## Tuesday, July 31, 2012

### How to draw cross sections.

Written by Chris Goodell, P.E., D. WRE

Cross sections must be perpendicular to the flow lines at all locations.  And they cannot intersect with each other.  That is why it is common to see cross sections snap at different angles outside the main channel (we call this doglegging).  The trick is to keep them from intersecting, while also staying perpendicular to flow lines.  In the figure below, the dark blue line represents the main channel.  The brown lines represent the edge of the flood plain.  The light blue lines are my impression of the flow lines through this terrain, if water were flowing appreciably in the floodplain.  The green lines are cross sections.  Notice that the cross sections are drawn so that they are not only perpendicular to the main channel, but also to my perception of the flow lines in the floodplain.  It can be very helpful to draw these flow lines before cutting cross sections.

It takes a little bit of practice to do this correctly, and most of the time some trial and error, but as long as you remain perpendicular to the flow lines and don’t intersect, you’ll have a good set of cross sections.
Where it can get tricky is at a junction.  The following RAS Bloggery article will help with junctions.  http://hecrasmodel.blogspot.com/2009/02/how-to-best-model-junction.html

1. Do the cross sections always have to begin at the same point on a line? I.e 0 is always the same distance from an imaginary centreline down the middle of the river.

If this start point varies how does HEC-RAS know how the channel meanders?

Also where are LOB and ROB downstream reach lengths measured from?

2. Hi Robert. Great questions. No...the cross sections do not have to always begin their stationing the same distance from the imaginary centerline. One cross section could start its stationing at 500, and the very next one can start its stationing at -200. Most people will start their stationing at 0, but you don't have to. Some people like to call the "0" point the channel centerline, and so all station points to the left the centerline will have negative stationing and all points the right will have positive stationing.
Now...how does HEC-RAS know the meandering characteristics of the river? The ONLY effects of meandering that HEC-RAS accounts for in the computations is the varying reach lengths (LOB, Channel, and ROB) for a given cross section. For example, if you have a river bending to the left, the LOB reach length will be the shortest and the ROB reach length will be the longest. Sometimes you have a fairly straight floodplain, while the main channel meanders around within that floodplain. I that case the main channel will have longer reach lenghths than the overbanks. The instructions for measuring the LOB and ROB reach lengths is to follow the perceived "center of mass of flow" of the respective overbank. I'll use contour maps to help guide me, but in the absence of good terrain information, I'll usually measure about 1/3 the distance from the main channel to the waters edge.

1. why am i getting the same flood extent for different discharge values in hec ras/ ras mapper

2. Thank you for all of the resources you provide, Chris!

I was wondering, do you still apply this 1/3 rule if the water's edge is within the main channel portion and not over the overbank station?

For example, on one side of the XS, the overbank station is well defined and there considerable flow within the overbank. On the other side, the main channel bank is steep and high with no grade break until the very top, say like adjacent to a roadway, and no flow is in the overbank portion?

For the steep bank side, is the OB length irrelevant because it never reaches the overbank station? Or is there some kind of doglegging considered while the flow is within the main channel?

I think it would be the latter but wanted someone knowledgeable to weigh in.

3. You're welcome! You are correct in your assumption, if the water never gets to the steep side bank, then what you do for the left overbank flow line is irrelevant. It will never compute flow in that overbank. You could even cut the cross section off at the top of that bank and not even have a left overbank. Just put the bank station on the end point. Good luck!
-Chris

4. Hi Chris,

I am not sure I understand your statement “I'll usually measure about 1/3 the distance from the main channel to the waters edge.” What do you mean by “the waters edge”? A side of main channel, or edge of the target flow in your HEC RAS model, such as 1:500 flood?

3. Thanks for your response Chris this is all very helpful info.

Just double checking is it only the LOB ROB and CHANNEL that affect the calculations not the stationing (i.e if you had one channel centreline at station 100 and the next cross section centreline at 200 but the LOB and ROB are equal then the calculations would not be affected)

Also I am currently modelling a long stretch of river which bends at almost a right angle. In order for cross sections to not overlap the cross sections must cut across the channel parallel and then change direction to go vertical and avoid overlapping? Will this give an accurate representation of the floodplain and flooding? Or can the river be split into different sections.

4. If I had a bridge and a road crossing the main channel at a skew, would I make my sections adjacent to the bridge, parallel to the bridge, or would I still have them be perpendicular to the flow at the channel?

1. I have the same question, a worked example of a skewed bridge and defining the cross sections to the approach would be very helpful. The manual has a small amount of information on this and shows the basic plan but in RAS I cannot get the plan view to look anything like how it is in real life and cannot be sure of the results.

2. Good idea. I'll work on a post for skewed bridges. Thanks for the suggestion

5. Your bridge bounding cross sections (Sections 2 and 3 in RAS terminology) should be parallel to the bridge. AND you should skew these cross sections (as well as the bridge) by going to the Options menu item in the Cross section editor.

