This article is the continuation of the previous article: Hydrology with a GIS, for dummies (that we are): calculation of the flow (1)
Now, we will discuss the result of the application of another flow calculation method.
Second method: Rho8
This is a modification of the previous method. It’s always a one-dimensional flow: all the water in the cell goes to one and only one other cell. If in the D8 method, the flow is calculated according to the largest slope between the cell considered and its 8 contiguous cells, in the Rho8 method we will introduce a random factor, calculated as of the orientation of the central cell and the direction of the two contiguous cells located in the same direction.
With QGis you will use, in the processing window- > SAGA -> Terrain analysis-Hydrology -> Catchment area ( parallel )
The result of this method on the DEM in semi-spherical form, for the catchment area layer is as follows:
If you now compare this image with the result of method D8:
You can observe that the concentration on the 8 main directions (0, 45, 90,135 …) is much less marked. On the other hand, since we are on a perfect sphere, water should flow in a homogeneous way and not in the form of “string”.
To better understand this comment, we will move to another method of calculation.
Third method: Multiple Flow Direction (MFD) also called FD8
Unlike the two previous methods, the flow from a cell is not in the direction of one and only one cell. According to the slope, all cells located below the cell concerned will receive a portion of the flow.
The following figure is sourced from the original publication of this method.
You can consult this publication at: Quinn, P.F.; Beven, K.J.; Chevallier, P.; Planchon, O.; The prediction of hillslope flow paths for distributed hydrological modelling using digital terrain models, Hydrological Processes, 5:59-79.1991.
To implement this method, with QGis , you will use the treatment window- > SAGA -> Terrain analysis-Hydrology -> Catchment area ( parallel )
The result of this method on the DEM of semi-spherical form, for the catchment area layer is as follows:
Here we have a globally uniform flow, on a uniform semi-sphere. If we zoom in, we observe that the structure at 8 prints some screening. Indeed, even if the flows are multiple, they are calculated from the centre of the central cell to the centres of the contiguous cells. The frame 0, 45, 90 … appears here, but when we look at a level of greater detail it disappears when we look at the phenomenon from above.
In the next article we will discuss the remaining methods of calculations.