[Tagging] intermittent vs seasonal
61sundowner at gmail.com
Fri Oct 2 11:50:56 UTC 2015
Slowly getting to a conclusion :-)
On 2/10/2015 7:50 PM, Mateusz Konieczny wrote:
> On Fri, 2 Oct 2015 15:14:55 +1000
> Warin <61sundowner at gmail.com> wrote:
>> At the moment the 'intermittent' wiki says;
>> In Description:/Indicates that waterway or water body is intermittent
>> (seasonal) /
>> In Key:intermittent:/Or that a water body (lake) disappears
>> The use of seasonal here limits it to things that have a yearly cycle.
>> I would rather remove the 'seasonal' references .. something like
>> In Description:/Indicates that waterway or water body is
> I agree with remove of "seasonal"
>> One could add the word/irregular/ or/random/ in brackets if
> I disagree. Waterway where water is missing with some regularity is
> stil intermittent.
And I disagree with your disagreement...
If the waterway is 'typically' seasonal then it is not intermittent but seasonal.
If something is tagged with seasonal then it 'typically' has water in it .. at least for one season each year. "Typically'.
If something is tagged with intermittent then it 'typically' cannot be expected to conform to some prediction.
If I go back to Lake Eyre .. I could not say when the next time it would be full of water .. might be next year , or in 30 years time.
I cannot even say that there would be any water in it next year or next decade.
There is an El Niño event happing in the Pacific Ocean .. and that would give an indication that rain fall is less likely.
But the water in the India Ocean is warmer than normal ...meaning rainfall is more likely.
SO the meteorologists are not saying anything with certainty (but they never do anyway unless it has already happened).
Not predictable = intermittent.
>> But intermittent does not imply seasonal to me and at least some
> I completely agree. Waterway may be intermittent without making it
> seasonal. intermittent=yes do not imply seasonal=yes.
>>>> Given the weather does not conform that well to the 'season's ..
>>>> I'd be more inclined to leave off the seasonal tag.
>>> It depends on location.
>> NO. It should depend on the features characteristic. But I think we
>> agree here.
> My "it depends on location" applied to "given the weather does not
> conform that well to the 'season's".
Yep. The weather never exactly follows the mean. However OSM maps the 'typical' rather than the exceptions.
So variations around the 'typical' should not trouble the tagging/map for weather things.
>> Prior to April 2014 the wiki for intermittent had no reference to
>> 'seasonal' in its description.
> And I would not be against removing them. I am only against restricting
> intermittent=yes to random lack of water.
> Stream that disappears every summer IMHO should be tagged as
> [waterway=stream, intermittent=yes, seasonal=yes].
If it disappears every summer then it is seasonal, not intermittent.
Intermittent does not mean seasonal... and should not be compulsory with it.
In fact the use of the two together implies that for some years/seasons there may be no flow.
We don't use oneway=yes with access=permissive .. saying that the permission is only given to travel oneway... that is what the oneway tag says.
Same situation here, the seasonal tag says that it only flows seasonally. No need to define/use intermittent to say the same thing as the seasonal tag.
> In case that using intermittent for "waterway does not have a permanent
> flow/waterbody does not have a permanent presence" is clearly bad we
> need tag is needed to cover this information.
That tag exists .. it is 'intermittent' as defined before April 2014.
>> I would say
>> "/In case of waterways or lakes becoming dry on specific times of
>> year the key" //seasonal
>> should be used alone, not with the
>> stream "sample" only flows in the summer season./"
> Why? What is reasoning for this?
It is the simplest concise and precise way of tagging it. It 'typically' only flows in summer.
>> Looking at the wider wiki too !
>> suggests replacement of
>> waterway=wadi and waterway=drystream
>> waterway=river or stream
>> The wiki for
>> waterway=wadi suggests (presently) using intermittent for seasonal!
>> waterway=drystream ...errr arr yes.. also suggests using intermittent
>> for seasonal.
> Seasonal waterways by definition have no permanent flow. Current
> practise is to tag such waterway as intermittent=yes. Is there a really
> good reason to invent a new tag to convey this information and
> deprecate current meaning of intermittent=yes (that is used over 900k
> times, with some data consumers already using it)?
The meaning of 'intermittent' on the wiki was changed in April 2014 ... before that it had no relationship to seasonal. And was correct!
The change is not a depreciation but a clarification.
To me and some others the use by some in OSM of intermittent is now used incorrectly. And that should stop.
According to taginfo there are 65,467 uses of intermittent with the tag seasonal. That is less than 7% of the uses of intermittent.
These are the ones that would need checking, even so .. it is not a huge error. Simply needs to be clarified.. is it 'typically';
yearly cyclic (seasonal)
random, no relation to seasons (intermittent)
may flow seasonally in a random fashion (both seasonal and intermittent)
Use of 'intermittent' for seasonal is not a great error, and can be corrected by local mappers over time, if they notice.
But new tagging and new mappers should be able to use the tags correctly without past practices telling them to incorrectly tag things.
More information about the Tagging