6. I'm having a hard time placing bridge bounding cross sections (Sections 2 and 3) parallel to my bridge; because they end up crossing the channel twice. Is there a way to post screen shots?

I read through the skew bridge blog but I'm unclear on how to proceed.

Please note that it is a railroad crossing where I am expecting that my high flow (500-year) will be at least 3 feet below the low chord. So I am not worried about weir flow conditions.

1. If you post a question in the forum, you can insert an image.

2. In the forum, I'm not seeing the "insert image" option. Should I use a particular web-brower? Do I need to be Logged in?

3. When you select "New Message or reply to an existing message, you'll see a text box to write some text. Above that text box is a button that says "Insert Image". Maybe you have to be logged in. Not sure.

7. Hi all,

I found it tricky to draw cross sections perpendicular to the flow line in some cases that I don't know the floodplain extent.

Can we draw the cross section perpendicular to the contours on the left and right banks instead?

K Nguyen

1. That would be a good approximation if you are unable to draw flowlines, but theoretically, it should be perpendicular to the flowlines, not the contours. Flowlines are not always perpendicular to contour lines, particularly in areas contraction and expansion.

8. Hi,
I did imagine that the LOB and ROB data scewed the next (downstream) cross section relatively to the modelled by decreasing and increasing the distance to the next LOB and ROB. So that is not the case? The cross section itself is doglegging?

But how do you model a cross section like the bottommost in the example picture above? On the right-hand-side of the main river, the cross section snaps twice.

Caroline

1. LOB, Main Channel, and ROB reach lengths simply give RAS a length over which to apply the friction loss. Doglegging teh cross sections provide the most 1-D hydraulically appropriate set of station elevation points. Two different things.

2. So how do you get the cross section to dogleg? Is there a certain feature in HEC-RAS where you can do that, or does that have to be defined from Autocad before exporting them?

If it's done through Autocad, how do you export those cross sections from CAD to HEC-RAS? What I've done before is expot an alignment line with cross sections 20' apart for example, and they come into HEC-RAS as straight lines across the stream and floodplain. Am I supposed to draw the cross sections (perpendicular and doglegged) then define them as their own profiles and export them?

3. Doglegging takes place in your GIS application. We use ArcMap, but I believe AutoCAD can do the same. Getting the cross sections into HEC-RAS from AutoCAD requires some 3rd party application, which I'm not familiar with. In ArcMap, we use HEC-GeoRAS to do this.

9. When creating river banks and flow paths with GeoRas, how do you handle a bridge or abutments that come into the reach? Do you draw the flow paths over the bridge as if it was not there and same for the river banks? Or do you follow the contouring around until the opening?

1. If you expect the bridge to be overtopped in your simulation, draw your flow paths and bank lines up and over the bridge deck, just as you would expect the water to flow. If everything is contained within the bridge opening, then draw your lines through the bridge opening.

10. I've been modeling (HEC-2 and HEC-RAS) too long....Need to retire. Recently, I was sitting in a live conference of HEC-RAS / HEC-GeoRAS was being presented. The presenter had a skewed bridge, so instead of utilizing the Skew Routine in HEC-RAS, the presenter, when cutting the X-Sect at the US and DS sections of the bridge, cut the X-Sections perpendicular to the flow In and Out of the bridge, and allowed his cut-line to cross the road. This was surprising to me. I've always paralled the road @ toe of slope US
and DS. Would like some discussing about this method. "When cutting x-sections, should you cross the road that the bridge is located or keep the x-section parallel to the bridge?"

1. That technique is not recommended. You are correct that keeping your bounding cross sections parallel to the bridge deck is a better way to represent the hydraulics. However, out in the overbank, I've crossed over the roadway approach to keep the cross section perpendicular to my perception of the flow direction (this is common when the roadway approach bends significantly relative to the alignment of the bridge deck. As I always say, when in doubt, keep your cross sections perpendicular to the flow lines.

11. Hi,
Is there anyway to calcualte the area of each cross section in HEC-RAS?

1. There are a few different ways. If you are interested in the flow area (i.e. the area of the cross section that is wet, you can get that from the profile output table. If you want the area of the entire cross section (end point to end point), you can get that from the hydraulic property plots/tables (View...Hydraulic Property Plots), and select Total Area as the variable. Also, and this is less precise, but you can measure the area with the measuring tool in the cross section plot. Just hold down the Ctrl key and click points around the perimeter of the cross section. Then when you've completed a polygon the shape of the cross section, release the Ctrl key and you'll see an Area value pop up.

12. Hi,

I am trying to remeander a reach of a brook, and I am trying to get HEC RAS not to show me a straight line in the geometry data editor, something more similar to the image at the top of the page. I have been trying to import it from AutoCAD and ArcGIS, unsuccessfully though. I have also tried to import the GPS coordinates of a few keypoints along the reach (from a CSV) but it is not working either.

