Proposed features/Landcover Barren

From OpenStreetMap Wiki
Jump to: navigation, search
Barren Land Cover
Status: Proposals without post-vote cleanup
Proposed by: LorenzoStucchi
Tagging: landcover=barren
Applies to: area, relation
Definition: Natural area without vegetation
Drafted on: 2019-04-05
RFC start: 2019-04-06
Vote start: 2019-04-26
Vote end: 2019-05-10

Proposal

The proposal is from Eng. Gorica Bratic and from the idea of mapping deforestation and proposes this tag as a general class of an natural area without vegetation and with bare soil.

Rationale

The tag barren is meant to delineate land without or with very sparse vegetation (known also as bareland or barren land). The main objective of introducing this tag is to track deforestation in Amazonian rainforest. Currently available tags (i.e. natural=bare_rock and natural=sand ) do not encompass bare soil that is a result of vegetation loss (i.e. soil where vegetation was removed by human, or due to some human-caused or natural processes affecting the soil fertility). As a matter of fact, if the bare soil contains organic matter (as if it case with deforested areas), most probably the vegetation will grow back unless it is further modified by human or affected by other processes like erosion. In any case, mapping this transient period is important as it might last couple of years. An example of long-lasting transient period is shown on the two Planet Scope Scene captured with 3 years difference. It is possible also to check on Bing imagery how the vegetation is extremely sparse on the same area.

Barren soil in Amazon in 2016
Barren soil in Amazon in 2019
Barren soil in Amazon in Bing Imagery

Other than deforestation, area can be characterized as barren whenever the vegetation loss takes place due to the contamination, fire, droughts, soil erosion, etc. The tag barren can generally encompass all the natural surfaces where the vegetation is missing due to natural or man-caused absence of nutrients, natural disasters or human impact like deserts, cliffs, beaches, burnt and deforested areas. The tag barren is very broad in order to make up for lack of local knowledge due to limited access to the rain forests. Sometimes it is hard to distinguish certain types of barren land like natural=scree and natural=shingle even when the satellite imagery is very good. Broader tag also meets the needs of the community to have a tag which is reusable. More specific tags such as natural=sand and natural=beach are very welcome, but in their absence any other (less specific, but correct) information linked to Amazon deforestation is appreciated in understanding underlying processes in order to prevent and stop them.

Tag is related to land cover, and not land use. As previously mentioned, local knowledge is scarce, therefore we have to rely on satellite imagery for mapping in rain forests. Land use is often only known locally, while land cover is visible from imagery.

Examples

Examples for the barren land areas in the forest where due to excessive clearcutting the vegetation is removed. The most evident examples are in Amazonian rainforest, but also forest in Democratic Republic of Congo, Indonesia, etc. Forests are also influenced by fires. In example of contaminated sites, locations of barren areas are usually in industrial zones of a city.

Tagging

landcover=barren is proposed in order to support mapping of bareland areas by means of visual interpretation of satellite imagery. The visual interpretation is especially convenient for mapping larger areas, like it is the interest here (deforestation of Amazon rainforest which has area of 5,500,000 km2). The existing tags as natural=bare_rock and natural=sand do not describe barren land that is a result of deforestation in the first place, and also they might be too specific to be recognized from satellite imagery. Observing if areas are vegetated or not might be much easier. Due to the unpredictable duration of barren land cover, it is suggested for the tag landcover=barren to be accompanied with key {tag|fixme}} to point out uncertainty about duration of this land cover type. i.e. landcover=barren, fixme=”Prone to changes. Revise feature if older than 6 months” or key end_date=* to let user’s know when the tag is might not be valid anymore and needs to be checked in order to confirm its validity. In this way, if feature is revised regularly, period of revegetation can be observed as well, and estimate how serious the consequences of deforestation are.

Applies to

Area, Relation

Rendering

Light brown area

Features/Pages affected

Comments

Please comment on the discussion page.

