In hindsight, it probably wasn't the best idea to try and do an entire continent in a single patch. I thought it wouldn't take too long to upgrade
the Antarctica WIP I mentioned last time, however access to better sources than were available to me two and a half years ago revealed that said WIP, while accurate at broader scales, was actually fairly inaccurate at smaller scales, and dead wrong in a few places where the 8K-BAM is wrong as well. Also, RL got in the way quite a lot over the last month, which slowed down progress, but that can't really be helped. Honestly, the better move would've been to cut Antarctica into two or three wedges along lines of longitude and post each of those as its own patch. However, by the time I realised that that's what I should've done, I was already too far in, so I just decided to push on to the end.
A few notes on the patch.
My primary sources are the datasets made available by the British Antarctic Survey
here, covering coastlines (showing true coastlines, ice shelf fronts and grounding lines), rock outcrop and lakes (because, surprisingly, Antarctica does have a few lakes, though more on them later). The datasets were ported into QGIS and reprojected to fit the R-QBAM, then used as a reference base as I extensively re-worked the old WIP in paint.net.
The main secondary sources were these two series of maps made available by the
Australian Antarctic Data Centre here and
here.
The first series of 22 maps covers the Australian Antarctic territory, French Adelie Land (a wedge enclaved within Australian Antarctica) and bits of the Norwegian and New Zealand claims.
The second series of 21 maps was produced as a collaborative effort by the Antarctic mapping agencies of multiple countries over time. It can be a bit inconsistent in places but on the other hand covers the whole continent.
Both map series have their flaws. The first one was produced as recently as December last year, and is incredibly detailed, but is only regional in scope. The second series, while overall useful, has some more annoying problems. As mentioned there are some notable inconsistencies where two maps produced by two different mapping agencies overlap, i
n Map 15 the map key covers up most of the south-eastern coastline of the Ross Ice Shelf,
while map 9 for some reason isn't accessible through the AADC, and I can't for the life of me find an alternate copy elsewhere on the internet, so is effectively missing.
Because these maps have their flaws, I was largely basing my work on the BAS data, which is internally consistent, shows everything I wanted to show, can be easily manipulated in QGIS and covers the entire continent. Also, more often than not, those maps are in a different projection or orientation to the R-QBAM's Robinson, and while the AADC does offer the option of downloading a georeferenced TIF file that would allow importing and reprojection in QGIS, this isn't available across all the maps. As an aside,
as several of the maps in the second series were produced by the BAS themselves as their contribution to the collaborative mapmaking effort, those maps align exactly with the BAS datasets as you'd expect.
However, as mentioned, there are a few places where the two datasets disagree, with the worst offender being the
Bear peninsula in West Antarctica. As you would expect from the name, when it was first sighted as an ice rise from the air it was assumed to be a peninsula.
Thus it is shown as a peninsula in map 11 from the continent-wide series, compiled by the Polar Geospatial Centre at the University of Minnesota, and
wikipedia follows convention by naming it a peninsula. However, the BAS data shows it as an island,
as does map 4 in the same map series, compiled by the BAS, while in
this picture also on wikipedia it really looks like an island if you focus on the grounding line. In this case I ended up siding with the BAS as they were the primary source, but I thought the discrepancy with other sources was notable enough that it needed to be mentioned.
I also had a bit of trouble with the Australian-Norwegian border (which was a sentence I never expected to have to type out). Some sources (
notably wikipedia) say that the border between the Norwegian and Australian Antarctic claims follows 44°38′ E,
however others claim that it follows the 45th meridian east, half a degree of longitude over. I ended up going with the 45th meridian,
as the Norwegian Polar Institute defines this meridian as the border,
while this map produced by the Australian government also shows the border following the 45th meridian. I think what happened is that some wikipedia editor mistook the edge of
Enderby Land as defined by historical convention with the official territorial claims. Someone read that the western edge of Enderby Land is at 44°38′ E, and that Enderby land was in the Australian claim, so assumed that this must be the border of the Australian claim, leading to confusion and inconsistency between articles.