Is there a way I could get the reach looking not like a straight line but how it is planned to look?

Thank you!

1. Yes, you have to set it up in GIS and then import. The GeoRAS manual steps you through the process very well and should help you do this.

13. Thanks for the post, Chris.
I am undertaking a assessment for a creek for existing and proposed cases. In the post case, two sections of the creek will be re-aligned, due to encroachment of a new road. The re-alignment of the creek is pretty much shift the creek about 2 meters and straighten the creek. My question is, do we normally keep the same XS lines for both the existing and proposed cases? if we do, that means some XS lines won't be perpendicular to the flow direction in one of the case? but we don't keep the same XS lines, it will be hard to compare the water levels between the existing and proposed case? (which is the main purpose my modelling).

1. In these situations, I realign the cross sections so that they are perpendicular, but I try to keep the intersection of the cross sections with the stream centerline at the same river stationing. That way, they will be consistent on the profile plot and you can do a direct comparison. But you are correct, it will not be entirely consistent, since the cross section orientations have shifted from existing to proposed. You just have to keep that in mind and communicate that clearly to your client and other stakeholders.

14. Hi Chris,
We are developing the cross cut line for a large river basin in Hec-Geo RAS. Due to meandering nature of river line the automated process of XS line construction, generate irregular cross cuts, which need to be correct manually for Hec RAS run.
Do we have any other automatic way to generate cross cut for such large river basins?

Thanks,
Prince

1. I don't. I've always drawn in my cross section cutlines by hand.

I wonder, if we could find a approach for it. Otherwise it would be time consuming exercise for large basins.

Many Thanks,
Prince

15. Hi,
I am using HEC_GeoRAS in ArcGIS for extracting the geometry data. Now, the river I am modeling does not have much of a floodplain so I have the dark blue as the river/flow path centerline and dark brown bank lines. But when I extract the data to Hec-RAS, LOB and ROB reach lengths are missing. I have extended my cross-sections beyond the bank lines. Can this be a reason for it? I have already went ahead with the modeling. Is there a way this can be corrected now?

1. Make sure you have drawn flow lines for the overbanks for each river/reach and that you have assigned a subsection to them (LOB, ROB).

16. Hi,
How do you change the view angle of the whole schematic plot ?..for presentation i need north of the plan to be towards top of my sceen

1. You can't in HEC-RAS. But you can export an image and rotate it in your favorite image handling software.

17. whether negative sign can be used in HEC-RAS geometric data at meandering portion of river when cross section criss-cross each other.

18. how to enter centerline of river in HEC-RAS from AutoCADD file of GIS file.

19. negative numbers can be used in cross section stationing. However, surveyed cross sections, or cross sections cut in GIS should never criss-cross or overlap. Sometimes, just due to the schematic representation in HEC-RAS, your cross sections may criss-cross in the geometric data editor, but they shouldn't criss-cross in a georeferenced project.

20. Hi Chris,
I have had a few problems with generating cross sectional data from Inroads into HEC-RAS v4.1.0. The first problem is that I am trying to use our survey alignment for stations within HEC-RAS but since we have station equation adjustments the cross section stationing within HEC-RAS does not match like it should. It appears that the .geo file that is generated takes cross sections at intervals based on the linear geometry rather than our station values (our survey data is adjusted about 90-95 feet depending on its location). The other issue I am having is that the export does not seem to recognize arc lengths for the stream alignment and makes a line connecting the 2 points that formed the arc. Do you happen to know of any fixes for these problems?

Sincerely,
Randy

1. Randy, I don't have experience with Inroads, so I'm afraid I can't help with that. As far as arc lengths, best way around that problem is to use more vertex points to define your stream centerline.

21. Hi Chris,

I don't know if my last post went through so sorry if this seems repetitive. I am currently having issues with generating geometry from inroads into hec-ras and am looking for some assistance. The first problem that I am trying to rectify is the stationing that appears within HEC-RAS. I am trying to have our survey stations be the same in HEC-RAS but our stationing has a few station equation adjustments along it (our stationing is adjusted 90-95 feet depending on its location). The generated .geo file appears to take cross sections at the given intervals based on the length of the alignment rather than at our station markers yet the "Generate Water Surface Data" recognizes our station naming when asking for information regarding the alignment. The other issue is that the generated geometry does not recognize arcs within the alignment and substitutes them with a line connecting the start and end of the arc. If you know of any fixes for these that would of great help.

Sincerely,
Randy

22. Hi Chris. Is is typical to have cross-sections on one reach intersect with cross-sections on another reach? My gut answer is "it depends on the flow regime and how close these reaches are to each other."

1. No. You should avoid this. Overlapping of cross sections (unless purely schematic) causes overestimation of volume.

23. Hi Chris. Is it typical to have cross-sections on one reach intersect with cross-sections on another reach? My gut answer is that it depends on the flow regime and how close these reaches are to each other.

24. Hello Chris when i create the tributary in HEC-RAS it give message that these section are not geo referenced, so when export to arc-gis ,the downstream reaches of the junction do not export to arc???

1. Anything not georeferenced will not properly export to ArcGIS. You must make sure everything is georeferenced to be able to map.

25. Hello Chris when i create the tributary in HEC-RAS it give message that these section are not geo referenced, so when export to arc-gis ,the downstream reaches of the junction do not export to arc???

26. How does one draw a long section in Hec Ras

1. Same as a short one. Just longer.

2. What I mean is I have a pipeline about 40 km the pipe crosses about +- 20 dry streams, which I used Hec Ras to run my floodplains now I need to show the flood level for each river on my vertical alignment (civil 3D) where the pipe crosses the river.

3. Better to model each stream individually then, rather then with one long cross section.

27. Ok thanks Chris, how to do I find the flood level @ intersection where my pipe crosses the stream, because all cross sections have different flood levels.

1. If you have a cross section located where the pipe crosses the stream, you can get a flood level from that cross section. Maybe I don't understand what you're trying to do.

28. Hei Chris

I am working with HECRAS 4.1. I want to ask you one question about calculating downstream reach length, LOB, Main Channel, ROB. The downstream main channel length is well understood. but what does downstream reach lengh for example LOB mean ? Is it distance measured between two consecutive left bank stations or between water edges ? And how are they measured in Autocad where c/s are produced ?

1. The overbank (LOB and ROB) reach lengths should be measured along a line in the respective overbank that follows your interpretation of the center of mass. So it should be somewhere between the bank station and the edge of water. It is a very subjective thing to do, but fortunately, it is not very sensitive to the results, normally.

29. Chris,

I have a river that runs parallel to a road (pretty much), then crosses the road (almost perpendicular), and continues on the downstream side parallel to the road.
I am struggling with how to draw my bounding cross sections. On the upstream side I have been drawing the x-sects perpendicular to flow which resulted in the x-sects crossing the road. If I draw my bounding x-sects (sects 2 and 3) parallel to the road I end up doglegging section 3 across the road such that I don't cross an upstream cross section, and also to contain the flow within the cross section (I suspect that the WL will exceed the top of road).
Question 1. Is it acceptable for the x-section to cross the road?

If I do this then I also end up doglegging and crossing roads on the downstream side 3 more times (where I have 3 crossings pretty close together).
Question 2. Would it be better to draw the bounding x-sections perpendicular to flow (and hence crossing the road at both bounding sections 2 and 3) at the upstream culvert, thereby eliminating the need to cross the road at the next 3 downstream culverts?

1. Yeah, sometimes you have to cross over roads. That's okay as long as you are keeping the cross sections perpendicular to flow lines.

2. Thanks Chris.
Do you have any suggestions on how to enter the road deck elevation if my bounding cross sections are not parallel to the road?

3. It's tricky. Try to keep the road deck on top of the bridge at least while in the main channel. In the overbanks, if the approach road turns significantly, you can cross over it. The bounding cross sections should likewise be parallel to the bridge. Make sure to skew the bounding cross sections. All other cross sections should be kept perpendicular to flow lines. Dog leg them as necessary.

30. Chris,

I am assisting in a widening/restoration project that is looking at incorporating an existing water quality basin that runs parallel to the creek and I was curious as to the best method for modeling the additional area since there will be ineffective flows that should only fill up when the remaining separation is over topped and then doesn't recede back into the creek. It is almost like a storage area but there will be an area above the ineffective area that will flow at certain wsel. I have included a drop box link with the existing condition cross section and the proposed version. I also include some comments regarding my proposed methods. Any help would be much appreciated. https://www.dropbox.com/s/w3fdyvgtk5446xh/Cross%20Section%20Info.pdf?dl=0

Sincerely,
Randy

1. You might want to consider breaking out the area to the side and model it with a storage area. You can connect the cross section(s) to the storage area with a lateral structure.

31. Good afternoon.
I have a question about the Flow Path Centerlines. What is the most correct way to draw these lines?

1. The center flow line should generally follow the stream channel invert. The overbanks should follow your approximation of the center of mass of the respective overbank. A common method is to go 1/3 from the bank station to the water's edge.

2. Hi Chris,

I'm hoping you can clarify that pesky "generally" for me. I have a reach which is well-confined, containing a well-defined floodplain where the channel meanders considerably. As the focus of the project is the 100-yr event, a large proportion of the flow is expected to just shoot over the meanders. I have the overbank flowpaths drawn to reflect this. It has been suggested that I also use a "flood flow" main channel length which would approximate the downstream distances. This makes some sense, but I don't know if it is fully justified. My understanding of HEC-RAS's computation methods makes such an adjustment seemingly unnecessary, but I don't know if doing so is actually erroneous. Looking for literature guidance, I have not found anything to suggest deviation from the centerline - aside from pesky mentions of "generally".

