The Influence of Antecedent Soil Moisture Conditions on Inter-rill Soil Erosion

The Influence of Antecedent Soil Moisture Conditions on Inter-rill Soil Erosion Ilja van Meerveld Ilja.van.meerveld@vu.nl

Acknowledgements p Andrew Barry p Graham Sanders p Jean-Yves Parlange p Seifeddine Jomaa p Marc Parlange

Introduction p The soil s inherent ability to resist erosion varies temporally due to differences in n Temperature n Time since ploughing n Microbial activity n Soil organic matter content n Soil moisture n Etc. p However, n Few studies have looked at the effects of antecedent soil moisture conditions on inter-rill soil erosion n In many soil erosion models soil erodability parameters are considered constant

The EPFL erosion flume

Flume experiments p 11 high intensity experiments p Loam soil p Pore pressure (3 locations, 4 depths) p Overland flow, vertical and lateral drainage n 2 l buckets with water level recorders

Flume experiments p 11 high intensity experiments p Loam soil p Pore pressure (3 locations, 4 depths) p Overland flow, vertical and lateral drainage n 2 l buckets with water level recorders p Sediment concentration in overland flow

Results: Influence of antecedent conditions Sediment concentration (g/l) Overland flow (mm/hr) Pore pressure (m) 45 4 35 3 25 2 15 1 5 5 4 3 2 1. -.2 -.4 -.6 E1: dry E2: wet E8: nearly saturated at.5 m below the soil surface 1 2 3 4 5 6 7 8 9 1 11 12 Time since start of experiment (min)

Results: Sediment concentrations Peak sediment concentration (g/l) 8 6 4 2 2 4 6 8 1 12 14 2 4 6 8 1 12 14 25 2 15 1 5 Steady state sediment concentration (g/l) Slope (%) Slope (%)

Results: Peak sediment concentrations Peak sediment concentration (g/l) 8 6 4 2 2% slope 7% slope 12% slope -2. -1.5-1. -.5. Pore pressure at start of the experiment (m) 2 4 6 8 1 Steady state infiltration rate (mm/hr)

15 2-2 µm 1 5 1.2 2-5 µm 1..8.6.4.2..8 5-1 µm.6.4.2. 1 2 3 e since start of the experiment

Influence of soil moisture and slope on size distribution at peak concentration Percentage of total sediment <5 µ m during peak sediment concentrations (%) 1 9 8 7 2% slope 7% slope 12% slope 6-2. -1.5-1. -.5. Pore pressure at the start of the experiment (m)

Simulating these results: Parlange et al., 1999 JoH p Decoupling of the short and long time behavior. p Very short time solution for the processes associated with the rainfall impact p Longer time solution for the behavior controlled by convection

Hairsine-Rose model Figure from: Heilig et al., 21, JoH p Water conservation: p D q + = R t x Sediment conservation: DC t M t i di qc + x = d i i e = e + e di i di d i RC i p p p Rainfall detachment: ap e i = [ 1 H( t, x) ] I Rainfall re-detachment: M di edi = ad P H t, M Deposition: d =V C i i i dt ( x)

Optimization of the unknown variables p Four unknown variables: n a d n α (= a d DP/RM dt* ) n K (=a d /a) n D n Affects the peak and steady state concentrations n Affects the peak sediment concentration n Affects the rise of sediment concentrations n Affects the sharpness of the recession (fine sediment)

Model results 2 2-2 µm 1-315 µm 2. 15 1.5 1 1. 5.5 Sediment concentration (g/l) 5 4 3 2 1 15 1 5 Exp. 4 Total concentration Observed Fitted < 2 µm 3 2 1 1.5 1..5 2-5 µm 5-1 µm 315-1 µm >1 µm. 1.5 1..5. 1.5 1..5 2 4 6 8 1 12. 2 4 6 8 1 12 Time since start of the experiment (min) 2 4 6 8 1 12.

Effect of antecedent moisture conditions on model parameters p apd d α = RM dt* 5 4 2% slope 7% slope 12% slope 3 α (-) 2 E1 1 E9-2. -1.5-1. -.5. Pore pressure at the start of the experiment (m)

Incorporating the effects of antecedent moisture conditions into the model p Incorporating the effects of antecedent moisture conditions (and slope) into the model p First step simple parameter regression a d = 2.14S+.69, r 2 =.93 α = 281.6ψ + 113.9.5 r 2 =.73 S = slope [%] Ψ.5 = pore pressure at.5 m at start of the experiment [m]

Model results: Total concentration Sediment concentration (g/l) 4 3 2 1 Exp. 4 Total concentration Observed Fitted α and a d from regression 2 4 6 8 1 12 Time since start of the experiment (min)

Model results: application of the regression 2 2-2 µm 1-315 µm 2. 15 1.5 1 1. 5.5 Sediment concentration (g/l) 5 4 3 2 1 15 1 5 Exp. 4 Total concentration Observed Fitted α and a d from regression < 2 µm 3 2 1 1.5 1..5 2-5 µm 5-1 µm 315-1 µm >1 µm. 1.5 1..5. 1.5 1..5 2 4 6 8 1 12. 2 4 6 8 1 12 Time since start of the experiment (min) 2 4 6 8 1 12.

Conclusion: Experimental results p Antecedent moisture conditions have a large effect on peak sediment concentrations, especially near saturation p Wetter conditions lead to a disproportionate large increase in transport of fine particles n Important implications for simulating the transport of sorbed particles (phosphorus, pathogens, etc) p Effect of antecedent moisture conditions on inter-rill erosion is possibly different for dry soils

Conclusion: Model results p Experimental observations do agree with the Hairsine- Rose theory n Initial transport of fine particles and coarsening with increasing time n Steady state composition reflects the original soil composition p The α parameter in the model appears to be influenced by the antecedent moisture conditions p Using a simple relation between initial pore pressures and the α parameter yields good results regarding the influence of antecedent moisture conditions on sediment concentrations and composition

Thank you From: http://www.public.asu.edu/%7emschmeec/rainsplash.html