On another note, contrary to what you might expect for a continent that is almost entirely covered by ice, there are a couple of lakes in Antarctica. Firstly, there are a couple of places where ice shelves abut the coast which are mostly ice free, such as the
Edisto Channel between the
Bunger Hills and the Highjump Archipelago. Some of these far inland such as
Beaver Lake are regarded as lakes, and I decided to show these cases as open water. Then there's
Radok Lake,
a true glacial lake nestled in one of the larger patches of open rock in the Prince Charles Mountains (notice Beaver Lake just to the north east in that link). It's frozen, but it's still very clearly a lake, earning itself two pixels on the map.
There are also about a dozen hyper-saline lakes in the
McMurdo Dry Valleys near Ross Island, by far the largest ice-free area on the continent. Many of these lakes have a higher salinity than the dead sea, and as such can remain liquid even when the temperatures drop well into the negative during Antarctic winters. Of these lakes, I judged that one of them,
Lake Vida, was just about big enough to show (the slightly smaller
Lake Vanda just about didn't make the cut). On a related point, I have good news for
@Rac98 - Antarctica has probably only a dozen or so '
rivers' (in reality little more than seasonal streams), most of which flow through the Dry Valleys into the aforementioned hypersaline lakes. It's by far the easiest continent to add rivers to.
Some final quick map notes concerning Antarctica before I move on. Before anyone says 'you accidentally coloured one of the islands off the Antarctic peninsula as ocean',
Deception Island genuinely looks like a horseshoe. In the same vein, the single pixel island to the east of the Balleny Islands isn't a misclicked pixel, it's actually
Scott Island (named for the ill-fated explorer) a small volcanic island in the middle of nowhere. Finally, there are some places on the edges of Ice shelves where major fissures penetrate quite far, marking the site where future icebergs will one day calve away. Said fissures are occasionally shown by a couple of pixels of open water near the coasts of the ice shelves, which I wanted to point out before anyone says that I missed a pixel.
It should also be mentioned that quite a lot of the nunataks in the southern Transantarctic Mountains aren't actually as big as they appear. This close to the bottom of the map however, projection distortion goes nuts, which would've been a fiendish challenge if I hadn't gotten the hang of QGIS over the last few months.
Oh, and one final bit of trivia, but I was amused and in no way surprised to learn that at least a few countries have found a loophole in the Antarctic treaty - existing territorial claims are frozen and new claims are banned,
but the treaty says nothing about maritime claims.
Long story short, Australia has apparently filed all the correct paperwork to get an EEZ off the coast of their Antarctic claim under the UN Convention on the Law of the Sea, sidestepping the Antarctic Treaty. Other states that claim chunks of Antarctica have pursued or expressed an interest in pursuing a similar approach,
which has led to some very messy maritime border claims and disputes.
This patch doesn't just add Antarctica. I also lightly revised several of the subantarctic islands, added a couple of remnant ice shelves north of Ellesmere island (more on that below), and added the first new bits of Australia for over a year, namely the
Cocos (Keeling) Islands and
Macquarie Island (administered by Tasmania).
I mentioned ice sheets before, but I'll quickly restate the problem. The conventional way of treating ice shelves by counting the grounding line as the coast works well for Antarctica, but falls flat in Arctic North America and Greenland, where quite a few fjord-constrained glaciers have extensive floating ice tongues that should count as ice shelves under the Antarctic rules. Problem is, nobody applies the Antarctic rules consistently worldwide, and treat the ice fronts of these glaciers as a proxy for coastline. I don't think I've ever seen a map that treats all the ice tongues of Greenland as ice shelves and extends the coast far inland along those fjords. A couple of months ago when I noticed the problem I decided to side-step the issue and put off considering the problem till my hand was forced by Antarctica. Having reached the frozen continent, I now had to make a decision.
I eventually settled on following convention and treating the two areas differently, largely because if I tried to keep things consistent then I would royally mess up the northern hemisphere and make it look wildly different from basically every other map of the area. The task of finding data and maps showing the grounding lines of all those glacier tongues, not just in Greenland, but on Baffin Island, Ellesmere Island and in Alaska too, may not even be possible. I tried looking, and for many such glaciers data is either untrustworthy, incomplete, out of date, nonexistent, locked behind a paywall or some combination of the above. Sticking with convention means I can avoid that hassle. It helps that there is some academic disagreement over whether these ice tongues count as true ice shelves or whether they should be classed as something else, which gives me the wiggle room I needed to make this call.
