Effective way of dealing with a large number of lakes in a DEM?

Question

Hi,

I am working on an area in with a large number of lakes and interested in performing hydrologic modeling using ArcHydro. I was wondering what is the best approach for handling lakes when processing a DEM with ArcHydro? I tried different methods and many failed to give a reasonable result; flow direction with value ranging from 1 to 255, small number of watershed, etc. 
Here is the best approach I had so far: 
1.I Leveled the DEM using the lake polygon (the lake don???t have any level or height data), performed the DEM reconditioning using the vector stream data available, filled all the sinks in the resulting DEM, then run the regular flow direction and flow accumulation , etc. 
I also tried the following: 
2.I Leveled the DEM using the lake polygon, performed the DEM reconditioning using the vector stream data available, filled the sinks in the resulting DEM using the deranged polygon option and using the lake polygon shapefile, then run the regular flow direction operation. The result was a flow direction raster with values ranging from 1 to 255 instead of an esri type coded flow direction raster. I then run the flow accumulation many time but the operations never finished. 
I finally tried filling all the sinks in the DEM except the lake areas and run the flow direction, flow accumulation, and so on, but the result was similar to 2. 
I was wondering if someone could describe a better working approach for handling DEM in area with large number of lakes and depressions.

Thanks in advance

LornaMurison · Answer

Yea, I find the way ArcHydro talks about sinks is very complicated with "real" sinks, and pits and all that, but I think I'm starting to understand it.
Even though "real" sinks can exist I've never been able to get a working flow direction grid out of it.  And I'm also trying to understand whether a lake that actually has an outflow is considered a "real" sink, or whether it should be filled.
Anyway, there's a document attached to this thread: http://forums.esri.com/Thread.asp?c=141&f=1415&t=296006 that I found fairly useful.  There might be something in there that you haven't tried.  Let me know if you figure it out.

CheibanyOuldElemine · Answer

Have you tried just running fill sinks without any other preprocessing, and running the flow direction on that?
I've also been having the same issue, where no combination of preprocessing will result in a normal flow direction grid.

Thanks Lady_Jane! 
I did and obtained a working Hydro DEM, but this result is not compatible with the topography of my study area. Many of of the sinks identified are actually lakes, therefore legitimate natural features that I would like preserve in the analysis.

Thanks again

LornaMurison · Answer

Have you tried just running fill sinks without any other preprocessing, and running the flow direction on that?
I've also been having the same issue, where no combination of preprocessing will result in a normal flow direction grid.