Watershed delineation

[sdevin]

I just tried using the watershed analysis routine. It left me somewhat confused. The model is in mountainous terrain and was developed from 24k contours. I would like to be able to pick a spot on the surface and have the software delineate the drainage basin discharging at this point. Can this be done?

Steven C. Devin, P.E.
Consulting Engineer
Civil 3D 2007
Dual Xeon 3.2 MHz
3 GB RAM
Nvidia Quadro FX3400 256MB

[Anonymous]

sdevin wrote:
> I just tried using the watershed analysis routine. It left me
> somewhat confused. The model is in mountainous terrain and was
> developed from 24k contours. I would like to be able to pick a spot
> on the surface and have the software delineate the drainage basin
> discharging at this point. Can this be done?

It's on the wishlist.

--
Jason Hickey

Civil 3D 2007, SP3
Dell Precision M70
2 GIG RAM, 256 MB nVidia Quadro FX Go1400
Intel Centrino 2 gHz Processor

www.civil3d.com

[Anonymous]

Hi Jason,

On the wish list or not, it's a fantasy. Any one who tells you they can do
it has no knowledge of TIN models.

Unless you pick a point which is exactly on a line in the model where the
two triangles on each side of the line slope towards the line forming a
valley, the catchment area to a point is a line. And, as you know a line
has zero area.

You could make some sort a guess and get a catchment area which drains to
the triangle which you pick - or some other convention such as an assumption
of water depth at the point, or a nominal length of gutter at the point, but
would you rely on something like this without knowing what the convention
was and whether that would be realistic in the circumstances of your design?

The watersheds defined for the TIN model are areas bound by "ridge lines" in
the model. They are a guide to the user in as much as water from outside a
catchment cannot get into the catchment because it would have to flow over a
ridge line to do so.

Obviously in practice, water in flat country can cross catchment boundaries
as the water depth can exceed height of the ridge line above the invert in
the catchment area.

As an example of this, I once had to analyse a "maximum probable rainfall"
event. A rail embankment varying from 3 to 7m high (nominally horizontal
for the trains) ran across several valleys with 6m by 4m culverts at the
valleys. These culverts were no where near adequate for the event I was
analysing and water would have flowed between the valleys had the storm
event occurred.

At it's peak, it is likely the water would have flowed about a metre deep
over the rail line over a length of about 6km and washed significant
portions of the embankment away.

--

Laurie Comerford
CADApps
www.cadapps.com.au
www.civil3Dtools.com
"Jason Hickey" wrote in message
news:5462837@discussion.autodesk.com...
sdevin wrote:
> I just tried using the watershed analysis routine. It left me
> somewhat confused. The model is in mountainous terrain and was
> developed from 24k contours. I would like to be able to pick a spot
> on the surface and have the software delineate the drainage basin
> discharging at this point. Can this be done?

It's on the wishlist.

--
Jason Hickey

Civil 3D 2007, SP3
Dell Precision M70
2 GIG RAM, 256 MB nVidia Quadro FX Go1400
Intel Centrino 2 gHz Processor

www.civil3d.com

[sdevin]

I believe that HEC-GeoHMS has some sort of routine for delineating watersheds. It pre-processes the surface model and burns in flowlines. It's been a while since I considered using that software package but I remember that this was one of the features that intrigued me especially for a large basin (>100 mi^2) with multiple sub-basins.

[fcernst]

USGS StreamStats has a nice watershed delineation tool I use. I picked the outlet point at the top there where I drew the red arrow. It delineated this watershed in about 9 seconds. One pick... no tedious waterdrops required.

I would like Autodesk to put an inquiry to the USGS about this technology for use in Civil 3D terrain models.

[alexandre_ssp]

good software Aquaveo WMS 

[rubaru]

Fred --

That looks like a really great tool...where is it on the Stream Stats website?

Jason

[jmayo-EE]

It's the little black dot button. This is a great tool. We have have it in NJ for a couple of years now.

[Joe-Bouza]

Hey John, Happy new year.

Stream stats a tool to get?

I have always understood watersheds to be nebulous so what about catchments?

[jmayo-EE]

Happy New Year Joe!

Stream Stats is free. It is available in NY. Other users can check here to see if it's available.

http://streamstats.usgs.gov/ssonline.html

It's very nice. Just as Fred said. Click a point and the watershed is delineated. Peak flows for a range of storm events can also be computed for you.

[Matt.Anderson]

StreamStats computes basins using preprocessed DEM data and streamlines. Those two computations are a bit computationally intensive.

Catchments operate on TINs, and the calculations are quite a bit different than the ArcHyro D8 calculations that StreamStats uses.

[dseatter]

Great site Fred, thanks for sharing.

I just generated a watershed boundary & imported into C3D in about 5 minutes.

Looking forward to this saving me the tedious task of tracing a watershed boundary by hand.

[jmayo-EE]

Thanks for the additional info Matt. The thing we like most about Stream Stats is the DEP has been accepting the data. No more methodology games. 🙂