Problem encountered in delineating an endorheic river basin
Hemant Servia
In the present case, we are trying to delineate a sub-basin which is an endorheic basin. When we are trying to delineate using the stream delineation, it is not depicting the correct direction of streams. This may be due to relatively lesser slope gradients. The DEM opted initially was SRTM90, which was then replaced by SRTM 30 from USGS, but the results obtained were no better.
The next option was to use the existing watershed feature of QSWAT+. The 3 main inputs were provided as under:
The sub-basins were obtained from HydroBasins Level-12, provided by HydroSHEDS
The basin was also obtained from the same source by merging the sub-basins above
The channel data was obtained from Hydrorivers data, again provided by HydroSHEDS. It is important to note here that the river data is corresponding to HydroBasins Level-12 due to which HydroBasins Level-12 were selected at the first place. Moreover, all the 3 shapefiles were adjusted as per the SWAT+ requirements, such as the outlet channel to have DSLINKNO as -1, the last sub-basin to have the next sub-basin as 0 and other naming conventions. These were executed as per the SWAT+ manual (v2.0 April 2021).
The link to all these files and the entire QGIS model can be found here.
As a result of the delineation, all the 83 sub-basins were generated. Please see Figure 1.
Figure 1: Delineation of watershed
Even the next step “Create HRUs” became active. Please see Figure 2.
Figure 2: Activation of the next step
Now when the landuse and soil data, available from the global database as available from the SWAT website were used as input, an error saying “WARNING: 0.0 percent of the watershed has defined landuse values. If the percentage is zero check your landuse map has the same projection as your DEM”. Please see Figure 3.
Figure 3: Error in HRU creation
The same error occurs for the soil data. Whereas the projections of both the datasets have been checked extensively and repeatedly (with many trials) to be the same as DEM (EPSG: 32719), there seems to be some technical issue, either with regards to the SWAT+ or there is something that I am missing.
