Hacker News new | past | comments | ask | show | jobs | submit login
Declassified U-2 spy plane photos are a boon for aerial archaeology (sciencemag.org)
230 points by pseudolus 15 days ago | hide | past | web | favorite | 43 comments

Desert kites: they're mentioned in passing in the article and I was curious so I looked them up. Apparently they're 5000 year old traps for gazelles that can be kilometers long: https://en.wikipedia.org/wiki/Desert_kite

Thanks! Also new for me.


"Interestingly, the walls of the kites are not high enough to actually block the animals. Rather, they just seem to channel herds in the right direction. Modern wildlife managers in the same region have used a similar approach by laying pipes on the ground to direct gazelles into a corral, Avner reports."

Soft control, local minima, convenience, creator intent, fatal consequences.

Or in other words, surveillance capitalism social media.

A similar concept would be used in Canada and the US by the Native Americans, although they liked to make the end of the run a cliff so the people didn't have to do the killing by hand. Such constructions were called Buffalo Jumps.

A famous one in Alberta is the Head Smashed In Buffalo Jump: https://headsmashedin.ca/

I'm not sure "constructions" are the correct word. Buffalo jumps were generally natural cliffs and the heard was chased over them. The animals were corralled towards the jump by people on foot and by lighting fires, one of the few things that would make buffalo run rather than stand their ground. (Remember, no horses at this time)

They also included a runup where stone walls were built to corral the buffalo into a narrow point so they all ran off the cliff in roughly the same place. Without the walls the herd would turn unpredictably and avoid the cliff face.

Funny how they were named by "pilots who first saw them from the air in the 1920s" and not by the people who built them.

Since they were abandoned about 4000 years ago, I'd imagine that it's quite hard to find any records of their original name...

Presumably the people who built them didn't call them "desert kites." I think it's pretty clear the author was referring to that name when pilots in the 1920s were said to have named them.

Writing and history were pretty new at the time too so we probably don't even have a record of what they were called by the people who used them if they ever wrote it down to begin with.

Hammer and colleague Jason Ur, an anthropologist at Harvard University, created a systematized index of several thousand U-2 photos taken on 11 reconnaissance missions throughout the Middle East

Fantastic bit of nominative determinism:

Jason Ur is Professor of Anthropology in the Department of Anthropology at Harvard University, and director of its Center for Geographic Analysis. He specializes in early urbanism

As well he might:


I wonder what his favourite game might be. One I still have an unfinished implementation in C lying around.

The U-2 is the only U.S. surveillance aircraft still flying a "wet film" camera called the Optical Bar Camera, or OBC. Ironically, this relatively old technology keeps the U-2 mission from being completely transitioned to other aircraft like Global Hawk, due to the film imagery being easily releasable and able to be declassified.

Around 2007, a PC-104 embedded computer running Linux was integrated into the OBC camera to allow correlation of flight data with the imagery, as well as supply an OBC camera status page on the cockpit multi-function display.

Why is film easier to declassify and release than digital imagery?

Film's properties are known and understood and there aren't any "super films" that are better or more capable than another country's. Lenses are also understood.

With electro-optical (EO) sensors, great care must be taken to reduce the quality of the final product when it is publicly released so that adversaries do not gain a complete understanding of the what the sensors are capable of.

Film creates "better" images, but modern EO sensors are more capable in certain circumstances.

There is all kinds of computational and electronic trickery one can do to obtain images that may be impossible to capture on film that you want to keep secret, like fusing short wavelength IR with visible light or using it to discipline visible light to correct or reduce atmospheric distortion. Other EO technologies can determine what an object is made of from great distances.

Technologies like that you want to keep secret.

In a hypothetical Cuban Missile Crisis set in 2019, US analysts would have visible, near- and short wavelength-infrared, thermal, and pan-chromatic imagery to look at, but the 2019 version of Adlai Stevenson would still only show the visible images at the UN.

