There are numerous intermittent waterways mapped in the Carmel region. I’ve physically sampled some of them, and many have no trace of a waterway, or have really negligible water-flow to the point I don’t think it’s worth mapping.
I would like to remove most intermittent waterways that have no name tag, and add the name to a few that lack them (such as נחל נשר). Significant waterways typically have names so this sounds like a good cleanup heuristic to me. If a mapper encounters a nameless but significant stream, they can re-add it later.
I’ll try to be more careful in the more downstream parts (as they might have collected enough water to be significant), but I think the nameless upper branches can simply be mass removed.
The original mapper, @BMM994 , said they were created long ago using a contour map and not a survey, and are probably not entirely accurate. He has no objection to this plan.
Please raise any objections or comments if you have them!
I have no objection. A couple of years ago I saw such a stream going through my neighborhood, mapped even with culverts that are completely made up. In the end I corrected the parts of it that I could, but upstream I have not checked it.
The only drawback if you delete them is that future mappers won’t see them like I did and go “that’s wrong, let’s fix it”. If that false stream wasn’t there, I never would have mapped the real drain that’s actually there.
But “keep bad data as motivation to fix it” is perhaps flawed reasoning, so make of that what you will. Especially if you only delete streams that you know aren’t actually there, then there is no such concern.
In my neighborhood there are mapped streams with no name, but they do have water in them when it’s raining, so it makes sense to leave those that do “carry” the water isn’t it? In any case I don’t think I’ll miss them much, but they do help understand the terrain better sometimes.
They were added heuristically based on a contour map, and not based on a survey. So I wonder if the approach of “keep until it’s proven there’s no water” is upside down here, and should in fact be “delete them all, and map the water if found”. I suppose it depends on how many of them are falsely mapped.
I am not planning to work on this now. Just sharing thoughts.
I think it makes sense to keep them if they do accumulate water in winter.
It seems there is a gray area of extremely thin branches barely accumulating anything. I think the original contour based mapping was overly loose with the cutoff threshold.
I have another data point. Mapper AvivBenSha reported they deleted waterways that do not exist. I checked their histories and they all trace back to @BMM994 , so, apparently, the contour line based waterways are inaccurate outside of Carmel region too.
Hi. As you can see I am new to this whole thing but I find it very interesting and fun. This is an area I have grown up in. I can assure you there are no waterways in any shape or form in any time of the year so I deleted them…
Why is this Wikipedia distinction imported to OSM when OSM does not make such a distinction, AFAIK? Have you suggested a new OSM tag for ephemeral streams?
For the sake of completeness, summarizing the private correspondence I had with @zstadler a while ago.
We’ve agreed that the issue is simply that of verifiability: The mass-import has introduced many false-positive streams that are simply not verifiable and are not visible on the ground. In other words, two mappers visiting the spot wouldn’t necessarily map the same stream or would not see/map one at all. There is often no water nor any indication right after the rain, and the mapped “stream” is really just rain pouring down, and sometimes not even that.
The definition of intermittent/seasonal/transient streams/rivers, though interesting, was a bit of a red herring.
In some areas (e.g. בקעת אלון), the intermittent stream is very distinct and objectively verifiable downstream, but it becomes a gray area as one travels up in elevation, and at some point it’s absolutely a mapping of something nonexistent.
רציתי לטפל בבעיות crossing ways אבל בלי לראות ספציפית crossing ways עם ה- waterways ש-BBM994 הכניס.
בעיה נקודתית נוספת היא ש-osmcha לא מאפשר סינון של “NOT” עבור crossing ways זאת אומרת, אם רוצים להתעלם לחלוטין מ-crossing ways אבל לא משאר ההתראות צריך להפעיל את השאר אחד אחד (מה שגורם ל-URL ענק ולשגיאות state כשמבקרים בו שוב).
ככל שאני חושב על זה יותר אני חושב שצריך למחוק. זה פשוט מידע שהוכנס אוטומטית ואיכותו לא משהו.
הבנתי. אני לא מתכוון לעשות mass-deletion בלי הסכמה יותר רחבה.
כן מודה שכשאני עורך אזור מקומי, אני כבר די trigger happy על המחיקה. אני בגישה שזה כנראה שגוי ומקסימום אני אוסיף על סמך survey אחר כך (אני קודם מוודא שזה של BMM994 ובלי שם)