Tree cover loss – 2001-2020 | Hacker News Reader

Tree cover loss – 2001-2020

globalforestwatch.org

126 pointsitstaken31 comments

Cool to see this on HN - this is a product from the World Resources Institute[0] (my employer). I can potentially relay some knowledge from that team or try to direct someone here for questions and comments. The important thing to know for these maps is that Tree Cover Loss is not directly "deforestation". On a pixel by pixel basis (each covering ~30m x 30m) tree loss can occur without the major ecosystem destruction of deforestation. This is still measureable. That is why the visualization of tree cover loss is parameterized by percentage for greater inspection and nuance.

I will take this opportunity to point out the joy and satisfaction that comes along with making these kind of data for impact products. We get to work on globally critical problems, making real measured impact, solving interesting technical problems, and developing features and tools for real and appreciative users across the world. We have dedicated and talented staff (product and engineering teams, in addition to tons of researchers and engagement staff) who are passionate about their work and our collective mission. It's a great environment to work in each and every day. There are always challenges - project/institutional revenue is driven mainly by grants, our salary can not compete with the cash+equity offers of big SV tech companies and VC backed startups. But when I wake up in the morning and come to work (we are a very distributed/remote organization) I know that the time and effort I spend is directed towards a global good and I can easily say the same for almost all of my colleagues. That is invaluable.

If this kind of work interests you feel free to reach out to me (email in profile). We frequently have job openings[1] for PMs, SWEs, and many technical roles.

Edit: Senior Software Engineer position currently open (and would be working on Global Forest Watch) [2].

[0] https://wri.org

[1] https://jobs.jobvite.com/wri

[2] https://jobs.jobvite.com/wri/job/oeglifwR

Thank you for your work at Data Lab! I've been following WRI and GFW for a while and it's important stuff.

Also working in a global public-good project [0] I strongly resemble your challenges - uncertain funding, difficulty competing for top tech talent, and relatively small teams for huge projects. But so much more importantly the amazing benefits of working on something meaningful resonate with me. The outsized positive impact each of my team members can make on the world is a pleasure to be a part of every day, despite the challenges.

Keep up the good work!

[0] https://dhis2.org

I immediately recognized the title of the HN post, and it always makes my day when I see GFW in the wild.

Thanks so much for your comment, as someone whose name appears in the list of contributors [0] and was in the presentation of the project when it first launched 8 years ago I feel very grateful being part of it and couldn't have expressed the work better.

[0] https://github.com/Vizzuality/gfw

I’ve lived in newer suburban neighborhoods where some of the homeowners would cut their developer-planted trees down to branchless stubs, just as the trees were about to grow. I wonder what would explain the aversion to a tree. I suppose they don’t want to clean leaves during fall. The beauty of nature be damned, if even one is aware of it in this instance.

I’ve also lived in the city of San Francisco. The Department of Public Works has a tree database of all the tagged city trees. Now and then there would be a 311 call about a tree getting destroyed. If you try to plant a tree seedling it too would get pulled. Some of the reasons are that trees are equated with gentrification and raising the values of that block, thus raised rents. The beauty of nature be damned, because blight is much better.

Have you ever driven through neighborhoods with trees and ones without and think of the difference?

Trees - Joyce Kilmer

I think that I shall never see A poem lovely as a tree.

A tree whose hungry mouth is prest Against the sweet earth's flowing breast;

A tree that looks at God all day, And lifts her leafy arms to pray;

A tree that may in summer wear A nest of robins in her hair;

Upon whose bosom snow has lain; Who intimately lives with rain.

Poems are made by fools like me, But only God can make a tree.

in LA, it's that time of the year where they butcher all the street trees for no discernable reason (the stated reasons like safety make no real sense). it's infuriating. we need more cover, not less, in this drying desert area.

Those suburbanites were probably doing some variation of coppicing their trees, to keep them stunted and bushy for aesthetics.

