SIMULATIONS DESCRIPTION

TEST IMPLEMENTATION: see that the matrices, eigenvalues, and output is correctly done
rectangular domain, 1x1 m^2, mixed-size, BC=downstream lowering, IC=equilibrium
	I1 2 size fraction, degradation into coarse substrate 				(well-posed)
	I2 2 size fraction, degradation into fine substrate 				(ill-posed)
	I3 3 size fraction, degradation into coarse substrate 				(well-posed)
	I4 2 size fraction, degradation into alternate fine-coarse layers 	(ill-well-posed)

TEST LINEAR ANALYSIS: see whether initial imposed oscillations grow or decay
rectangular domain, 1000xB m^2, unisize, BC=contant, IC=equilibrium + infinitesimal alternate bars
	L1 B=22, sec.flow=N (growth)
	L2 B=17, sec.flow=N (decay)
	L3 B=22, sec.flow=N (decay)

RESULTS OF LINEAR ANALYSIS: see whether it is ill-posed or well-posed
rectangular domain, 100x10 m^2, unisize, BC=constant, IC=equilibrium

bed slope effect
	B1	yes	(well-posed)
	B2	no 	(ill-posed)
	
secondary flow effect
	S1 no diffusion (ill-posed)
	S4 diffusion 	(well-posed)
	
ASHIDA
experiments by Ashida et al. (1990), sine-generated flume. 

secondary flow
	A1.1 (L013) no secondary flow 	(unrealistic)
	A1.2 (L015) Dh=1.6e-4 			(ill-posed)
	A1.3 (L016) Dh=0.50				(well-posed, too flat)
	
bed slope effects
	A2.1 (L017) Talmon				(ill-posed, but no oscillations)
	A2.2 (L018) Sekine				(well-posed, but no oscillations)
	A2.3 (L028) N=10				(unrealistic)
	
finer substrate
	A3.1 (L027)