Voting

First Voting

  • I oppose this proposal I oppose this proposal. Deforested land is usually dedicated to other uses such as farming or agriculture. Land that's left without immediate use for other purposes will quickly be overgrown by grass, shrubs and other conquering species and thus no longer be barren -- that process starts after only a few days even here in Central Europe. The kind of mapping you suggest sounds more like a thematic map featuring short-term changes in the landscape, rather than a long-term map as OSM should be understood. --Grimpeur78 (talk) 10:05, 24 April 2019 (UTC)
Deforestation happens also for the reason illegal timber harvesting. In such a case, there’s no intention to use land for anything, just to profit from the timber. In case that timber harvesting is on a large scale, this makes the soil prone to erosion, and therefore, it is hard for vegetation to recover if the soil is often eroded. Erosion is highly probable due to the heavy rains in the rainforest. Central Europe is less affected by this type of rains. Furthermore, soil contaminated due to mining activities becomes infertile. This is the reason for vegetation loss, and vegetation on these sites does not recover quickly. As mentioned in the talk page, I think this tag should be accompanied by the key fixme to point out uncertainty about duration of this land cover type i.e. landcover=barren, fixme=”Prone to changes. Revise feature if older than 6 months” or key end_date=* to let user’s know when the tag is might not be valid anymore and needs to be checked in order to confirm its validity. —Gorica7 (talk) 21:32, 25 April 2019 (UTC)
  • I approve this proposal I approve this proposal. I understand the proposal is purely about landcover, not landuse. --Peter Elderson (talk) 10:37, 24 April 2019 (UTC)
Can you please explain why you vote no? Thanks --LorenzoStucchi (talk) 07:11, 26 April 2019 (UTC)
  • I oppose this proposal I oppose this proposal. As I mentioned on the talk page, "At OSM we want to map specific features that are real and current and that individual mappers can verify on the ground or with good quality aerial imagery. Landcover=barren does not meet these standards. It's a mix of several different features that already have established tags: natural=sand, natural=beach, Tag:natural=scree, natural=shingle, natural=mud, natural=glacier and natural=scrub. In most tropical regions, if a patch of forested land has been recently cut or burned, it will quickly regrow into the previous vegetation type." --Jeisenbe (talk) 14:23, 24 April 2019 (UTC)
It is very hard to obtain on-site information (wild-life environment), and satellite imagery has high cloud coverage that affects availability of high-resolution imagery. This could imply that this areas won’t be ever mapped. We think that having at least general features is better than having no features. Also, regrowth is affected by erosion (please check my previous comment), so it might not happen so quick. It is explained more in detail in our reply on the talk pageGorica7 (talk) 21:32, 25 April 2019 (UTC)
  • I oppose this proposal I oppose this proposal. There are several problems here. For start, name is mismatching definition, why "barren" would include just "Soil area without vegetation" and exclude rocks, sand etc. Also, "soil without vegetation" is extremely temporary. Either it gets overgrown or eroded away. Proposal claims "it might last couple of years" but it is highly surprising to me that it would take so long for grass/herbs to start growing - can I get some link to research confirming this? In my experience in Europe plants start growing in soil within days/weeks, and I would expect it to take years only in areas that are dry (and would not be deforested due to lack of forests). There are also several other problems, some of them mentioned above and some will be probably mentioned below. Mateusz Konieczny (talk) 15:31, 24 April 2019 (UTC)
