Ukraine Update: Back of the Envelope Calculations, Digital Evidence, Maps, Scenarios

Ukraine Update: Back of the Envelope Calculations, Digital Evidence, Maps, Scenarios

Facebook
Twitter
LinkedIn

[ad_1]

By Lambert Strether of Corrente.

Having become “kinetic,” the situation in Ukraine has also become overly dynamic; I hope readers will supply up-to-date and vetted information as events progress, because I feel that whatever detail I go into here will rapidly become obsolete (or turn out to have been based on bad data; see below)[1]. Meanwhile, here are a few back of the envelope calculations to help scope the matter.

First, kilometers kilometers. Search tells me that Kyiv, Ukraine is 327km from Gomel, Belarus, near the border. Converting to the imperial system of measurement — ironic name, that — for my own comprehension, that’s 203.188 miles. Assuming conservatively that a tank can move at 25 miles per hourthat’s about 8 hours for Russian tanks to reach Ukraine’s seat of government, assuming little resistance. In fact, the trip seems to have taken about 24 hours. From the Financial Times:

Russia’s onslaught breached the suburbs of Kyiv on Friday as armoured columns battled towards Ukraine’s two biggest cities in Europe’s largest military offensive since the second world war.

Kyiv residents woke to loud explosions shortly after 4am and reports of Russian armoured vehicles advancing into the northern Kyiv district of Obolon, bringing street fighting to the 3mn-strong capital.

Western officials said that stronger than expected Ukrainian resistance on Thursday slowed Russia’s two key objectives of encircling Kyiv and capturing Kharkiv in the north-east, Ukraine’s second-biggest city.

Ukraine’s president Volodymyr Zelensky delivered an early morning address to his country as his army warned Russian troops were “moving at speed” towards central Kyiv followed by a column of military trucks.

(I’m assuing the troops heading toward Kyiv are coming South from Belarus, not West from the Donbas, or up from Crimea. I’m also assuming there won’t be a heroic last stand in Kyiv.)

Second, troops involved. Ukraine has an army nominally sized at 209,000. Russia’s troop count for the operation is opaque, but calculating from BBC figuresa little over 60,000 are involved. (60% of Russia’s troops are within 50km of the border, and half of those are “tactically deployed.”)

Third, casualties.From WikiPedia’s updating page:

Wars are bad, even small wars. At least that is the principle I think leadership should begin from. Of course, a small war can have large geopolitical effects, as this war doubtless will; the lightest touch can collapse a house of cards. And if you have a bet riding on that house of cards, a light touch can have enormous leverage. Nevertheless, as measured by the standards of our own wars in Iraq, Afghanistan, and Libya, or our sponsored wars in Yemen or Syria, this is a mercifully small war, both in troops and casualties. And if a war ends when a combatant’s capital is captured, this war will end soon (although I’m not sure it will, for reasons I will briefly explain below).

Back of the envelope calculations complete, I want to look at three topics. First, an old favorite: Digital evidence. Second, maps. Third, scenarios for what happens next.

Digital Evidence

First, my standard for visuals in war reporting is something like this (By Flag on the Reichstag, Fair use):

I don’t know if the Russians will be crass enough to raise a Russian over Ukraine’s Parliament, but if they did, that would be excellent proof they had really captured Kyiv, as opposed to, say, a Tik Tok video of the event.

Much of the digital evidence you see online is faked by amateurs. From Gizmodo:

Gizmodo has found at least ten viral photos and videos currently being spread on social media that are completely fake. In some cases, the videos and photos are years old. In others, the images are clearly not from Ukraine. There are even two examples of videos on Twitter today that are actually from war-themed video games, something Russian state media has previously tried to do on multiple occasions.

I don’t know why anybody would Tweet out a picture of an old war for the current war, but apparently people do. (I mean, one would expect official fakery to be more adept, although when we think back to Colin Powell’s vial of white powderperhaps not.)

Other digital evidence you see online can be staged:

At least we have the Reuters sigil, but from the commentary, these troops do not seem to have taken up actual combat positions.[2]

A second example:

The account (@brycewilsonAU) is real. Loading the photo into Google Image Search, I find the use in this tweet is the onlh example. So, interesting. However, I do think if a photographer had taken a wider angle shot of the Ukrainian National Guardsmen above, from farther away, the same theme would have emerged.

