From OpenStreetMap Wiki
Jump to navigation Jump to search

Verifiability issues

It is unclear where this tag is supposed to be used in contrast to waterway=river and waterway=stream in combination with intermittent=yes. Having two tagging variants for the same feature would be counterproductive.

This tag could be used to indicate sporadic ephemeral waterflow as opposed to seasonal waterflow. This is not without problems however since the term wadi/oued is used in the North African/Middle East context for both types as well as for valleys which have historically been formed by water but do no more carry even sporadic waterflow today (i.e. fossile waterways). And there is of course already seasonal=*. --Imagico (talk) 14:22, 9 October 2013 (UTC)

Perhaps the default usage for people doing imagery tracing should be intermittent=yes, reserving wadi or seasonal for use after survey or based on local knowledge? --Ceyockey (talk) 00:12, 16 October 2013 (UTC)
Does not seem like a good idea - it would still be two taggings for the same feature with the only difference being the method of survey (which should go into the source tag of either feature or changeset). Note on-the-ground surveys are not necessarily more reliable than images for determining seasonal waterflow so such tagging could not even be used to assess the reliability of the data. For precise mapping of the waterflow situation waterway=river and waterway=stream in combination with intermittent=* + seasonal=* seem to be way more precise and clear than waterway=wadi which according to the description is more like any valley that possibly carries water occasionally but where waterflow is probably non-permanent. --Imagico (talk) 12:45, 31 October 2013 (UTC)
What do you think about the idea using waterway=wadi for riverbeds only that are permanently dry, so intermittent=no would be implied? May be we should then rename "wadi" to "dry_valley" --Markus59 (talk) 08:30, 18 May 2014 (UTC)
Well - the key does not fit then, it would be more like natural=valley. But in principle a tag for indicating a wadi as a valley originally formed by waterflow independent of current presence of water could be useful. --Imagico (talk) 20:56, 19 June 2014 (UTC)
It is also problem with rendering - should it be displayed like an intermittent stream or like an intermittent river? Mateusz Konieczny (talk) 14:31, 14 January 2015 (UTC)

Suggestion for Wadi definition -

A dry riverbed/delta in an arid region that is expected to be dry year-round, except in cases of extreme weather, in which case the wadi serves as a natural channel for dangerous flash flooding.

Contrasting with seasonal which may continually have water during the appropriate season (rain / melt), or intermittent, which is dependent on melt/weather as well - but wadis are exclusively in an arid setting (this is a desert feature AFAIK), and where the presence of water is pretty much exclusively a flash flood danger (which is severe.)

As there is some overlap with intermittent, I think the major defining feature is the *expected perpetual* lack of water, except in flash flood conditions. There's no stream, no fish, no passable channel - just a muddy soup that appears and rapidly disappears vary rarely - cutting away or depositing up to 1m of sand as it cuts into the desert. (google image search "borrego springs flooding").

As these are a desert feature, and usually devoid of water, and used for offroad access (as they are often flat and sandy, as opposed to the impassible craggy areas they have cut trough), rendering in a large amount of blue, even a dashed blue line, would be somewhat of a lie (you would never go to a wash searching for water, as it's almost never there), it should be something denoting it's rough status - like a track - maybe blue and brown alternating segments of a dashed line (like a track, but thicker like a stream or river) down the the center of the wadi. It is quite difficult to map the "area" of the wadi, as it is often defined by other features (cliff walls, valley sides), and in the open desert it's exact edges change year over year as it eats away and alters the edges via erosion on a very very quick scale, though they usually stay in the same valley or flood plain area - so way mapping is quite easy via arial imagery.

Speaking of arial imagery - a wadi or wash is usually an active flood plain, probably at least once a year, and the contrast is easy to see - the surrounding area slowly builds up sun-baked rocks and vegetation (however slight), which is in contrast to the fresh (less than a year old is fresh) cuts exposing different colored material and fresh sand.

we might also want to have a major/minor tag (wadi=major wadi=minor) using the same rough standards for river size, as (in my experience with washes) many many small washes feed into large, eventuall massive ones (100m across), and denoting these large ones from smaller "stream" ones would be useful. Most of the named washes I have ever seen are major ones.

I think people who know what a wadi or wash is can use this tag, and if we emphasize the lack of seasonal water (not melt) and that the intermittent presence of water for a waterway=wadi in OSM means only flash flooding, I think most mappers looking to use this tag would be able to properly use it. Javbw (talk) 09:59, 15 January 2015 (UTC)

First of all any such definition is in conflict with currently practiced use of the tag which is frequently applied to waterways with seasonal waterflow.
Generally using the term 'expected' in the definition is not a good idea regarding Verifiability - expectations are subjective and can vary. We map based on observable facts, not on expectations. The good thing about intermittent=yes is that it makes a clear claim that there is at least sporadic waterflow. If you don't know you should not apply a waterway tag.
The alternative tagging for ephemeral waterflow based on currently established tags would be intermittent=yes + seasonal=no. If you want to further distinguish between short and longer lasting waterflow additional tags might be appropriate although this will usually be very hard to classify reliably.
--Imagico (talk) 22:06, 17 January 2015 (UTC)

wadi+highway combination?

Wadi's very frequently serve as tracks. RicoZ (talk) 12:17, 18 November 2013 (UTC)

Most probably yes, but I think these should be traced separately: the effective track will wander around the rocks and mudfilled holes. As well the exact path of the wadii will vary largely in an unclear intermittent waterbed. Each flooding will move rocks and sandbanks around. Generally the tracks are parallel, on the most stable soils and will also follow the natural tracks left by animals, with some areas cleaned by domestic animals feeders or nomads. Trucks have difficulties to follow the central path and will draw parallel tracks. In desertic areas, where rainfalls are rare, it's not exceptional that wadis invade large areas to flood them. Then in the next few months this area will be invaded by an exploding vegetation and animals will seed it all around. Trees will appear at random place but people living tend to keep these plants and will move around to preserve them. As all this is extremely variable across seasons, high precision of tracks is not needed and we can perfectly draw separate highways and waterways even if they self intersect frequently, without needing to place precise nodes on their intersection (for this we absolutely need the "intermittent=yes" tag on the waterway, so that intersections are not always requiring an explicit bridge (at different layers) or an explicit ford (on the same layer when in fact there's no stable point for the ford when the wadi is flooded: even the borders of "riverbanks" will be very variable and not clearly visible once water is gone and sands/dusts are erasing everything and animals have walked in accumulations of mud now dried, or digged holes in them to find the last water resources, and when vegetation that florished during wet periods has moved all banks, or animals have been eating these vegetations out of their ground).
So wadis are just indications of a direction to follow but not systematically followed when used. Shortcuts will be used frequently instead of following the irregular curves of wadis in water and the many small branches going nowhere. — Verdy_p (talk) 15:34, 17 March 2017 (UTC)