There were superfilms, film stock sensitive in the infrared/UV spectrum and filters to optimize such film. There was also film sensitive only to very specific colors. Releasing images from these, at any resolution, would indeed give away much of the program's abilities/goals.

The ability to more easily declassify film stock is due less to the technology and more to the bureaucracy within intel communities. The film stock is owned by a single agency and so the declassification authority is relatively straightforwards. Digital imagery is shared instantly with a host of different agencies, many of whom still do not talk to each other regularly, and is stored in countless archives. Declassifying a digital file is therefore an administrative burden in comparison to a roll of film kept by a specific agency.

These were all known since 1920s-1960s and very little progress has been made since. Everything you named is bog standard film photography practices.

Yes but the use of a particular technique in a particular location/time would divulge the specific collection goals of an operation, something that often remains classified long after the operation itself has been acknowledged. So while the existence of UV film is no secret, knowledge that it was being employed over a specific site at a specific time can be.

Less opportunity for sigint types to deduce information about the image source from the images and less opportunity for reversing whatever algorithm/processes were used to reduce quality (in order to make a "releasable" copy) in order to re-create that information.

I probably shouldn't have said it's "easier" to declassify. It's more of simply being a matter of deciding that the NIIRS (basically resolution) rating of the imagery from the OBC is low enough relative to other sensors and collection platforms that it won't provide much of an advantage to an adversary.

The embedded Linux system that was added to the OBC also provided a continuously variable Velocity over Height (V/H) control that allowed the OBC to collect imagery at lower altitudes than what it was originally designed for. This improves the quality of the imagery but at a lower area coverage rate as a tradeoff.

Film from the OBC is digitzed, ortho-rectified, and exploited as soft-copy. Back in the day it was done on mechanical light tables.

A good video that shows some of the film and other details is located here: https://www.youtube.com/watch?v=uemrNDEWgzA

For those interested in aerial archeology I also heartily recommend what Roger Agache did in (mostly) Northern and Central France starting with the 1960s. His aerial photos helped discover hundreds of locations for Roman-era villas and castra among other very interesting things. The website documenting his work can be found here [1], and for example aerial photos of Gallo-Romanic structures can be found at this link [2] (the website looks like it hasn't been updated in quite some time but it's still functional)

[1] http://www.archeologie-aerienne.culture.gouv.fr/en/

[2] http://www.archeologie-aerienne.culture.gouv.fr/en/

If this triggered your curiosity, have fun with http://www.apaame.org/

> The Aerial Photographic Archive for Archaeology in the Middle East was established in 1978 by Professor David Kennedy under the patronage of Prince Hassan bin Talal of Jordan. The archive now consists of over 91,000 photographs and several hundred maps of a dozen countries. The vast majority of this material can be viewed online in our digital archive at Flickr.

>In recent years, the Islamic State group gutted archaeologically important Iraqi sites in Mosul and Raqqa and Syrian sites in Aleppo and Palmyra.

Not to forget Buddha of Bamiyan, Afghanistan.

It is disheartening to think that, these sites & artefacts which were preserved for thousands of years are destroyed during our lifetime.

Perhaps because of its familiarity we value it higher, but powerful people in several parts of the world have been systematically destroying historical objects of value to a culture or civilisation to push forward their ideology.

Edit : Edited to be succinct.

Not specific to time, place, or ideology.

Example, Qin Dynasty China, Burning of books and burying of scholars (213 BCE):


Or the destruction of the Mayan codices by Bishop Diego de Landa in July 1562:


Not to forget, even if it does not fit that little theory, Ur Ziggurat and the many other important loses in Irak (see for instance http://www.uruknet.info/?p=58169).

Of course in that case, the ideology that prevailed was "one can't make an omelet without breaking some eggs", which is more palatable to us that Isis ideology of course, but in the long term that makes little difference I'm afraid.

> Not to forget Buddha of Bamiyan, Afghanistan.

That was the taliban.

Yes it was, I added it as example of sites of historical significance destroyed in our life time.

> Perhaps because of its familiarity we value it higher, but powerful people in several parts of the world have been systematically destroying historical objects of value to a culture or civilisation to push forward their ideology.