Regards, Tyler

3. typically what is done is if you are looking at an event simulation (low flow to high flow back to low, you would maintain the main channel reach length following the low channel and then set your overbank reach lengths to go with the high flow. but if you are strictly focusing on the 100 year flood, then I would set the main channel lengths equal to the flood flow lengths. I think this is the most appropriate way.

4. It is Steady Flow model so it looks like the flood flow lengths are most appropriate. Thanks for appeasing my conscious with your advice!

Regards, Tyler

32. Hello Chris,

I have looked around, and can not find a reasonable explanation for this. Why can't two cross sections intersect?

1. Quite simply, if you overlap or intersect cross sections, then you will be double counting volume and your results will be in error.

33. Dear sir,
I am trying to make a project using Arc 9.3 Hec-GEOras 4.3.93 and Hec-Ras 5.0.1 and is about dam failure.
I have a lot of questions and I would appreciate if you could help me. I will use photos to my questions so it will be easier for you to help me. How can i upload here one?

1. The forum would be a much better place to submit photos and to solicit help for your model. https://hecrasmodel.blogspot.com/p/hec-ras-bloggery-forum.html

34. Hi Chris,
Thanks for this amazing blog. I am new to this world and I have learned a lot while reading your comments/replies. But I cant find solution to my own problem.
I have developed a hec ras model for the steady state but I cant get it to run. It stops computing when it is developing "cross section interpolation layer". The model has 520 cross sections along four tributaries of total 42 km in length. The model works fine for 2-3 km reach in each tributary but the model stops computation, if I try to run it for the entire tributary. I cant ungerstand why only a few km reaches are runing when I have used the same approach for obtaining cross sections throughout tributaries. Any idea what could be going on?

1. Not really sure what's happening there. You might try to use a slightly different max spacing and reinterpolate.

35. Hello Chris,

Thanks a lot for all your pro-bono HEC-RAS work. I am fairly new to this software and have some fundamental questions.

In response to a query above regarding overbank flowpaths, you responded stating that they should approximately follow the center of the mass of flow and that generally is 1/3rd the distance from the channel centerline to water's edge.

1)If you are modeling a previously unmodeled stream, what's the best way to judge where the water's edge would be?
2) I have read varying opinions on where the bank stations should be set but I haven't been able to come to a conclusion. I found two very short forums a) http://www.eng-tips.com/viewthread.cfm?qid=343019 and b)http://hec-ras-help.1091112.n5.nabble.com/bank-stations-td190.html that briefly discuss this topic but it would be very helpful if an expert like you can parse the information.

Could you please browse through those two links and provide a short summary of how to locate bank stations?It would be great if you are able to write a separate detailed article on this topic. Thanks,

VB

1. Hi VB-

1. That's where the 1/3 comes from. It's actually 1/3 the distance from the bank station to the edge of water. You'll have to take your best guess as to where the edge of water will be. Don't worry though, the results are typically not overly senstive to the overbank reach lengths. All the action is happening in the main channel.
2. The most accurate way to locate the bank stations if you want to stay true to the conveyance distribution method used in RAS is to locate them where you have significant change in conveyance between the main channel and the overbanks. No one knows where this is exactly without running the model first, so typically modelers will put them at the obvious change in grade (or top of bank). A change in roughness is also a good place to key in on. And typically, your main channel width shouldn't change drastically from one section to the next. Reviewing the comments on the threads you link to, they're going into a lot of detail and discussion, most of which seems logical and there's certainly a lot of "industry standard" practice being quoted, but if you want to know for sure you have it correct, put it at the change in conveyance. If you want to double check where this is, do a velocity distribution with a lot of small slices and run the model. For most cross sections of typical rivers/floodplains, you'll see clearly where this is. But like with most things in computer modeling, there are always exceptions to the rule, so don't take it as gospel. I like to tell people to understand how the computations are being made, know your modeling objectives (i.e. what questions are you trying to answer), and use common sense, and you'll get it right every time.

2. Chris- Thank you for your detailed responses. It helps to know these things so that I don't start doubting my model-setup if it doesn't conform to one school of though or the other.

