Jump to content
TUFLOW Forum

MFedak

Members
  • Content Count

    11
  • Joined

  • Last visited

Community Reputation

0 Neutral

About MFedak

  • Rank
    Member
  1. I do not know about SMS and viewing multiple photos at once. If your images are overlapping and properly georeferenced, you should be able to mosaic them. You can do this with GIS packages like ArcGIS, ArcInfo, GRASS, PCI Geomatica, ERDAS, etc... The result will be one image. Hopefully this helps.
  2. I've narrowed down the problem to machines running Windows 7 home premium. When I try to launch TUFLOW it brings up the error that WIBUCM61.dll is missing from your computer. On my workstation, which runs Windows 7 Ultimate, there are no such problems.
  3. I personally use tuflow_to_GIS and export results as an ESRI GRID file, for ease I run the program in the same directory as the mesh and dat files. For exporting multiple timesteps or from multiple simulations , I usually use a batch file.
  4. Hello, I have attempted to use TUFLOW-2010 AC with SMS 10.1 The 32-bit executables work perfectly but the 64-bit do not because of licensing issues. Is this a bug or are users of SMS restricted to the 32-bit executables? Thanks in advance.
  5. Sorry, reading over your first post again; since you are 5km from the coastline upstream you definitely need to get some cross sections of your estuary. To save time and space, you may be able to model some stretch of it with 2D and the rest as 1D.
  6. Sounds like you need to create a merged DEM between the bathy and LiDAR points. Using a TIN works but the transition between the sparsely spaced bathy points and the close LiDAR points is usually quite coarse. I would recommend merging the points, then interpolating a new DEM from the merged point layer using something like a spline interpolator, which should create a smooth surface. If that doesn't work then just use the TIN, that tends not to create the kinds of artifacts an interpolation function might when you are going from densely spaced to finely spaced data. Also, I haven't tried importing a 3D shapefile in 10.1 yet but in 10.0 I found that it was dismally slow and unstable when dealing with a large number of points. I always found it easier just to export a .xyz file from arcgis and import that into SMS.
  7. The Read MI PO command is in a .tcf file (Manual A.7). You can create a separate .tcf file containing the PO commands and just point SMS to it using the read external file command in the Misc tab of model control. Be sure to add the type of outputs you want as an attribute in the "Type" field. I have been working on this lately so hopefully this helps.
  8. I have been trying to do a similar thing and have had to create a PO .mif file.
  9. I am having trouble with obtaining time series output from TUFLOW using the PO/LP commands (manual 4.8). I create a .mif file with the profiles I want to extract information for and then I point the model to the files (read MI LP/PO commands from manual A.7). I do not get output as .csv files or a _TS.mif file. So what have I missed here?
  10. Thank you very much for the responses, they certainly helped. Firstly, I have changed the location of the tidal BC to make it perpendicular to the direction of flow entering the model. I now have the 1D channel embedded within the 2D model and transfer of flow between the models seems to be occuring, the connection cells around the HX boundary all show up within the _1d_to_2d_check.mif. The issue was that the tide BC was in a different coverage than the HX lines, putting them in the same coverage makes perfect sense and I should have realized it earlier. However, there are a few new problems. 1. Flow within the channel is not represented inconsistently, Im guessing this is due to the configuration of the water level arcs, I probably need to insert more of them. 2. The model now becomes critically unstable about 5 hours of simulation time. I am looking at the _messages.mif file and there seem to be several Negative V depth warnings before the UNSTABLE 2999 warnings. These warnings seem to be outiside of the 1D domain with the exception of the 1D node at the downstream end of the network. From reading the manual, I gather that I should be investigating the DTM I am using. However, for purely 2D simulations forced only by the tide, these errors do not seem to occur. Adding the flow BC into a simulation seems to cause these errors, so I am wondering if modifying the grid, as suggested, will just be a waste of time if the 1D network is responsible for the problem. What is the best way to deal with these errors using SMS? Will I need to smooth out elevation values? if so, then perhaps I should try a different interpolation technique for creating the grid or should I just change them manually? -Additionally, I have been looking at the _MB1D.csv file and in the H V In and H V out columns there are no entries so I suspec that the flow BC is not applied correctly; should the 1D BC also be in the same coverage as the 2D? 3. An error does not seem to be created so I am assuming this is not an issue at the moment (I have left the invert values as the default -99999). Thanks, I figured as much after running a model for a week and watching flooded areas disappear gradually. Is there a type of observation coverage I could apply to assess how much water is actually lost in a given area from an initial amount? Anyways, sorry about all the questions but you have been very helpful so far so I figured I would ask. Thanks, Michael
  11. Hello, Instead of stating a new thread on this topic I thought I would comment here. I am currently working on a model of a macrotial estuary and the forum/manual/tutorials have been very helpful. However I seem to be somewhat stalled attempting what Phillip Ryan has described. I am attempting to see the interaction between the tide/river flow (hopefully my image illustrates that). Right now I am mostly working on familiarizing myself with SMS/TUFLOW so I am using a fairly coarse grid (LiDAR data re-sampled to 60m resolution). So far I have managed to use a HT BC along with a flow vs time BC from the river, the results have been alright and the model stable. The step I am currently on involves trying to use a 1D/2D model to give additional detail within a portion of the river channel. This has not worked for me so far, the model only seems to recognize the upstream 1D BC, I have connected the downstream HX boundary with the network thinking that flow/water level would be transfered between the 1D model forced by upstream flow and the 2D model forced by tides. The tidal forcing does not seem to be taking place. Perhaps the only way to do this is to link models (like Phillip Ryan shows)? If so, then where could I find information about how to do that? Also, this is a bit of a different topic but how exactly are hydrologic losses initialized using SMS? I have defined an initial/continuing loss using the GUI but this does not seem to be having an effect (material set not shown in image). Is this something I need to edit into the tfc file after exporting from SMS? Thanks,
×
×
  • Create New...