We use cookies to make your experience of our website better. You can delete or block cookies, but some parts of this website won't work without them. By using this website you accept our use of cookies.

chasward

01/11/2016 12:29:41

Currently FMP does not have Easting Northing fields for structures. This means that FMP cannot site structures on a georeferenced background (map). Instead the .gxy only situates the model nodes that do contain East/Northings in their correct respective positions. The structures are remain left out of position in a conceptual model space. More than occasionally these are left 'stacked' at the model origin. For some applications (i.e. dense model structure) it might be convenient to have structures out of the way (but in FMP these it can be opted to show channel icons / non-channel icons). Of course it would sometimes be useful to see structures assume their correct positions. For units such as weirs, it should be possible to see the extent of these and for spills, ideally, read the underlying grid for elevation data. For your consideration. Thanks.

Sign in to vote

0
-
+7
7
+

Bob

Flood Modeller Support

01/11/2016 16:55:55

Hi

The gxy file that is part of every Flood Modeller 1D network holds georeferencing for all model nodes in a 1D network. However, the problem you describe can occur when you load in a legacy ISIS model which does not have all structures georeferenced, i.e. it was never displayed in the old ISIS gxy visualiser and so it has no gxy generated for it. Flood Modeller therefore has no information as to where to display these structure nodes. It could try to position them between any connected river sections but this would just be a guess and so invariably they would end up incorrectly positioned. So instead Flood Modeller positions these nodes offset to the side of the main network and requires the user to move each node to correct location, i.e. using the move nodes functionality, possibly in conjunction with base mapping or background GIS layers.

Move nodes is accessed by right clicking on the map view (you can move individual nodes or groups of nodes - see user guide for details). Once all moves are complete then saving the network will update the network gxy file with all new locations for your structures and other node types. Thus, in future sessions the model will be displayed fully georeferenced.

For any new 1D network created in Flood Modeller a gxy file is generated as an integral part of the network and new nodes (of any type) are added by placing them on a map. Thus the entire network will be georeferenced.

I hope this provides useful information with respect to the georeferencing of structures in Flood Modeller Pro.

Regards

Bob

 

 

chasward

02/11/2016 09:39:23

Hi Bob: re legacy ISIS model which does not have all structures georeferenced; that is exactly the case. I wasn't aware that this info was held in the gxy, but then I've not really looked at the gxy outside of ISIS / FMP. The number of nodes that require to be moved (on just one model I'm looking at) doesn't make for a particularly entertaining afternoon! If FMP was to position the structure closest to the RS where distance to next = 0, that would be a good approximation; even stacking, say culverts at the orifice location would be an improvement.

You suggest that any new model would be georef'd; would this be by virtue of the user placing the node? As we now request surveyors to deliver data in 'ISIS' form, most nodes are imported, but presumably not where the unit doesn't hold easting/northing? Also, importing units from other models would also deposit the imported node at the origin.

I note that the gxy is text; perhaps a script could be used import and capture the salient details there.

Bob

Flood Modeller Support

02/11/2016 11:37:41

Hi Chas

Just to clarify, when I said that all new models will be georeferenced I meant that whenever a 1D network (dat file) is saved in Flood Modeller then its associated gxy file will also be saved and if a gxy file does not exist then one will be created (based on current node positions in the map view). So this does not mean that networks will all be (initially) georeferenced correctly, but map locations are now always stored for all nodes.

I agree that moving of nodes for legacy models containing many structures might be a bit tedious, but it does only need to be done the once and then the new locations will be saved. The positioning of non-georeferenced units to the left side of the map view is seen as a first iteration towards correct georeferencing. I think the ideas you suggest are doable and possible improvements that could be investigated for possible incorporation into the next release. However we would need to check to what degree of efficiency they provide - as it could be that a user still has to spend time further moving every non-georeferenced node to the correct/desired location. So the initial loading of a legacy model might look a little "tidier" but the amount of work to make it correct would still be the same (in which case why do it?).

Another option that already exists in Flood Modeller is to import a comma separated text file containing fields easting (or x), northing (or y) and node label to update the gxy file of a 1D network. The text file can be an ISIS ixy file or a text file with extension ".txt" or ".xy". So if survey information can be requested to include these data then once the model is loaded in Flood Modeller the gxy file can be updated by right-clicking on the network table and selecting (from the displayed menu); File > Import visualisation from other file.

Hope this helps.

Regards

Bob

chasward

07/11/2016 11:04:15

A draft model recently delivered 226 nodes geo-ref'd just leaving 547 to move. Again no .gxy but perhaps this will be remedied when the current year's crop begins to deliver with the new Scope. Or will we just receive a gxy with the above relationship still intact? Only time will tell.

Sign in to create or reply to forum posts.

Privacy & Terms

Copyright