QSWAT+ predefined watershed: fel.tif not created/used during Create Landscape

[Thamiris Fontoura]

Hi everyone,

I am working with QSWAT+ using a predefined watershed and channel network, not automatic delineation. My subbasins are based on USGS HUC10 polygons. I prepared the required shapefiles:

• subbasins.shp
• wshed.shp
• channels.shp
• outlets_inlets.shp

The shapefiles pass the QSWAT+ predefined watershed checks: WSNO, LINKNO, DSLINKNO, BasinNo, PolygonId, and routing relationships are valid.

However, I am having trouble during Create Landscape, especially related to the pit-filled DEM.

At first, QSWAT+ gave this error:

File C:/QSWATPlus_Projects/ACT/act_simplified/Watershed/Rasters/DEM/ DEM_30m_basin_nodata9999fel.tif for TauDEM input -fel to dinfflowdir does not exist.

But in the TauDEM log, QSWAT+ appears to run dinfflowdir using the original DEM as the -fel input:

dinfflowdir -fel C:/QSWATPlus_Projects/ACT/act_simplified/Watershed/Rasters/DEM/DEM_30m_basin_nodata9999.tif -slp C:/QSWATPlus_Projects/ACT/act_simplified/Watershed/Rasters/DEM/DEM_30m_basin_nodata9999slp.tif -ang C:/QSWATPlus_Projects/ACT/act_simplified/Watershed/Rasters/DEM/DEM_30m_basin_nodata9999ang.tif

So it looks like:

1. QSWAT+ expects DEM_30m_basin_nodata9999fel.tif to exist.
2. But dinfflowdir is being called with DEM_30m_basin_nodata9999.tif as the -fel input.
3. Therefore, slp.tif and ang.tif seem to be derived from the original DEM, not from the pit-filled DEM.
4. Later, QSWAT+ stops because the expected fel.tif file does not exist.

After that, I manually created the pit-filled DEM and named it:

DEM_30m_basin_nodata9999fel.tif

After creating this file, the Create Landscape tab became available. I first ran the buffer channels method. It completed, but the result did not look hydrologically correct to me.

Then I tried the DEM inversion method. However, QSWAT+ stopped with this error:

Dead end at (1006604.9999999998, 1402965.0) from (1006634.9999999998, 1402965.0). Landscape not created.

Even after manually creating the fel.tif, QSWAT+ still did not appear to use this file to create ang.tif and slp.tif; the TauDEM command still showed the original DEM being passed as -fel.

My questions are:

1. In a predefined watershed workflow with Create Landscape / DEM inversion, should QSWAT+ automatically create and use DEM_30m_basin_nodata9999fel.tif?
2. If I manually create the pit-filled DEM, should QSWAT+ use it as the -fel input for dinfflowdir?
3. Why would QSWAT+ still call dinfflowdir using the original DEM instead of the manually created fel.tif?
4. Could the later “Dead end” error be related to ang.tif and slp.tif being generated from the original DEM rather than the pit-filled DEM?
5. Is there a recommended workflow for using predefined HUC10 subbasins and channels with Create Landscape, especially with the DEM inversion method?

Any guidance would be greatly appreciated.

Thank you!