36. Dear Chris,
I am working on flood delineation using HECGeoRAS with ArcMap 10.2. while doing flood inundation in mapping unit I faced error stating "system.Runtime.IteropService.COMException(8x80041098):ERROR010316 Unable to open the input raster:....Error in Executing grid expression at ESRI.ArcGIS.SpatialAnalysi.RasterConditionalOpClass.SetNull(]GeoDataset ConditionalRaster.IGeoDataset False Raster) at HECGeoRAS.clsMFloodGridIntersection.CalcDiffGridMonolith(IGeoDatasetPWSDS,String sUniqueDepthGridName) how can I overcome this problem? Thanks

37. Hi Chris ,while studying backwarer study on Hec Ras we came across problem like length of few cross section is about 15km the reason for this is that dam located in such a area where one side(say left) is elevated while other side(say right) is amost flat,getting ground level above 3to 5 mtr.of dam/full reservoir level is quite difficult also lots of tributaries associated with main river. How perform backwater study in such a scenario by using Hec Ras.

1. If you're running a steady flow model, and most of the 15km portion of the cross section is ineffective flow, you could ignore it (omit the ineffective portion from your cross section). Tributaries can be included as their own reaches, connected to the main stem with junctions, or as lateral inflows.

38. hi chris,

i've been wandering around as i encounter a problem while trying to Export RAS Data in Hec-GeoRAS but not much help. the problem says c:\Program Files(x86)\HEC\HEC-GeoRAS\bin\NewRasControl.xml and in the dialog message box "target location not specified..."fyi i've re-drawn my xs for times and it fix other problems but not this one. could u plis advice where i probably went wrong? Also is there a specified distance between two xs and i extended my xs very far away from flowpath....is it okay? i really appreciate if u could help..fyi im using arcmap 10.4 and hec-georas 10.2..thank u so much

39. Dear Sir,We are Working With Project Where We Have to estmate backwater level by using HECRAS v4.1.0 but we are facing a problem as main rivers accompanied by nearly 17 tributaries (10 tributaris on right side,7 tributaries on left side of river respectively) we have to study upto 24 km.it seems that cross section of river goes into tributaries &tributaries cross section goes into main river in objective to get GL 3 to 5 mtr.above The FRL.Also is it possible that manning value 'n' is same upto 24km? Please help us in this regard -bvs

40. i have done finishing cross sections and exported to HEC-RAS. but the cross section profile in hec-ras i.e the left and right bank are not unique. How to edit that cross section geo,etry in HEC-RAS.

41. Hi, Chris! I am working on a dam break project and I've recently received the survey data. Problem is that, once I've georeferenced, the elevation points do not follow a straight line. Some points are displaced. I've uploaded a print screen in the imgurl to illustrate.

http://imgur.com/JFBjRYc

Can you tell me if there might be some kind of problem by inserting this kind of cross section? Because I've run the model and have checked the results and apparently there is nothing wrong, however, I'm not 100% sure that it's ok.

1. Matheus- That is strange. I've never seen that before. Must just be a graphical thing since the results look okay. You might consider manually moving the points into position if you want it to look more true to reality.

42. Hi Chris,

I just wanna ask. What if I have an existing project and I need to work on it again but with bigger extent of study (i.e. river basin considered increased/enlarged thus extent of hydraulic study for the main channel and tributaties will be farther upstream), can I still use the existing centerline, banklines, & flow paths shapefiles from the previous project and just extend it further upstream (up to the portion of channel included in the new&bigger basin) when I digitize the model with HEC-GeoRAS in ArcGIS?

Do you (somehow) know if this process will cause any problem that I might encounter?

Thanks for the insight! I appreciate your help.

-Dylan

1. Dylan, yes you can do that. Just make sure that where you have increased your model extents, the river stationing, reach lengths, etc are all consistent with what you already have in your model. Should work fine though. Give it a try!

2. Thanks for the response. I'm not sure I quite understand what you mean. Because I thought those parameters will change since the area of study (channel, flow paths,bank lines) will extend farther upstream and I'll have to repopulate the attribute table for this parameters before exporting them to hecras. Am I right? Also will it work for xs lines as well?

To be honest the reason I'm not entirely sure about this process is because I'm still waiting for a more refined DEM so I havent tried doing it yet. I'm mostly relying to instinct and answers from those who have encountered and done this before.

Thanks

- Dylan

3. Yes, you are correct. Once you've defined your new model extents, just go through the GeoRAS steps from start to finish. That will redo all of the attribute tables and prep your data for exporting to RAS.

43. Hi Chris!

Have you worked on a project where in the geometric cross section data are all surveyed data points formatted in CSV files (nor things, eastings, elevation) and directly export the to HECRAS? If so, do you know how to work around this kind of problem and would you be kind enough yo give some kind of step-by-step guideline?

I learned the it was briefly mentioned in the reference manual but I can't seem to grasp the process and perform it correctly when I tried following it because I keep getting warning message saying 'cross sections not georeferenced'.

I hope you can enlighten me on this dilemma of mine. Any help will be much appreciated.

Thanks a lot!

