You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
There are a lot of issues that relate outline/outfall derivation. Furthermore, this choice has an enormous impact on the derived network. There is a sense that the current method needs improvement. In this PR I will explain the current processes and bring together existing issues. Here we can also discuss proposed improvements.
Three current options exist (two on swmmanywhere - which primarily take place in graph_utilities.py/clip_to_catchments):
Clip to subbasin (i.e., parameters.py/SubcatchmentDerivation.subbasin_clip_method='subbasin'). Where subbasin is derived either using pyflwdir.snap or, if a non-None parameter is provided, using SubcatchmentDerivation.subbasin_streamorder=<int>. In this method, all potential pipe carrying links that span two subbasins are removed - thus flows always stay within a subbasin.
Community threshold (SubcatchmentDerivation.subbasin_clip_method='community'). Subbasins are derived, as are communities using nx.louvain_communities. Communities with less than subcatchment_derivation.subbasin_membership proportion of nodes in a subbasin have their links to all other nodes in that subbasin removed.
Bypass this issue by providing a user defined outline, and removing all nodes/catchments outside of that outline (see trim_to_realhere ).
In addition to the processes above, if there are rivers near the street network, then further processing to identify which outfalls should be used takes place in graph_utilities.py/identify_outlets.
Grouping these issues below:
General DEM-related features (these will inherently interact closely with subcatchment_derivation since both rely on DEM, but it is important to note that subbasins relate to the overall drainage of the network, while subcatchment refers to the individual catchment that drains to a given manhole):
There are a lot of issues that relate outline/outfall derivation. Furthermore, this choice has an enormous impact on the derived network. There is a sense that the current method needs improvement. In this PR I will explain the current processes and bring together existing issues. Here we can also discuss proposed improvements.
Three current options exist (two on
swmmanywhere
- which primarily take place ingraph_utilities.py/clip_to_catchments
):parameters.py/SubcatchmentDerivation.subbasin_clip_method='subbasin'
). Where subbasin is derived either usingpyflwdir.snap
or, if a non-None
parameter is provided, usingSubcatchmentDerivation.subbasin_streamorder=<int>
. In this method, all potential pipe carrying links that span two subbasins are removed - thus flows always stay within a subbasin.SubcatchmentDerivation.subbasin_clip_method='community'
). Subbasins are derived, as are communities usingnx.louvain_communities
. Communities with less thansubcatchment_derivation.subbasin_membership
proportion of nodes in a subbasin have their links to all other nodes in that subbasin removed.trim_to_real
here ).In addition to the processes above, if there are rivers near the street network, then further processing to identify which outfalls should be used takes place in
graph_utilities.py/identify_outlets
.Grouping these issues below:
subcatchment_derivation
since both rely on DEM, but it is important to note that subbasins relate to the overall drainage of the network, whilesubcatchment
refers to the individual catchment that drains to a given manhole):clip_to_catchments
#165outfall
identification behaviour #314The text was updated successfully, but these errors were encountered: