Hopefully I’m in the right place but feel free to move it if there is a more appropriate place. Whilst the query relates to a specific area it raises general principles such as the origin point of a watercourse and where surface water pipes should be coded or not shown in OSM if they are part of the urban drainage pipe network.
A (large) area of farmland ( Way: 72721848 | OpenStreetMap ) has been (being) developed which is on the watershed of two distinct drainage basins (Ouse and Thames). New ditches and pipes have been constructed and all the new elements in the past were named Ouzel Brook (as a stream) to the intersection with Houghton Brook.
Quite comfortable to remove the “Ouzel Brook” name from the stream elements that drain into Houghton Brook. The watershed seems to occur at the Lidl access road which is confirmed from the Defra water catchments and levels derived from Lidar although I have not seem the pipe, don’t have access to pipe details (invert levels) to verify this (or even understand why a pipe would be constructed to join ditches on either side of the road if this is the watershed !!).
Now the questions :
What defines the source / origin of a stream. Is it the historical location shown on OS maps and adopted by Defra, or the furthest extension of newly constructed ditches and pipes (which in this case are dry except in times of heavy rainfall ?).
When subsequent development removes the ditches and it is not known whether the drainage course has been piped should this be coded as ‘tunnel=culvert’ ie a pipe ?.
My personal view would be to record the drainage course to the point it disappears underground and doesn’t resurface which seems the usual approach in an established urban area which has not been developed in several phases over a period of time. In this particular case where the ditches connecting to Ouzel Brook seem to be mainly providing for ground run-off a ditch classification may be more appropriate. Further upstream they have disappeared entirely and may have been provided as a temporary measure to drain unpaved land prior to development and the installation of permanent drainage.
Life would be easier if construction drawings were available. Whilst I hesitate to remove details I believe this would make it similar to the way in which other drainage is coded in urban areas in that the normal storm water pipes draining roads, buildings and adjacent urban areas are not coded. Hopefully, the attached sketch illustrates the issues and I look forward to the views of more experienced mappers.
I can’t really answer your two questions, but sometimes 1. will relate to springs and collections of springs.
I might have an answer to joining ditches either side of a watershed. This was not my water profession, so planning timing and nation may be important as well as my understanding. (I no longer have day to day access to my ex-colleagues to check ideas why a ditch joins both sides, but it might fall both ways from a central highpoint.) As part of the planning process, a requirement was recently introduced that required that no additional rate of water run-off would enter surface water ways and drainage after a development as occurred before the development. This would reduce the risk of downstream flooding. If an area of grass shed 1 litre/second during rainfall, then if that grass was replaced by asphalt, then it could only shed 1 l/s during similar rainfall. So an easy solution is to build a detention tank of a size based on typical rainfall events with a pump to the drainage systems that would operate in a manner to keep within the 1 l/s limit.
On a watershed, my thought would be that you would not create an independent solution for each part, but would gather both volumes into one tank, then maybe pump or percolate into both water systems after the rain has passed.
The intermittent reservoir south of the Lidl warehouse looks like a detention pond as part of a similar Sustainable Drainage Solution (SuDS) for the warehouse. I suspect the adjacent pond may be a retention pond that overflows to the detention pond.
You might be able to access design drawings via the local planning portal, but it’s likely that as-built drawings would be needed as there would probably be some adaptation to conditions as found during construction. Then there’s any licence issue that might to attached to a drawing, there is usually at least a copyright of some kind.
More generally, unless you have as-built drawing information, I’d not map underground pipes and tanks. You can’t assume straight lines between manhole covers, even if you have the foul and surface water separation correct. It is possible for several manhole covers to be access points to different parts of underground tanks with grids, overflows or syphons within.
Someone has mapped some local to me that are not the same as the drawing records held by the asset owner & maintainer. Unfortunately the information is closed data, so I could not use it as a source to correct OSM when I did have access to it. Unfortunately NUAR is unlikely to ever be open, but maybe some parts of it might in time. Government pace time…
You are right there are numerous balancing ponds in the area which confuses the drainage picture. I had considered using digdat but decided against the £30 or so fee. It seems gone are the days of being able to walk into an office to view plans or make informal enquiries.
I am always hesitant to make changes in case the originator is more knowledgable. However, in this case a mistake had been made extending Ouzel Brook to join with Houghton Brook – understandable as on plan it all looks part of the same system. I waded in when GoogleMaps also assigned the Ouzel Brook legend to the streams joining Houghton Brook (which it initially labelled River Lea !!) and then the issue arose of what elements should be labelled Ouzel Brook and the direction of flow for the ditches and pipes.
In trying to clarify the watershed and extent of Ouzel Brook I made enquiries with Defra / Environment Agency who provided a “helpful” response but unfortunately no response from the local drainage engineers (Flood Risk Team) who would have been familiar with the local network. Hence the suggested solution to only show details on OSM which have little doubt but quite open to suggestions from those with more experience.