You are right. There is an error in the definition. Barren is not supposed to include only soil. This can be seen throughout description of the tag.
As you may notice in the previous comments, I stressed the effect of erosion on deforested land vegetation regrowth.I could not find a research exactly about re-vegetation period. However, I found this paper that is saying about soil loss and how it reduces soil fertility. The point here is that erosion can affect the period of regrowth depending on it’s severity, therefore the period of re-growth is not certain. As I mentioned in the talk page, I think this tag should be accompanied by the key fixme to point out uncertainty about duration of this land cover type i.e. landcover=barren, fixme=”Prone to changes. Revise feature if older than 6 months” or key end_date=* to let user’s know when the tag is might not be valid anymore and needs to be checked in order to confirm its validity.—Gorica7 (talk) 21:32, 25 April 2019 (UTC)
  • I oppose this proposal I oppose this proposal. This would not map what is there, but what is not there i.e. trees. Map the presence of trees and the deforestation, which is what the proposer wants, can be obtained from the changed areas over time. Warin61 (talk) 22:02, 24 April 2019 (UTC)
The aim of this mapping is also to identify what replaced the forest (i.e. agricultural areas, barren land (unused)...). We’re hoping in this way that the collected data will be useful for the analysis of causes and help in defining mitigation measures.—Gorica7 (talk) 21:32, 25 April 2019 (UTC)
Landcover=barren does not say landuse=unused - and that would have the same opposition. If there is rock - map it. If there is soil - map it. If there is cloud cover, chose a different image, sometimes zooming in/out will change the image too. If there is agriculture then map it - landuse! Still no from me. Warin61 (talk) 06:10, 26 April 2019 (UTC)
This is exactly your idea if we want to map just the soil what is the correct tag? --LorenzoStucchi (talk) 07:11, 26 April 2019 (UTC)
  • I oppose this proposal I oppose this proposal. The OpenStreetMap landcover tag is about what the surface is made of, “barren” does not fit well into this concept, because it describes the absence of something.—Dieterdreist (talk) 07:35, 25 April 2019 (UTC)
Barren is indeed surface (land cover). Absence of vegetation is a way to describe barren land cover, but the land cover itself is present (not absent). To give an example, rock is natural non-vegetated surface. The lack of vegetation here does not mean that the rock surface is missing.—Gorica7 (talk) 21:32, 25 April 2019 (UTC)
IMHO the value for this case would be something like „rock“, not „barren“. This is exactly what I meant.—Dieterdreist (talk) 21:54, 25 April 2019 (UTC)
But if there is not rock but only soil, what is the correct tag? --LorenzoStucchi (talk) 07:11, 26 April 2019 (UTC)
if there is only soil you would wait some weeks for spontaneous vegetation to grow ;-) —Dieterdreist (talk) 13:04, 26 April 2019 (UTC)
But if there is some type of pollution and nothing will grow, we just have to wait and map like there is grass and we are not sure if it will be in future --LorenzoStucchi (talk) 15:42, 26 April 2019 (UTC)
  • I oppose this proposal I oppose this proposal. As other's have stated repeatedly, the same could be achieved using existing tags --Fizzie41 (talk) 02:19, 26 April 2019 (UTC)
What are in your opinion the correct tag to be used for this area? --LorenzoStucchi (talk) 07:11, 26 April 2019 (UTC)

