Flow Direction Artifacting?

KHugeback

I am preparing to run a hydrologic model for a watershed in Central Iowa. When subsetting 3s flow direction from the North America/Central America dataset, I noticed an odd “diamond” shape when opened in QGIS. Having lived in the area for almost a decade and researching in hydrologic modelling in this region, these flow directions seem fishy. Doing my due diligence, this same artifact was present in the n40w100 3s flow direction and the 15s North America/Central America rasters as well.

Rough coordinates of the corners of the “diamond”:

Top: 42.10398,-93.86271

Bottom: 41.333457,-93.900664

Left: 41.71953,-94.33986

Right: 41.751607,-93.421709

Link to screenshot from QGIS

Any ideas on what’s going on here?

guen

KHugeback Hi KH, not exactly sure what is going on here, but this is our theory: at the time the data was produced, certain areas were classified (e.g. nuclear launch sites, etc.), and the high-resolution data was substituted with a coarser resolution dataset. This was done by the original data producers, not by the HydroSHEDS team. The HydroSHEDS team processed the DEM as it was provided, generated flow direction, and carried out manual corrections as needed. So despite the odd pattern, the data should be consistent and hydrologically correct. PS: It may be that in certain cases like this, the areas are no longer classified, so newer version of the SRTM data no longer show the artefacts. Hope this helps,

Best regards, Guenther