On the other hand, a couple of small ice shelves in the Canadian Arctic that everyone agrees are conventional ice shelves do deserve to be shown now I'm showing the Antarctic ones.
There appears to have been one single ice shelf hugging the northern coast of Ellesmere island at the end of the 19th century, as documented by early Arctic explorers, however that ice sheet had cracked into about six successor ice shelves by the early 20th century. Things then remained pretty stable till the 90's, when they began to fall prey to a warming global climate. Some detached from the coast becoming icebergs, others essentially disintegrated and collapsed in place. Two ice shelves survive,
the Ward Hunt ice shelf in the far north (the long thin one on at the very northern tip of Ellesmere Island on the map) and the
Milne ice shelf nestled in the peninsula of the same name, however both have seen substantial losses over recent decades.
I tried to check much of the rest of the Arctic to see if there were any other ice shelves to add, but to be honest, there really aren't that many ice shelves in the Arctic at all. I tried really hard to get my hands on the full text of
these two papers, up to and including getting a few old IRL friends in the sciences to check if they had institutional access without success. Nevertheless the abstracts alone confirm that there are apparently no ice shelves in Iceland, Svalbard and Novaya Zemlya, while the small ice shelves of the Franz Joseph Archipelago are apparently too small to worry about. Outside northern Canada, and excluding floating glacier tongues, the only ice shelf of any size in the northern hemisphere was the Matusevich ice shelf in the Severnaya Zemlya archipelago, which collapsed back in 2012, so I don't need to worry about showing it here, and is outside the scope of the current map anyway.
Final point. I wasn't sure what the best way was to show the Antarctic Treaty system and the thoroughly opaque status of territorial claims on the continent, so I made quite a few variant maps of Antarctica then compiled them together to present some options, in addition to a few more useful variants, presented at the bottom of todays patch.
Starting from the top, the first map shows the base Antarctic geography with no borders at all, while the second map is basically the same as the first, but with the colour of the nunataks exaggerated to make them more visible. Map 3 shows only borders over land, map 4 shows borders over land and ice shelves (because in many ways, the ice shelves of Antarctica kinda are an extension of the land), and map 5 shows the full claimed sectors up to the 60th meridian south.
I ran into a bit of a problem trying to show who claims what. The initial idea to colour ice free land by the country that claims it falls flat for two reasons. Firstly, the Antarctic peninsula is disputed between 3 nations, and secondly the small French claim doesn't actually have any significant rock outcrop so has no pixel of land to recolour. In the end, I went with creating a new layer showing the extent of various claims with a block of colour, hashed in the case of disputed territories, then pulled the opacity down and overlaid it on the main map. Thus map 6 shows the full sector claims appropriately shaded, map 7 shows only the land and ice shelves shaded, while map 8 shows only the land borders and the land shaded. Map 9 (my personal favourite), shows borders and shading over land and sea ice but nothing more
Finally, map 10 is one for the completionists, updating the ice coverage layer to add the ice sheets and ice shelves of Antarctica.
After the long slog through Antarctica, I'm hoping progress will speed up over the next few months. Next up on the schedule is Australia, in four chunks (Western Australia, then central Australia, followed by the southeast and ending with Queensland and the Coral Sea), followed by New Zealand and a bit more of the Pacific before changing tack and moving west to tackle South East Asia and Indonesia.
There have been a few comments since my last post, but this write-up took ages to put together, so I'll make my comments later. In the meantime, on with ...
Patch 122 - Antarctica;
- Added Antarctica (massive job, but it's done) featuring some alternate suggestions for showing the opaque nature of territorial claims under the Antarctic Treaty.
- Tweaked other Subantarctic islands
- Added the Ward Hunt and Milne Ice shelves north of Ellesmere Island
- Added the Cocos (Keeling) Islands (Australian external territory)
- Added Macquarie Island (to Tasmania)
View attachment 908693
Antarctica variants, in two files;
View attachment 908694
View attachment 908695