Importing partially georeferenced ISIS model to ICM

[Martin Spiers]

I have an ISIS model which contains XYZ data at cross sections. However when I import this into ICM, the cross sections are not geo-referenced at all.

There are some structures within the ISIS model which are not geo-referenced, but my understanding from the help file is that coordinates should be imported where present.

Any idea why this isn't working?

[Kristian Ravnkilde]

This is a common problem, and one I expect more and more of us will be having. You need a .gxy file to go with it to complete the georeferencing, which in a nutshell gives you the location of each cross section. One may be available from the original model builder. If not, and if there's a TUFLOW model to go with it, you can extract the locations from the relevant SHP file. You need an ISIS modeller to help you, there is nothing in ICM that does the job. Best of luck.

[Martin Spiers]

This is a common problem, and one I expect more and more of us will be having. You need a .gxy file to go with it to complete the georeferencing, which in a nutshell gives you the location of each cross section. One may be available from the original model builder. If not, and if there's a TUFLOW model to go with it, you can extract the locations from the relevant SHP file. You need an ISIS modeller to help you, there is nothing in ICM that does the job. Best of luck.

I don't have a GXY file, but I do have a GIS file of the node locations and of the cross section lines. Does anyone know how to create a gxy file from that? (I have ISIS available but am not that familiar with it)

[Will_Rust]

If you're importing cross sections into ICM, EACH VERTEX of your cross section (each chainage elevation) needs an XY. ISIS (unless its fully georeferenced) only has XY for the centroid of the cross section - which sounds like what you have. Its totally possible to get in into ICM with the data you have. You'll need to export the ISIS cross sections to CSV, and extrapolate your XYs from the center point to each chainage elevation (you can assume that the XSs are horizontal to do this easily, or work out their rotation based on the previous / next XS XY). the XY for cross section centroids may not be in the same file as your XS chainage elevation, so you will need to do some matching and indexing in excel.
import into ICM needs to be:
XS_NAME,X,Y,Z,Mannings
So essentially, listed XYZs, grouped by the Cross Section name! (hope that makes sense!)

Will

[Martin Spiers]

If you're importing cross sections into ICM, EACH VERTEX of your cross section (each chainage elevation) needs an XY. ISIS (unless its fully georeferenced) only has XY for the centroid of the cross section - which sounds like what you have. Its totally possible to get in into ICM with the data you have. You'll need to export the ISIS cross sections to CSV, and extrapolate your XYs from the center point to each chainage elevation (you can assume that the XSs are horizontal to do this easily, or work out their rotation based on the previous / next XS XY). the XY for cross section centroids may not be in the same file as your XS chainage elevation, so you will need to do some matching and indexing in excel.
import into ICM needs to be:
XS_NAME,X,Y,Z,Mannings
So essentially, listed XYZs, grouped by the Cross Section name! (hope that makes sense!)

Will

Not all of my cross sections are georeferenced, but those that are are fully georeferenced like this:
RIVER
SECTION
01_10337u
0
20
0.000 109.830 0.100 1.000 999335.425999371.749 0
5.000 109.870 0.100 1.000 999337.428999367.168 0
9.100 110.070 0.100 1.000 999339.071999363.412 0
9.770 110.070 0.100 1.000 999339.339999362.798 0
9.960 110.080 0.060* 1.000LEFT 999339.416999362.624 0
9.960 109.480 0.060 1.000 999339.416999362.624 0
9.962 109.280 0.060 1.000 999339.416999362.622 0
10.480 109.240 0.060 1.000 999339.624999362.147 0
11.760 109.230 0.060 1.000 999340.137999360.974 0
12.540 109.230 0.060 1.000 999340.449999360.260 0
13.500 109.240 0.060 1.000 999340.834999359.380 0
14.340 109.240 0.060 1.000 999341.171999358.611 0
15.240 109.240 0.060 1.000 999341.531999357.786 0
15.241 109.480 0.060 1.000 999341.532999357.785 0
15.242 109.670 0.060 1.000 999341.532999357.784 0
15.370 109.670 0.060 1.000 999341.583999357.667 0
15.380 110.060 0.100* 1.000RIGHT 999341.587999357.658 0
16.850 110.020 0.100 1.000 999342.176999356.311 0
18.890 109.940 0.100 1.000 999342.994999354.442 0
19.320 110.010 0.100 1.000 999343.166999354.048 0

(co-ordinates edited to preserve client confidentiality)
Yet the import doesn't work.

A sample .gxy file from one of the ISIS examples looks like this:
[QTBDY__0]
X=341
Y=512


[RIVER_SECTION_0]
X=339.5
Y=487.6125


[REPLICATE__80]
X=339.5
Y=468.225

so only references one point per cross section. I've tried creating a .gxy file from the GIS data I have, but that doesnt seem to work either; possibly because it doesnt reference all the cross sections?

[Duncan Kitts]

Feel free to send the file to support@innovyze.com if you feel ICM is not importing the georeferenced information.

[Martin Spiers]

Following Will-Rust's suggestion, opened the DAT file in Excel and pulled out the x,y,z data for each cross section
Then used the open data importer to import these as if they were surveyed data.

[Martin Spiers]

Following Will-Rust's suggestion, opened the DAT file in Excel and pulled out the x,y,z data for each cross section
Then used the open data importer to import these as if they were surveyed data.

For future reference, I think the problem here is the lack of a space between the eastings and northings. Editing the DAT file to include a space here makes the import work (but may break ISIS/FloodModeller?)