reason for tight meander bends?

[Carolyn Cload]

good afternoon 

I am trying to replicate a set of flume experiments using CL. The data I have on the flume experiments is from a 2007 paper (ref Bertoldi & Tubino River bifurcations: Experimental observations ) . One difference between my models and the original flume results is that the meanders formed in the model are much tighter - it seems that the original meanders had a wavelength of at least 2m whereas mine are generally 0.8 - 0.9m long. The model starts with a plane bed and creates its own channel pattern.  

Because this is all very small scale, I have used a very small grid cell size of 0.03m to get a reasonable number of cells across the channel, which is 0.35m wide. I am aware the model was not designed for cells this small, so that may be affecting its behaviour. But another modeller working with my supervisor has found a similar outcome when modelling a  section of a real river - the meanders have a shorter wavelength in the model.  

Is there a parameter in the model which is controlling the meander size and is it possible to adjust?

Thanks for your help, 

Carolyn 

[Tom]

Sorry for late reply... 

You can change the 'number of passes for edge smoothing filter' which will reduce the pointyness of meanders.. If you have 10 cells across the channel - and with a meander wavelength of 3-5 times the channel width, this figure will need to be at least 30-50 if not more... 

Personally, I'd use a larger grid cell size so the channel is represented by 2-3 cells across. The model won't give you any more detail by having 10 cells across, but will just take longer to run...

Tom

[Carolyn Cload]

hi Tom

thanks for the suggestions, I have tried them both. I increased the "number of passes" which did slightly increase the meander length, however it seemed to get unstable at about 60.  

I tried incresing the cell size so I only had 3 cells across the main channel. It certainly made the model speed up a lot, but at this resolution I cannot see any meandering effects. The sediment outflows are around twice as high. 

I am also looking at the balance of flows between the two bifurcates. To estimate the flow in each channel, I choose a cross section and mulitply the velocity in each cell by the water depth and the cell width, and sum these to get a calcuated discharge. At the larger cell size this estimate becomes very inaccurate. So for my purposes the smaller cell size is more useful.  

best regards

Carolyn