Second Voting

After the first vote we found some misunderstood in the description so we improved it, also with some graphical example.

Instructions for voting
  • Log in to the wiki if you are not already logged in.
  • Scroll down to voting and click 'Edit source'. Copy and paste the appropriate code from this table on its own line at the bottom of the text area:
To get this output you type Description
  • I approve this proposal I approve this proposal.
{{vote|yes}} --~~~~
  • I oppose this proposal I oppose this proposal. reason
{{vote|no}} reason --~~~~ Replace reason with your reason(s) for voting no.
  • I abstain from voting but have comments I have comments but abstain from voting on this proposal. comments
{{vote|abstain}} comments --~~~~ If you want don't want to vote but have comments. Replace comments with your comments.

Note: The ~~~~ automatically inserts your name and the current date.

Can you please explain why you are against the proposal? Thanks --LorenzoStucchi (talk) 21:40, 26 April 2019 (UTC)
Note that users are not obligated to provide reasoning for vote and badgering them to provide one is a poor form. Especially in cases where there are many comments explaining what is wrong with the proposal Mateusz Konieczny (talk) 07:09, 27 April 2019 (UTC) misleading, see link below
Sorry but users have to provide a reason for the vote, they don't have to provide a suggestion. This quote is from the rule page "People should not just vote "oppose", they should give a reason for their proposal, and/or (preferably) suggestions." We want to improve our proposal and make it the best for all, so most different views we have better will be the proposal. --LorenzoStucchi (talk) 07:25, 27 April 2019 (UTC)
Sorry, I was not aware about this. Thanks for a link Mateusz Konieczny (talk) 07:28, 27 April 2019 (UTC)
  • I abstain from voting but have comments I have comments but abstain from voting on this proposal. Told you so - your thinking does not match the mainstream OSM mapper's view --Peter Elderson (talk) 21:46, 26 April 2019 (UTC)
  • I oppose this proposal I oppose this proposal. I don't see a reasonable use case for this. Consider the following aerial views: [1][2][3]. All three look fairly similar, but only the first is "barren" as defined in this proposal (it's a pumice field); I know this because I've observed all three on the ground. Since "barren" isn't something that can be reliably surveyed from the air, and ground surveys give you the information needed for more detailed tagging, we don't need this tag. --Carnildo (talk) 04:39, 27 April 2019 (UTC)
Thanks for this example, I check them also on OSM and the last two are not mapped, as you said the first should be a barren soil, but this is the only feature that is mapped on OSM and is mapped as natural=scree it's correct or should be more correct that this is barren land? --LorenzoStucchi (talk) 07:47, 27 April 2019 (UTC)
Scree is...not too wrong. It's a pumice deposit from an eruption of Mount Mazama, and has almost no vegetation because of how well water drains through it. Something like natural=pumice_deposit or the little-used natural=gravel would be more accurate, while landcover=barren would be more generic. --Carnildo (talk) 19:21, 27 April 2019 (UTC)
But since the tag that you proposed are not existing, should be better to describe it as landcover=barren that is generica but not wrong --LorenzoStucchi (talk) 07:37, 4 May 2019 (UTC)
This is the view in iD editor you can see some rare spot of grass inside the land, it's not a lake because there is water near and looks very different and you can suppose that is not farmland because is far from every building and have a strange and irregular shape. --LorenzoStucchi (talk) 07:47, 27 April 2019 (UTC)
Strange and irregular farmland is common around me, because fields are often bounded by the contours of slopes too steep to safely plow. I would have guessed it as a meadow, because I live in a fire ecology area and vegetation re-grows rapidly after a disturbance. --Carnildo (talk) 19:21, 27 April 2019 (UTC)
Yes but in the middle of the forest this can be a parameter to consider if the area is agricultural or not. --LorenzoStucchi (talk) 07:37, 4 May 2019 (UTC)
  • I oppose this proposal I oppose this proposal. My comments above have not changed. I will add that a tag like "natural=clay" or "natural=bare_soil" could be useful for desert areas, where it is possible to find clay, silt or mixed soils without vegetation and not in a wetland (We already have natural=mud for wet areas of silt and clay soils, and natural=sand for sandy soils). However, there is no need for a general tag for areas that could be tagged more specifically as natural=sand, natural=gravel, natural=scree, natural=shingle, natural=bare_rock. If the aerial imagery is good enough to show that there is no vegetation (not even grasses or small shrubs), then it's usually also good enough to distinguish between sand, small rocks and big rocks. --Jeisenbe (talk) 14:05, 27 April 2019 (UTC)
Yes can be enough good to show distinguish the type of bare land, but if just bare soil doesn't exist a tag for it, so this is general but describe this situation --LorenzoStucchi (talk) 07:37, 4 May 2019 (UTC)
