Bacis Step-By-Step

[Anthony_Passaro]

I have used Autocad for some time, but just starting to use Civil 3D.

I have 3 sets of data for a town:

1. DEM Files for the contours across the town, which I downloaded from the county GIS website.  The are added to the drawing with the Add Surface...DEM command

2. Orthophotographic "pcitures" (in .sid) format from the same county GIS website.  These are done with map import.

3. A .txt file of information showing manhole numbers, northings, eastings, rim elevations, etc...These are added with an add points command

I have been able to import all 3 into a drawing, but, I am not confident that I am doing it correctly.  The contours appear to be in the correct locations becuase all of the embankments along roads and bridges appear to line up.  I have been able to import the .txt file with the manhole infomation, and the northing and easting put the point very close (within 6" in plan) of the actual manhole locations.

BUT, when I check the elevation of the surface near the manhole, Autocad is telling me an elevation that is anywhere from a few inches to several feet different than the surveyed mahole elevation.  And the difference is not consistent from one manhole to the next.

I would like to know if there is a correct "sequence of operations" for loading these three items into a drawing and if I am doing something wrong while loading them.  Thanks for the help...

[Jay_B]

The DEM data is considered rough in terms of accuracy and will vary. several Feet isn't uncommon.

If the text file you have was surveyed then that is very likely of a much higher accuracy.

If you need a high level of accuracy you may need more survey information collected but only you can answer that?

If your satisfied with the DEM data in terms of a surface and you only have random survey shots on the manholes then
You may want to consider removing those anthills and valleys from the surface as they may do mare damage then good
given the vertical differences.

[Neilw_05]

I suspect your problem is with the DEM. DEM's are typically not as accurate as a surface model built from topographic survey data.

So the first step would be to find out if the manhole elevations are supposed to coincide with the DEM. If so then the next thing to look at is whether the DEM truly represents the survey.

Typically a DEM just a grid of sample points of the true surface at some interval and thus they do not pick up significant changes in the terrain between the points. You'll want to get the points and breaklines from the survey to build an accurate surface.

[neilyj666]

As with the two other responses, I wouldn't use DEM data for anything more than a feel for the general terrain i.e useful for hydrological analysis at the catchment scale but not much use for detailed work as the accuracy just isn't there as breaklines etc are absent

[AllenJessup]

Also. Check the MetaData for the DEM. It should be available on the same website. That should tell you the expected accuracy of the DEM. A DEM can vary quite a bit depending on it's source. Some older ones are from digitized USGS Quad maps. Newer ones can be more accurate. It's always best to check.

Allen Jessup

[Anthony_Passaro]

Thank you all for the replies.  That makes sense.  I should have realized that the DEM files are not too accurate.

Now, can I add survey data (N, E, Z data) to the surface and make it more accurate?

[Neilw_05]

You'd have to convert the DEM to XYZ points if you wanted to merge more points in. Instead, create new surfaces from your supplemental data and paste them into the DEM or a composite surface.