- Alexis Morales

1. Alexis. Yes, that's a common thing. It just takes some geometric calculations to convert northings/eastings to stationing. The cumulative distance between points on the cross section is the stationing. The distance between two points can also be considered the hypotenuse of a right triangle where the difference between the 2nd point's northing and the first point's northing is one side of the triangle, and the difference between eastings is the other (non-hypotenuse) side of the triangle. So, the distance between the points (on the hypotenuse) follows Pythagorean's law. Distance = ((y2-y1)^2) + (x2-x1)^2)^1/2

A simple spreadsheet will allow you to easily compute stationing from northings/eastings. But keep the northings and eastings. While not required, you can use them to georeference your cross sections in the GIS Tools menu item. It makes for a more realistic looking model.

2. Hi Chris!

Thanks for the reply! I think you understood my question differently seeing that I didn't mention that my surveyed data has complete details, meaning there is stationing as well as the northings, eastings, and elevation. Lucky me! :)

What I'm actually having problems with is because when I tried importing the data points, all the cross sections are seem to follow the correct positioning/location of each cross section only the main channel and tributaries doesn't have nodes/junctions. Also, warning/error message says cross sections are not georeferenced.

I totally understand that this is the case because I haven't established the centerlines yet. However, when I tried digitizing the centerlines in ArcGIS from a DEM and exporting it to HEC-RAS, it still doesn't seem to fix it. The warning/error message is still there and the stationing as well as the reach lengths don't adjust themselves. And I'm quite confused because this is the process that the reference manual states when creating the model geometry from csv file format.

I hope I made my problem a bit clearer this time. Thanks for your help!

- Alexis

3. Alexis-Not sure why that is happening. All I can suggest is that you follow the directions closely in the GeoRAS manual for creating your geometric elements in GIS and importing them to RAS.

4. Chris,

Thanks for the response and tips! Follow up question, I just thought about after you mentioned referring to the manual because it wasn't see there. If I have different files of the surveyed data of the main channel and several tributaries, should I merged them into one CSV file before importing to hecras or should I make individual CSV file for the main channel and each tributary?

Thanks a lot and more power to you and this BL!

-Alexis

5. To be honest, not too many people use the import by CSV file method. That would explain why it is so buggy. I haven't done it myself, I like to import using GIS format. If I were to try importing by CSV, I'd probably do one river/reach at a time.

6. Oh I see! The main reason why I'm doing it is because the cross section data from bathymetric survey is quite a lot since the main channel is long and there quite a number of tributaries. Aside from that, the channels are not adequately represented in the DEM of the catchment because of their depth. I was kinda hoping that by directly importing the cross section from CSV files will somehow speed up the process. Nonetheless, thanks for the insights! I appreciate them a lot! I'll try doing it thru GIS if I still can't make it work.

44. I have a bridge where the stream turns immediately downstream of the structure and runs parallel to the roadway. Is it ok for my downstream cross-section, cross-section 2, to cross the deck/roadway profile entered in the bridge/culvert data so that the cross section remains perpendicular to the flow lines?

1. You've highlighted one of the more difficult scenarios for laying out cross sections. Yes, you can bend your cross sections so that they remain perpendicular. However, you want to be careful about crossing over the approach roadway. Make sure you have a good feel for whether or not water will overtop the roadway during the simulation. If so, you'll want to avoid having your cross sections intersect your road profile. Unless you think water will be moving away from the bridge, in the direction of the road, then you can do that.

45. Hi,
i am extracting manning n values for crosssections in arcmap and it gives the message that N values are extracted successfully but the manning table is empty. i can't figure out the problem.

46. Hi Chris,

What is the proper way to lay cross sections over dams?

1. Best not to put them on or over dams, but instead lay them parallel to the dam crest at the upstream and downstream toes. Then insert an inline structure between those two cross sections to represent the dam.

47. Respected sir,
I have digitised the river reach in HECgeoRAS with stream centre line, bank lines, flow path centre lines and XS cut lines as the RAS layer. The geometry is exported to HEC RAS sucessfully but when i look into hec ras at each and every CX the value of ROB and LOB is missing!!! however the channel length is specified..
pls help me out

1. Hi Tejas-

All I can say is double and triple check that you haven't missed a step in the process as outlined in the GeoRAS manual. You can always check the attributes tables and make sure all of the data is there. A common mistake people make that will prevent LOB and ROB reach lengths from coming over is forgetting to designate them LOB and ROB in GIS. Also, if they don't intersect cross sections, lengths cant be computed. Keep trying...you'll figure it out.

48. Hi Chris,
What will be effect on channel water level depth when we increase or shorter the length of cross cut line?
Thanks

1. Prince-

The cut line itself has no bearing on the computations. However, the station-elevation data does. When you change your cross section cut line length in GIS and then re-extract it to RAS, the station-elevation data will change. If the flow area is impacted by this change, then the results will change. If you manually change the cut line coordinates inside the GIS menu in the RAS geometry window, nothing will change with the computations, since the station-elevation data won't change. However, there is an option to extend the cut lines and sta/el in the GIS Tools menu. That will change the sta/el data so the results may change.

