[cameron]

It depends on a few things to determine if you need to extend the cross-sections or not.

How much area are you missing?
Are you doing a steady or unsteady flow analysis.
Do you need to generate a floodplain?

If the area you are missing is ineffective flow and you are doing a steady analysis then the results probably won’t change much, except for the added wetted perimeter.

It is so easy to do with GeoRAS or other methods that there is little reason for it not to be done. I would recommend to always extend cross-sections as that is proper modeling.

[cameron]

WGS 1984 is in degrees and is probably why you are having a problem. I would convert it to UTM or Stateplane.

If you are in the United States, you would check the USGS National Map website which can have down to 1/3rd arc seconds I believe.

If you are out of the United States, your could try SRTM or Aster Data. Another place to look is Intermap

http://www.intermap.com/en-us/home.aspx

[cameron]

What projection is the data in and what projection did you convert it too? What happens if you use the 3D Analyst tool in ArcMap to extract the elevations at one cross-section, does it give funny results also?

[cameron]

another option would be to use RAS Mapper to generate the floodplains and then import the shapefiles to ArcMap.

[cameron]

you could either delete all the other cross-sections in the geometry you are importing in or during import only select the bridge to import.

[cameron]

You could just do the culvert, but would need to include a few cross-sections upstream and downstream of the culvert. Generally with structures you need at least two upstream and two downstream, I usually do a couple more than that though. If there is anything that would affect the tailwater of the culvert downstream (another culvert, bridge, ect..), you would need to include that in the model. Another option would be to use HY-8 instead of HEC-RAS to design the culvert.

[cameron]

You need to model them with a junction.

Can you explain the second question in more detail?

[cameron]

You need to make sure you have cross-sections at grade breaks, change in cross-sections areas, change in roughness, and around bends. Below is a link that discusses a way to determine how many cross-sections based on bank full depth and slope. It is also important to make sure that the downstream boundary (assuming sub-critical flow) is far enough away to that it does not impact the area of interest.

http://hecrasmodel.blogspot.com/2008/12/samuels-equation-for-cross-section.html

[cameron]

A – When I say burn, I mean combine the channel and overbank data. You can use ArcMap or other programs to combine the channel data with the other topographic data. If all you have is survey data then it is usually easiest to incorporate the data into the cross-sections in HEC-RAS using the Graphical Editor Merge Options.

B. It is always best to a have a georeferenced model. Since you have the xyz data for each cross-section you can go to the gis tool|xs cutlines table in the geometry editor and specify the xy coordinates for each cross-section (make sure it is left to right).

C. In arcmap, one option would to create TIN in arcmap from the survey alone and need to incorporate breaklines to connect cross-section to cross-section. Once this is done you could then mosaic the channel surface to overbank.

D. http://resources.arcgis.com/en/help/main/10.1/index.html#//0017000000m6000000

[cameron]

There is free software called QGIS or GRASS that can be used instead of ArcGIS and has similar features, just not as easy to use.

There is also a website called http://spatialreference.org/ where you can type in the coordinate system the DEM data is in and download a *.prj file (projection file) that HEC-RAS Mapper can read. The USGS National Map website should contain the projection that the DEM is in.

If the DEM does not overlay with the HEC-RAS model, you will need to convert the DEM to the correct coordinate system using software.

If the model only covers the 2D area, the model should be able to create the 2D volume curve data. If you turn off floodplain mapping it may or may not run, I have not tested this yet.

In Mapper you can actually create a surface from either the entire cross-sections or just from the channel (between banks). Once this is created you can than use the cross-section/channel surface and DEM to create one surface in Mapper.

[cameron]

Terrain resolution all depends on what you are trying to accomplish. If you are doing a dam breach study, than you could use very coarse terrain (USGS 30m DEM), you just may need to burn in a channel. I have done many studies using Intermap topographic data which uses a 5m DEM for 2D floodplain modeling. You would want to make sure the resolution picks up all important features or else modification will need to be done to the 2D grid to incorporate the missing data.

For 1D modeling it is best to get actual survey if possible, but it also depends on what type of study it is. Often what happens is you get survey for the channel, use lidar for the floodplain, and incorporate them together into a DEM. The terrain resolution needs to be good enough to pick up critical features. If it does not, it will need to be supplemented with levee features and modifications to the channels

As for importing shapefiles, you bring them in the same way as an image and can digitize the 2D area. Another option would to use HEC-GeoRAS and import the 2D area as a storage area and convert it to a 2D area. If you don’t have GeoRAS, you could first create a dummy 2D area, than extract the x and y coordinates of the 2D (in GIS) and import that into HEC-RAS using |GIS|Storage Area Outlines Table in the geometry editor window (needs to be less than 500 points though).

[cameron]

You can import data from other HEC-RAS geometry files using the File|Import Geometry|HEC-RAS Format in the geometry edit. If the stationing is different between geometries you would need to update the stationing of the bridge/culvert data.

[Author]

Posts