intro story Lake


Lakes and reservoirs are important resources for communities around the globe. They provide food, drinking and irrigation water and can also offer spots for recreation. Under certain conditions lakes can also represent a threat for local communities, such as flooding or limnic eruptions. The ecological state of shallow lakes may be hampered by suspended sediments due to wind driven currents and waves. In deep lakes, stratification can cause undesirable anoxic conditions in the deeper areas. Understanding these systems is crucial to enable a sound management of these water bodies.

Delft3D is the state-of-the-art modelling framework for hydrodynamics, water quality, ecology, waves and morphology.
Here, in this space, we share knowledge & experiences, and discuss issues related to modelling of lakes.






Sub groups
D-Flow Flexible Mesh
Cohesive sediments & muddy systems


Message Boards

Delft3D FlOW: Error NaN after n time steps

Ulrike Kobler, modified 4 Years ago.

Delft3D FlOW: Error NaN after n time steps

Youngling Posts: 3 Join Date: 3/13/15 Recent Posts

I set up Delft3D FLOW using z-coordinates for a reservoir, but somehow the model is only running for several time steps and then the model produces an error: ERROR NaN found after n timestep(s).
I already reduced the processes, which are considered, and included either wind or temperature calculations and I also removed all in- and outflows, but the error remains the same. Prior to the error are no additional warnings.

One comment on the morphology of the water body itself: the southern part is really shallow (max water depth up to 4 m), while the northern part is at its maximum 24 m deep. The water level in general should vary between an elevation of about 889 and 876m. In the south of the reservoir the majority of the inflows are located

Are there any suggestions how to proceed?

Thanks for your help,

Phil Shepperd, modified 4 Years ago.

RE: Delft3D FlOW: Error NaN after n time steps

Padawan Posts: 33 Join Date: 5/31/11 Recent Posts
The error may be in one of your input timeseries, but I'm surprised that that isn't picked up at the start. So I wonder if it's a z-layer issue? Have you tried using sigma-layers? Should be rather more straightforward.