I lived in San Francisco, too, and helped to plant dozens of trees with FUF (https://www.fuf.net/). You can't just randomly plant a tree along a sidewalk; some trees do badly in cities, attract pests, or drop inedible fruit that attracts vermin.

Pollarding (not the same as topping) is very common in cities to prevent trees becoming monsters that block out all light (that 40m oak might look nice from far off, but nothing else grows under that canopy, and someone might want to see sky from their garden), can't be maintained easily and eventually become safety issues with huge windage and heavy, old, brittle branches that can break off and fall (or the whole thing goes down in a storm and takes a house with it).

A pollarded tree looks stark and misshapen when it's done, but the foliage will return bushier and denser in fairly short order.

Pollarded trees can also live longer because not only do they not blow over so often, they stay in a juvenile state of young growth longer.

Some of the reasons are that trees are equated with gentrification and raising the values of that block, thus raised rents.

That's abominable. Does that really happen?

That pink/blue distinction is really hard to visually nail down. Is the Southeast US gaining or losing coverage?

Click a place name > "analyze" pop-up button > bottom analysis tab.

I know it's not really difficult, but just reading it spelled out like that makes it sound very non-intuitively difficult. Tech writing is hard!

i zoomed into some land that i'm intimately familiar with in mississippi. it got the tree loss right (when some logging took place), but doesn't show the later reforesting.

It has a tooltip on the loss section that mentions that the young trees need to get big enough to 'achieve canopy closure'

Looking in the area of the map East Texas and eastward toward the east coast, there is a definite blue tinge showing indicating gains in that area. Using the slider to narrow the range makes it even more clear.

I wonder if this is gains from natural regrowth or reforestation efforts like you describe. I was pleasantly surprised to see any positive direction on the map as I have a seeminly pessimistic view on the situation.

It helps if you remove the Green layer and zoom in (otherwise it just looks purple).

You can decrease the opacity of the purple/loss. It seems like a lot of the loss areas is the same areas where there's gain. It would be nice if there was some estimate of "unreplaced loss" or "predicted net loss"; the purple on top of the blue is visually difficult to discriminate.

There's no way to compare them because the data sets are for different time periods. The losses are for 2001-2020 but the gains are only for 2001-2012.

I don't know about specific regions, but forests worldwide are doing fine. In 1990, the world had 4.13 billion hectares of forest. In 2017 that number was 4 billion hectares.[1] That's not bad considering world population increased 40% in that time.

1. https://ourworldindata.org/forest-area#primary-vs-planted-fo...

I'd be careful drawing too many conclusions from that dataset. Many if not most forests are managed and logged; not all of that logging is necessarily on the public books nor does it imply complete deforestation; Legal logging typically requires a certain amount of regrowth to satisfy regulations (whether national or to satisfy carbon offsets). However, a sapling is not equivalent to a centuries old giant. Some afforestation targets permit a very, very low tree cover to be considered wooded. And if trees are being replanted; how is biodiversity doing? People have a tendency to pick cheap solutions (which is perfectly reasonable!), but that also means we should be careful we don't misinterpret simplistic coverage metrics as saying more than they do.

I looked (quickly, and therefore poorly, I'm sure), but I couldn't see anything in the ourworldindata sources that tries to distinguish the quality of forest left. I'm likely missing something, admittedly; I'll keep looking, if I find a link to their methodology I'll post an update.

They seem to be gather data from the UN FAO forestry project, and also other sources. I still haven't found any in those sources that account for the size of the trees; but ourworldindata does indirectly include some evidence that size may have reduced - the UN FAO source estimates forest coverage has dropped by almost exactly a third; but the tree number (where tree is defined as being 10cm in diamater at breast height) has fallen by almost half. However, since the sources use wildly different methodologies, I'm sure there might be other causes for the discrepancy, but in any case - it still highlights that interpreting these numbers isn't entirely trivial.

(Incidently, the same caveats likely hold for the globalforestwatch.org site; there too it's not obvious what they're actually measuring, nor how well they can do that, nor how close whatever metric they've got aligns with whatever you actually care about.)

