In the previous article, ArcHydro: determination of the watersheds of a territory (1), we analysed points 1, 2, and 3. In this article, we will discussed the last three points:
- – Segmentation of the network flow
- – Determination of the watershed of each section of the flow network
- – Fusion of the watershed of each element in order to build the targeted watershed size
Network Segmentation Flow
We must remember, we are working in raster mode. Each pixel (cell) can contain one or many values, and they are all “independent”. Unlike the Hydrographic network which is a vector type, and, therefore we have identified sections, in this case, we are not entitled to state that two contiguous pixels belong to the same hydrographic section.
We will proceed to perform a segmentation stage in order to fulfil this need: the request will look for contiguous pixels between two 
hydrograhic nodes.
The nodes taken into consideration are: the sources, the junctions and the outlets. Those contiguous pixels between two nodes will share the same section identifier. And each section will be numbered differently.
To create this network, we use the command” Land Preprocessing »-»->« Stream Segmentation “
The resulting raster is, visually, identical to the previous one. In order to appreciate the difference you can change the symbology to “values unique”. Then, you will visualize the pixels belonging to each section with the same colour.
Establishment of the catchment area of each section of the flow network
Well, now that we have the sections of the hydrographic network, we are going to establish the catchment areas of each section, i.e. the cells that discharge in each section. Unlike the classical concept of catchment area that takes into consideration the runoff flow of the banks of the sections, we are going to consider just the direct runoff of each section, disregarding the runoff of the tributaries of the section. We will use the following command “Land Preprocessing »->« Catchment Grid delineation “
The inputs correspond to the raster and the flow direction plus the raster with the hydrographic sections. The output corresponds to the watersheds for each section.
Now, we want to observe the watersheds as we know them, as a whole, not section by section, and that is what we are going to be doing in the next stage.
Fusion of the watershed of each element in order to build the targeted watershed size
Let’s leave the raster world to join the vector world. Even though there is a way of doing it with the rasters, it is much more easy to work in vector mode.
Therefore, we are going to convert in vectors the watersheds of the different section as well as the sections network.
To convert the sections network, we will use the command “Terrain Preprocessing » -> « Catchment Polygon Processing »
To convert the network of the different sections into polylines we use the command ” Land Preprocessing »->« Drainage Lines Processing “
Now you can observe that the characteristics of each section allow us to recognize the limiting nodes as well as the following downstream section.
We will now calculate the contiguous watersheds with the “Assistant Catchment Processing”.
This function generates the aggregated catchment, cumulative and contiguous for each catchment of the hydrographic section. In each catchment generated, the polygon defines the entire upstream zone contiguous to the section considered. If two sections flow to the section being considered, the polygon surrounds the three catchments of the sections, and a new catchment is generated.
We perform this process until no longer there are sections flowing to the aggregated catchment.
To execute this command we use Land Preprocessing »->« Assistant Catchment Processing “. The entries are the drainage lines, and the polygon layer is the catchment. The result is stored in a new polygon vector layer (adjoint catchment) that you have the chance to rename.
This layer will us allow to perform a series of very interesting calculations, but the command also adds an attributes’ option in the drainage line together with the identification of the catchment of the section.
This layer will allow us to achieve all a series of calculations interesting, but the command also added an attribute field in the table of Catchment “With the identifier of the basin downstream and another attribute field DrainageLine in the table with the identifier of thesection of the watershed
