I have four hydropower stations in a single polygon watershed. In order to run the wateryield model, each hydropower station must have own watershed ID which must correspond with the biophysical table watershed ID. I therefore snapped all the four hydropower stations using a basemap as per the advise i got from the NatCap team. I tried all options, eg creating a single vector point (hydropower station) and running it in delineateIT, and merging all four vector points in one single shapefile and running it in delineateIT. The following are my outcomes;
- That the watersheds gpkg outputs are too tiny when I have not SNAPPED points to the nearest stream.
- When i SNAP points to the nearest stream, that out of four points in a gpkg output, only one is independent. The rest are intersecting. The second dominates the entire first and the third dominates the entire second and first combined. Of course it is understandable because all the three points draw water from one major river. What disappoints is that i keep getting similar results however much i try to set the number pixels.
Now my question is, can I still run the wateryield model using these results? Will my output make sense?
Hi @Safari1 -
1/ It’s often the case that watersheds come out tiny if they are not snapped, given the imprecision of spatial data, such that the watershed points don’t exactly match the stream network generated by that particular DEM. If the point is not on a stream, then its watershed comes out tiny.
2/ I’m having a hard time picturing what you’re seeing here - can you post an image that illustrates the problem?
If DelineateIt is not snapping points to the stream location that you know is correct, since there are only 4 hydropower stations, you could manually relocate those points to lie exactly on the stream network generated by DelineateIt. Then you wouldn’t need to snap (or if you do snap by one pixel, it will be located correctly).
As you can see the attached, point black is the independent watershed. Point red dominates both point blue and point green (as you can see the black dots). Point blue dominates point green (as you can see the red net). Point green is the faded yellow watershed.
If i do not snap to the nearest stream, the watersheds are horribly tiny. As tiny as a dot!
If i snap and set the extent, as long as its above 10 pixels for each of the two entries, I get what you see in the picture.
Hope this’s clear
Hi @Safari1 -
It’s a little hard to tell with all those black dots, but it appears that the red, blue and green hydropower stations are all located on the main stem of the same river. If that’s the case, then the area that drains into the blue station will contain all of the area that drains into the green station, plus the additional land that’s between the green and blue points, which only drains into the blue station. Similarly, the area that drains into the red station contains all of the area that drains into the blue and green stations, plus the additional land between the blue and red stations, which only drains into the red station.
So this result looks good to me. If you know that your hydropower stations are not actually located on the same main river, then you’ll need to manually move those station points so they overlap with the correct tributary.
yes you got my point. All the three points (green, blue and red) lie along one major stream as I had stated in the first inquiry. My only doubt was as to whether this is acceptable in the water yield model and whether it can produce valid results.