I agree context and intricacies are everything. On your point about GFW, I know the team is very conscious and attentive to communicating what the data means.

This link might be helpful: https://www.globalforestwatch.org/help/map/faqs/

Thanks for the link; that's helpful. It also confirms that the worry that indeed it's quite tricky to interpret this data; even with the best of intentions errors are quite plausible.

It's still interesting, it just means we should see these efforts as narrow, potentially distorted perspectives on the ground truth; and therefore we need to remain open to the fact that other reasonable perspectives might show different trends.

I don't have the ability to give a highly researched and informed response to this currently, but I disagree with the claim that 'forests worldwide are doing fine'. I don't think you can make a well-educated argument from looking at global aggregated numbers of land classification for a ~25 year time period. Biodiversity, ecosystem connectivity, indigenous rights, agricultural competition, surface water quality, and legal regulations are all important framings to consider as well. Also, while 1990 is a while ago in human terms, it is hyper recent if you think about the destruction of forests over the past 12-thousand years of human-induced landscape disturbance.

I agree that limiting forest loss has been one of the more prominent victories in the environmental space, along with halting ozone depletion. Happy to pat some colleagues on the back for that success...

Dropping this resource from WRI: https://research.wri.org/gfr/global-forest-review

Some property I bought after it was clear cut around 2003 shows up on this map quite clearly ... including the increase in cover in the time since. So looks accurate for my single data point.

Part of my job involves satellite imagery (when I can get it), and I have occasion to compare the current year with imagery from, say, 2018 or so. It's a little depressing at times.

Will there be later years of data incorporated? The tree cover gain appears to only go to 2012 whereas the loss goes to 2020 which would underrepresent any gains.

So just to add some extra regional colour to this, it's not all down to deforestation for the purposes of farming or logging[0]. There's some pretty gruesome losses in the South Bohemia region of Czech Republic (around Telč and Dačice) which have an interesting backstory. Trying to recall an information board I read, so it may be a little bit spotty. But there was a big movement about planting spruce forests last century, which has now turned out to be a fairly vulnerable tree to a particular bark beetle (referred to as "kurovec" iirc) and ultimately wasn't very suitable in many of the areas it was planted anyway as tended to die and rot in the warmer climate of South Bohemia. This means that vast swathes of forest that previously appeared to be ok are now under threat and basically need to be felled.

So what you see when you go through there are pretty substantial areas of felled trees. We went a few years in a row to somewhere near a little village called "Zvůle") and the difference was shocking for the years 2019 and 2020. If you look both sides of the road here you get an idea of what I mean - https://www.google.com/maps/@49.0877897,15.2336907,3a,90y,61... Notice that basically all of the remaining trees have been marked, so they'll have been felled too.

[0] - I'm sure in this case they put the wood to good use, but I think they would have preferred to keep the forest as-is

edit: found the area I'm talking about on the map: https://www.globalforestwatch.org/map/?map=eyJjZW50ZXIiOnsib...

I imagine much of the green remaining there is just waiting for someone to come cut it down.

How is this reconciled with the NASA 2016 report that the "Change in Leaf Area" was largely positive from 1982-2015?

https://climate.nasa.gov/news/2436/co2-is-making-earth-green...

I have have big doubts on the methodology used to generate this maps. The map seems completely wrong for France, there is a big red mark on the biggest forest of France (Landes) but I can assure you that the forest is still there and will be there for a long time.

I guess I'm surprised that Finland & Sweden lost so much tree cover. Is that due to Ikea?

The vast majority of Swedish and Finish forests are managed for logging. All of the tree cover loss on this map in my area(except for some very recent that has just been logged/replanted) is today young forest which will be logged again in a few decades...

If you zoom in you see a patchwork of blue and pink, so it seems like just cycling through different areas for cutting and regrowth.

I looked at the data from some places I knew very well for over 3 decades (like around where I grew up and are still visiting parents regularly) and it looks noisy and mostly incorrect, nothing like real tree cover change.

Reply on news.ycombinator.com