This is a particular feature of utopian ideology; marxists, islamists, fascists, etc. seek to carve utopia from reality, and the artifacts which prove that there's something else are just in the way. A look at what's preserved in Taiwan should make you weep to know what was lost just a little bit to the west.

It seems far more generalised than that:


I guess a less violent form of it practised by every country, even the democratic ones is falsifying history text books from grade school.

Yeah, Canadian grade school history textbooks are a joke. Be a good parent, don't let your kids get all of their historical education from your school district's standard textbook. :- )

Specifics might aid your argument.

In my experience, there was no mention of the broader historical origins of Canada, the focus was generally on a) stories framed to make French Canadians look interesting and blameless (especially long sections on coureurs de bois), and b) stories of amicable arrangements with native Americans, conveniently away from the various times the public's government has broken promises to various communities in Canada.

I'm an adult, why would I have the worst history books I've ever read, which were on loan from the school when I was in school, in my personal library to share specifics from?

Thanks for the first.

I'd not made any demands to the latter point, though there are times when keeping a spectacularly bad example around for debate, discussion, or reference can prove handy.

I'd put specific store in rough order "steelman" or best-case defenses or arguments for the indefensible, officially sanctioned references (as with textbooks), or with particularly poular bad examples, even if not particularly cogent.

Knowing your enemy, testing your own beliefs and biases, and walking into battle fully armed, are all benefits.

You don't need to find endless such examples (see also: Gish Gallop), but a carefully selected few can be exceedingly useful.

This applies to other areas as well, tech included.

> Knowing your enemy, testing your own beliefs and biases, and walking into battle fully armed, are all benefits.

Sure, to be clear, I grew up with two older brothers and a younger sister, in three different major cities, in three different provinces in Canada, and the history curriculum has largely lacked much particular detail. My history curricula, and that of my brothers, included little or no international history, ancient history, or national history. The main topics of every history textbook (the only source for each curriculum, in my experience) I've seen in Canada (including my brothers', for years I didn't attend in a given school district) have been an obscure subset of clean indigenous stories, and a handful of stories about early Qu├ębec.

I'm not saying these history textbooks are especially bad among government school history textbooks, but that they are bad in a general sense, and fail to give much perspective on the origin of the tapestry of nations in Canada, or the story of our legal and governmental traditions.

A better job could be done with an in-depth reading of a mature historical author's work, the kind of thing you would read if you had a personal interest in understanding the history of something.

I'm working with some researchers who are using similar recently-declassified spy plane imagery from the '60s for an analysis of glacier size + movement changes over time, which is another cool use case.

Somehow I always thought the U-2 is like a gliding plane, quiet and slow. But then I saw one at an airshow last year and I was pretty impressed how loud and powerful the plane sounded. It also climbed extremely steeply which kind of makes sense considering that it has get up to 30 km altitude within a reasonable time.

Recommend the book "The Past from Above" by George Gerster for those who enjoy 'aerial archeology'. 500+ photos taken from lower altitudes back in the 1950s and 1960s. (Can be had in paper, if you can't find the coffee-table-size hardcover.)

Review: http://www.historyinreview.org/ggerster.html

ISBN-Book sources: https://en.wikipedia.org/wiki/Special:BookSources/0-89236-81...

Repost are usual https://news.ycombinator.com/newsfaq.html and sometimes the same link is ignored one time and gets traction other time. It's matter of the day of the week, the hour of the submission, the other post that are trending, the number of users reading the newest page, the phase of the moon, and many other factors. Just assume that it's a mater of luck.

Linking the old post is useful when it has some interesting comments (I sometimes quote partially the most interesting comment.) But in this case the older post got only 2 points and no comments.

I like this: "It's matter of the day of the week, the hour of the submission, the other posts that are trending, the number of users reading the newest page, the phase of the moon, and many other factors."

What I want is some of those lenses and put them on my Canon. Looks like "pretty good" pieces of glass.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact