As per the title, I am looking for a tool that can produce bathymetry of a river from cross sections. Although people who may know such tools probably will not need any explanation why this kind of interpolation is a bit complicated, here are a few words of explanation for those which may know and use such things for other purposes.

In order to receive a DEM of river valley with river bed bathymetry, I need to fence/mark with barriers the area on my original DEM where bathymetry is to be created. Basically it is the whole area of the wet river bed where TLS did not have a chance to measure the bottom surface (bathymetry). Rivers are in various shapes, sometimes more meandering, sometimes more straight, and also the depth of the bottom and its alignment are varying a lot, which makes interpolation more complex as it has to follow the centre line of the river and also conduct the interpolation vertically and horizontally.

Dr Venkatesh Marwade has prepared such extension to ArcGIS 9.2, though I did experience some problems and was just wondering if anyone would know a piece of software or code that could generate something alike.

(Edit: see graphics in this tutorial.)

  • Can you post a graphic for this? Upload to some place like imgur.com Commented Feb 1, 2011 at 14:55
  • 1
    I added a link to a pdf that has graphics in it. Commented Feb 1, 2011 at 15:17
  • Thanks. I'll remember to attach such information in the future.
    – Tomek
    Commented Feb 2, 2011 at 6:57
  • I also want to use Dr Venkatesh Marwade's tool to create river bathymertry from cross sections. But, I don't know how to generate my own input file (as the same format as the SampleData.mdb). Can you teach me how to do that? Thanks.
    – user3510
    Commented Jun 29, 2011 at 18:10
  • @peterlai: If the answers here are not enough, please open a new question.
    – underdark
    Commented Jun 29, 2011 at 18:16

7 Answers 7


Have you had a look at GRASS Wiki "Marine Science" yet? It discusses Bathymetry processing and the necessary steps.

  • Yes, I did look in to this. However you gave me second thoughts and I'll have a look at it again. In fact there was some interesting interpolation tools. i'll keep you informed.
    – Tomek
    Commented Jan 27, 2011 at 19:04

At this point I have to correct myself.

The solution I was looking for, is a tool incorporating a set of procedures, 2 linear interpolation methods, etc. rather than a single interpolation method/tool.

I did not find any tool able to do the task properly, other then Dr Venkatesh Merwade's tool.

I had to downgrade my ArcGIS for Desktop version in order to use it. The tool is still not bug-free, but it does the job.


I found an open source (python, R, QGIS, and PostGIS) workflow by Michele Tobias, based on Dr Venkatesh Merwade's 2005 paper here: http://www.slideshare.net/MicheleTobias/open-source-workflow-for-surface-interpolation-with-curvilinear-anisotropy

There is a Github project with documentation trying to turn this into a (very useful) Qgis plugin: https://github.com/MicheleTobias/CurvilinearAnisotropy


This is normal functionality of Civil 3D. Autodesk products.

To work with esri programs there are some free and paid for (listed)

User Conference presentations1
User conference abstracts
User conference abstract2
USACE documentation

MWH software

Bentley WaterGEMS

  • I am working on those. Will let you know if found what was looking for. Thanks.
    – Tomek
    Commented Jan 28, 2011 at 9:09

Mike11 has a module that does this.

  • As far as I am concerned, MIKE does simple linear interpolation, where results are shown within the river boundaries, but interpolation itself is carried in straight line cutting terrain etc. I may be wrong, but I think thats the way it is. However I may be not aware of some additional modul. I'll check if thats the case.
    – Tomek
    Commented Jan 28, 2011 at 8:55
  • What I am looking for is actually something similar to dr Merwade's tool, which basically interpolates new sections in between sections which are known and than is connecting all the sections with interpolated polylines. New sections are aligned to known centerline. In result you receive mesh of polylines. Its pretty basic describtion, but I just want to indicate that in this way you receive mesh of vectors instead of raster. Thanks
    – Tomek
    Commented Jan 28, 2011 at 9:07
  • I had forgotten mike.
    – Brad Nesom
    Commented Jan 28, 2011 at 12:20

It has been a few years since I looked at this stuff, but HEC-RAS from USACE is a powerful, if complex, suite of software -- widely used in government. There may well be hooks to ArcGIS by now, too.


The correct way to do what you need is to use a tool like v.surf.rst from GRASS. The problem of your data is that the data is not isotropic, like in general the interpolation of contours is considered. So there is need to insert a parameter for anisotropy There is an example in Neteler and Mitasova Open Source GIS: A GRASS GIS approach, at page 163 (2004 book which can be seen here).

Parameters for anisotropy are present in R package gstat, in which you can perform various types of kriging interpolation.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.