We use cookies to make your experience of our website better. You can delete or block cookies, but some parts of this website won't work without them. By using this website you accept our use of cookies.


22/01/2015 10:39:58

We are modelling a gauge just upstream of a significant meander system which is believed to control water levels at the gauge until it is drowned out.  We have tried using Relative Path Lengths but the losses generated by this method are not significant enough to replicate observed flows and stages at the gauge.  Is there a better way to model the individual bends and has anyone looked into modelling bend losses using ISIS?

We only have survey at the gauge US of the meander system, 1 section on the last bend of the meander system and the next section is ~300m downstream.


23/01/2015 10:13:44

Best answer

Hi there,

Off the bat, I can think of three options for you, though none sound ideal from your current situation. Firstly, there are bend loss units in ISIS, which would do a much better job than RPLs, though they are 2-noded units like any structure and therefore require cross sections specified either side. If you don't have sections through the meanders then you'll need to generate some (either survey, interpolation or engineering judgement).

The second, better, though more difficult option is to represent this in 2D, as this would capture the losses associated with bends/meanders much more accurately and reliably than a 1D model, and the transition between low flow and drowned will be determined by the model directly rather than by yourself. This would also have the added benefit of getting the storage in the area correct, which extended cross sections, even with RPLs, would not on a meander system. The difficulty with this is obviously the need for yet more data to define your channel.

The final and perhaps simplest thught is that sometimes people represent sinuosity through an increase in Manning's n. I wouldn't care to advise on how much, but if you do this then make sure you only apply this locally to the area where the bends are, not globally as I have on occasion seen! The gauge you have upstream would probably serve well to guide you on values. Not a perfect fit for the physicalities of the system, but may help under the data-sparse circumstances.

Good luck and do let us know how you get on!


Damian Debski

23/01/2015 17:43:16

The Cowan method for estimating roughness values for channels includes a correction factor for the effect of meanders in the channel. The factor is a multiplier on the estimated value of Manning's 'n' for a nominally 'straight' channel. See 'Open Channel Hydraulics' by Chow (section 5-8 in my edition) for more details. The recommended multiplier value is between 1 and 1.3 depending on the severity of the meander(s) - higher values for tighter bends.

Sign in to create or reply to forum posts.

Privacy & Terms