  • I oppose this proposal I oppose this proposal. First of all, we shouldn't vote on a new tag as long as there is no clear definition for the key. Secondly, the definition of the tag itself ("natural area without vegetation") contradicts the proposal rationale (deforested areas). E.g. oceans and glaciers are natural areas without vegetation, but not deforested; while a meadow is a deforested area but with vegetation. Apart from that, the fact that an area is deforested adds a historic context to the tag and by this means to the key as well, making it a competitor to the historic=* key. I agree with others that natural=* can and should be used for deforested areas. When you are armchair mapping from aerial images and all you can see is that there is no forest, then all you need to do is not to map a forest there (see Multipolygon on how to exempt an area from a forest). --Fkv (talk) 13:03, 29 April 2019 (UTC)
The previous definition talks about "soil without vegetation" and we have to change it because was too general and don't describe what we want, are just a few words I think that the important part is the concept of land that is without vegetation and that doesn't grow. Yes but multipoligon is always complex to be mapped and can be misunderstood by other users if they saw it with different images if there is also a description of the inner area they can better understand the previous mapping. --LorenzoStucchi (talk) 07:37, 4 May 2019 (UTC)
  • I oppose this proposal I oppose this proposal. If it is bare soil then tag it natural=soil or landcover=soil ??? Don't use barren as that does not indicate what is there. --Warin61 (talk) 05:13, 1 May 2019 (UTC)
Yes but this tags doesn't exist, so how to map? --LorenzoStucchi (talk) 07:37, 4 May 2019 (UTC )
Make a new tag value... just as you are trying for barren. Mush easier if it says what is there. Warin61 (talk) 04:46, 10 May 2019 (UTC)
  • I oppose this proposal I oppose this proposal. Barren land will change very quickly. Vegetation will find a way soon or it will be used for other purpose (farmland, plantation, buildings). So it doesn't make sense to map it. --TheBlackMan (talk) 14:00, 2 May 2019 (UTC)
Not in all the area this will change quickly, looks to the example in the description. --LorenzoStucchi (talk) 07:37, 4 May 2019 (UTC)
  • I oppose this proposal I oppose this proposal. i many countries like Sweden is it normal to have a foresty livecirkel that inkludes periodisk "barren". Trying to tag the Forest livecyjelstage vased by satelit imerage leads to wrong tagged areas - not only burren but also scrub. --Skinfaxi (talk) 13:14, 5 May 2019 (UTC)
The tag shouldn't be used in the area were is not needed like in the area were there is a process of cut of trees, like for Sweden but in other area were is needed. --LorenzoStucchi (talk) 08:05, 10 May 2019 (UTC)
Nothing in the tag to stop its use for an area in the process of cut trees. It is not a matter of the intention, but what is documented in the tag. If the documentation of the tag allows its used then it will get used that way. Warin61 (talk) 21:59, 10 May 2019 (UTC)
  • I approve this proposal I approve this proposal. I understand the intent of this proposed feature. The comments from others that the barren land will be quickly overgrown is refuted by the two images 3 years apart - nothing has grown. Note that Buildings under construction and other construction projects are mapped, duration for the construction phase can be shorter that one year. The fact that something could grow there one day should not be a reason to not map it currently. I also think that Sand is not Soil, therefore natural=sand is not applicable. As an alternative you could consider other tags that indicate that an area was deforested and is now bare soil, e.g. use the forest tag combined with deforested=yes. For this I suggest to keep start dated and image source on the area. --ConsEbt (talk) 02:51, 6 May 2019 (UTC)
    • "is refuted by the two images 3 years apart - nothing has grown" - I am not convinced by this aerial image. I am not sure that it is not arable land (shape is not unusual for field cultivated without use of heavy machinery) or mud or area with small plants or wetland or area where vegetation was cut down and harvested multiple times. I would be convinced by citation to a credible source (confirming that normal soil in wet climate may remain without vegetation for years), possibly also by depiction how it looks on the ground. Mateusz Konieczny (talk) 11:32, 6 May 2019 (UTC)
  • I abstain from voting but have comments I have comments but abstain from voting on this proposal. I find most of the arguments against grotesque (e.g. wait until the vegetation grows back to describe what will happen in 3 years). but the proposal is poorly documented. it should have shown why generic landcovers allow incremental data improvement. this avoids the current situation where many contributors put the most likely value when they do not have certainty about the right value. and this (precise but inaccurate values) is much worse than imprecise but exact values (as is done with many key=yes) --Marc marc (talk) 16:02, 8 May 2019 (UTC)