But perhaps your question is really "what will happen to the channel water level depth if we increase or decrease the channel width". Typically if you increase the width, the water level will go up locally, but will show a decrease in the backwater upstream of the change. If you decrease the width, the water level will go down locally, but will show a rise in the backwater upstream of the change.

49. dear sir,
we are trying to dam breach analysis. we have face to problem that inline structure is always displaced from its original place and it is located near second or third cross section.

50. I entered the cross sections and bridges. However, the water is in no contact with the bridge, it is higher and out of alignment with the bridges

51. I entered the CS and bridges However when running a steady flow. The water level is in no contact with the bridge and the water is above the bridge and out of alignment with the bridge

52. Hi Chris,
is there a method (in RasMapper/Hec-RAS 5.0.4) to draw XS perpendicular to river center lines?

Andrea Conca

53. Hi Chris,

I have a question about x-secs crossing the roadway that approaches the bridge. I’ve read earlier questions and comments, some of the replies could be more specific or clarified. Comments I’m looking at were posted 02/20/15 and 11/14/17.

I have a stream in a valley with a bridge crossing perpendicular to flow. Before and after the bridge, the approaching road is parallel to the stream/flow. Think of a harsh “S” shape. Originally I drew x-secs with middle portions perpendicular to flow and parallel to bridge, and outside portions parallel to the road and sometime parallel to stream. At the middle part of x-secs flood events will overtop the road, I have appropriate ineffective flow areas defined here. On the outside parts of the x-secs, I believe flood events don’t cross the road. In this case what do you recommend the cross sections do? Cross the road? Keep original shape but add ineffectively flow areas (temporary or permanent)?

Additionally, the DOT has suggested x-secs that only remain perpendicular to flow and crosses the roadway several times. Where the suggested x-secs intersect the road it’s overtopped during certain flood events.

1. When in doubt, always draw cross sections perpendicular to flow lines. The direction of flow lines are determined by you, so it takes close examination of the topography and things that "get in the way of the flow" like bridge roadway approaches. You've identified the tricky part of bridges-when to transition from the parallel bridge bounding sections to those same sections being perpendicular to the flow lines. typically I make this transition around the bank station, so the main channel is parallel to the bridge, the overbanks are perpendicular to the flow lines. Of course, every bridge is different so use your judgment. Understanding the basis (and limitations) of 1D modeling, and the computations behind it, will help guide you.

54. Hello,

First of all I would like to thank you for this blog, it helped me demystify the meaning behind a lot of the HEC-RAS capabilities and guided me in correcting and fine-tuning my first simulations. But I am still new to HEC-RAS and I have a problem to solve, one that I do not know how to approach. I am trying to simulate a bridge to be built on a bend of a stream. The problem is that the bridge is also on a road bend, exactly at the intersection with the stream. So there are two intersecting arcs, I am puzzled how to draw my cross sections in order to define the bridge piers. Should my cross sections be arcs parallel to the road – bridge arc? Or is this not acceptable? If my cross sections are straight lines perpendicular to the stream centerline then I cannot define exactly the piers beginnings and ends. The bridge is to have 3 piers and the cross section is going to be a two stage river channel. I will try to provide pictures as an example.

Also, I have run the simulation in HEC-RAS with arced cross sections and I get the following messages and notes for various cross sections but specifically for the one at the bridge:

“The velocity head has changed by more than 0.5 ft (0.15 m). This may indicate the need for additional cross sections.

The conveyance ratio (upstream conveyance divided by downstream conveyance) is less than 0.7or greater than 1.4. This may indicate the need for additional cross sections.

The energy loss was greater than 1.0 ft (0.3 m). between the current and previous cross section. This may indicate the need for additional cross sections.

Multiple critical depths were found at this location. The critical depth with the lowest, valid, energy was used.”

I can add extra cross sections at the other parts of the stream but not in the bridge area, is there any way to overcome that and have HEC-RAS solve it without warnings?

Any help would be greatly appreciated.
John

1. Hi John. In the river, definitely keep your bounding cross sections parallel to the bridge. Out in the floodplain, where you go from parallel to the roadway approach to perpendicular to flowlines will depend a lot on the terrain, the degree of skew of the flowlines to the roadway. Generally, if the skew is quite low (<20 degree approx), I'll keep the cross section parallel to the roadway approach. More than that, I'll generally bend the cross section back to being perpendicular to the downstream flow direction. This is definitely a very subjective exercise. There is no "right" answer. But just keep a couple rules in mind and try to stick to them as much as possible.
1. Draw cross sections perpendicular to flow lines. This may require you to dogleg your cross sections if the flow lines are not parallel with each other.
2. Keep your bridge bounding cross sections parallel to the bridge deck. Skew the bridge deck and bounding cross sections if they are not perpendicular to the flow lines.