Then, quis custodiet ipsos custodes? Neiman Labs has a list of resources for following the warbut hilariously lists Bellingcat, the Atlantic Council’s cut-out to the Intertubeswhich is a lot like putting the American Council of Pork Producers in charge of censoring fact-checking information on swine fever.

Finally, some common sense suggestions:

Be careful out there.

Maps

Of the several maps available, I will select the Financial Times, the BBC, and the New York Times.

First, the Financial Times:

You will note the lack of dynamism (no arrows) and the sketchy provenance of the data. Even a lowly site like PayDay report, when it puts a dot on their strike map, tells you where they got the data!

Second, the BBC:

Here we have arrows. It would be even nicer to have roadsso we could know if the tanks were moving by road or overland. The provenance is sketchy in the extreme, even worse than the FT.

Third, the New York Times:

Here we have both explosions for ground fighting and arrows (though the two seem to have little relation to each other). And the captions are handy. But the provenance is worst of all. We’re basing Times reporting on information from one of the combatants? Really?

These are not really maps. They’re graphics. designed to show what a child of six knows: That the Russians are attacking from the North, the East, and the South. (As with so much journalism, the point of these maps is not to convey information, but authority, or at least authoritativeness.)

Wtih relief, we turn to digital projects. Tnis is interesting:

Perhaps some readers would like to play around with the app. Of course, you should put your trust in digital evidence from these feeds as much as you would put your trust in any feed where war and peace are at stake; ie, not at all . Fun, though!

And then there’s Google Maps. From Vice:

At 3:15 AM local time, Jeffrey Lewis, an open source intelligence (OSINT) expert and professor at Middlebury Institute, saw a traffic jam in Belgorod, Russia, using the traffic layer of Google Maps. “Someone’s on the move,” he tweeted…. Lewis said this wasn’t as simple as them loading up Google Maps and seeing a traffic jam, assuming it meant an invasion was occurring. They were also using optical and radar satellite imagery taken days before. On Tuesday, they had a radar picture taken by Capella Space, a company that provides satellite imagery, showing Russian armored and heavy vehicles lined up in columns ready to move.

“We all looked at the picture, and was like, oh shit, it’s coming,” he said. As they were monitoring this unit, they noticed the traffic jam. “So it’s the prior work of knowing that there’s a giant Russian armored unit sitting right there that allowed us to say, like, oh, I know what that traffic jam is, they’re getting on the road.”

Google Maps live traffic data works by incorporating location and speed information from Android phones. This led many commenters responding to his tweets to assume the Russian soldiers left their phones on—which would be horrendous cybersecurity hygiene—and Google was picking that up. But Lewis said that almost certainly was not the case, and the traffic data Google was showing was ordinary people who couldn’t get on the road because of Russian military movement.

“We have developed incredibly data rich definitions of what normal patterns of life look like,” Lewis said, “And any deviation is immediately caught.”

Interesting, but sadly not happening in real time. There is also this Google map:

Here, at least you get sourcing when you click on a dot — yay! — but the map authorship is incredibly dubious: “This map is put together by a group of friends seeking to map the conflict.”

For all I know, there’s a dynamic map with full provenance out there on the Intertubes, but if there is, I haven’t found it. Readers?

Scenarios

Finally, in no particular order, I present some threaded tweetstorms on future scenarios for Ukraine. I take no view on any of them, but you may find them interesting to read and assess:

Scenarios (1):

Scenarios (2):

Scenarios (3):

Scenarios (4):

I won’t venture to predict what the new “security model” for Europe will be, now that NATO”s house of cards has been shown to be wobbly, to say the least. However, I’m long stupid generally, and long stupid here, at least for the actions of the United States, which I know something about. I think we will fund and support an insurgency within Ukraine, no doubt using Azov Battalion types plus mercs and dupes, in an attempt to bleed Russia, dreaming of a Color Revolution. I doubt very much that will work even in its own terms, because we have form (losing) and Russia has form (winning)[2]. And if that means hardening the border with the EU, then (from Russia’s perspective) so be it. Notice this policy has continuity with previous US policy.

Conclusion

To fair, those who meme aren’t entirely good-for-nothings:

Idea: Swap out Zelensky, swap in Juan Guaidó?

NOTES

[1] For example, as of this writing China’s Xi Urges Putin to Negotiate With Ukraine During Call. or Ukraine ready for talks with Russia on neutral status – official.

[2] Except in Afghanistan. We both lost there.

Print Friendly, PDF & Email



[ad_2]

Source link

More to explorer