of rainfall) versus measured outflow sediment concentration for each settling velocity class for an experiment used in the validation process (q=0.00083 m² s⁻¹; c_in=43 kgm⁻³).

Table 1. Experimental and input parameters for the Hairsine et al. net deposition algorithm in the presence of rainfall

Table 2. Parameter ranges of the sediment deposition data used for calibration and validation of the multi-class net deposition equation in the presence of rainfall (n₀=number of experiments)

Table 3. Mean square error and mean absolute error for each settling velocity class and for the total outflow sediment concentration using the sediment deposition data selected for the calibration (n₀=10) and evaluation (n₀=9) of the multi-class net deposition equation in the presence of rainfall