39.91 N, 127.55 E: Hamhung, Haunted City

In 1997, Washington Post correspondent Keith Richburg was allowed into the city of Hamhung, just inland from North Korea’s east coast, to try to find the truth behind fragmentary rumors of a famine inside the world’s most isolated country. Although Hamhung is North Korea’s second-largest city and a key industrial center, it was an isolated place with few foreign visitors, little commerce with the outside world, and at a great distance from any international border. This is what Hamhung looks like today. But in 1997, the reporter found utter economic blight and devastation there: idled factories, dark streets, dying and displaced people:

HAMHUNG, North Korea — A visit to this remote and desolate city near North Korea’s eastern coast provides a rare glimpse of the country’s near-total economic collapse. The crisis is over food — or the lack of it — but the country’s problems run much deeper, to the core of a socialist system that simply has ceased to function.

You can start at Hamhung’s local hospital, a dilapidated, cavernous 1,000-bed facility without lights, where the stench of urine fills the dark corridors, and patients recovering from surgery writhe in pain on dirty sheets in unheated rooms. There are no antibiotics, no intravenous drips and no stretchers, so workers carry patients on their backs. There were only 250 patients during a recent visit; few sick people bother coming, since the hospital has no food and no medicine.

“We have a shortage of anesthesia, so the patients have to go through pain during surgery,” said Dr. Lee Huyn Myung, as he points to a man gripping his mattress after a colon operation. Most of the patients have rectal, stomach or liver problems, the result of slow starvation, he said. Almost all are malnourished.

From the hospital, travel across this city past gray cement buildings that look half-finished or simply abandoned, past lots strewed with broken-down Soviet-era trucks that cannot be started because there are no spare parts. Then drive down narrow, winding mud roads until you reach the Hamhung orphanage and talk to its director, Choi Kwang Oak.

The orphanage is divided into several small rooms, with playpens for the smallest infants. Almost all the children are malnourished, with browning hair, bald patches on their scalps and sores on their heads and faces. The most severely malnourished are listless and unresponsive.

There are 198 children under age 4 at the orphanage, and about 20 percent are expected to die because they arrived too late to be helped. About 70 percent of the children here were orphaned when their parents died of malnutrition or disease, Choi said. The other 30 percent simply were abandoned and left for dead by parents too poor and too hungry to feed them.

“Some parents just put them outside on the street and leave them to nature,” Choi said. “Sometimes people pick them up and bring them here.” And other times? “They just die.” The orphanage is surrounded by high hills covered with graves and stone markers. It is an old burial ground, she said. But there are also many new graves. [Washington Post, Oct. 19, 1997, emphasis mine]

Similarly, this 1999 Reuters report described widespread malnutrition that had stunted 50% of the surviving children. Visiting a local orphanage, a reporter described it as “surrounded on all three sides by hills covered in graves.” We will return to the matter of those graves in a moment.

These were not the only reports of Hamhung’s silent devastation to leak out to Earth. In 1998, one former engineering student told Kyodo News Service that more than 10% of the city’s population, including his mother, had starved to death, and that another 10% had fled the city to find food. Some would have fled by train. Those who could not afford tickets and obtain travel passes are unlikely to have survived the long walk in their weakened condition. The student added that “the food is going to the troops and that party cadres are hoarding up to a year’s grain for themselves and their families.” Andrew Natsios, a former AID worker, USAID Administrator, and author of “The Great North Koreans Famine” described Hamhung as “the city most devastated by the famine” and reported that hunger had triggered an unsuccessful mutiny against the regime. The journalist Jasper Becker bought back the life of a former Hamhung resident who walked to China after her husband perished in the famine. Chinese police were paying bounties for North Korean refugees, but Becker outbid them to save her from repatriation to a North Korean gulag.

Today, the world knows that between 1994 and 2000, anywhere from 600,000 to 2.5 million people died in what is what Natsios called The Great Famine, or as the official propaganda describes it, The March of Tribulation. North Korea watchers have debated the question of mortality in Hamhung because of how that toll could impact the overall death toll from the Great Famine. If in fact the death toll from the area around Hamhung was worse than what most refugees described — and most refugees came from the border areas further north — then this would suggest that the death tolls these refugees reported would still understate higher death tolls in places like Hamhung.

But is there any empirical evidence that would corroborate stories of such devastation? If those reports are true, we would expect to see numerous grave sites in the hills around the city. Koreans traditionally bury their dead under spherical mounds on high ground, and Korean cemeteries have a distinctive appearance in satellite imagery. This image of the National Cemetery in Seoul provides a baseline comparison for the size, shape, and general appearance of Korean graves, though one seldom sees North Korean graves laid out in such neat rows.

Less than a year ago, Google Earth published the first high-resolution images of Hamhung. In the hills around the city, one can see a distinctive rash of small mounds packed haphazardly into the steep slopes. What begins as a small, well-tended cemetery consumes an entire hill …

graves-hamhung-well-tended.jpg
[Click image for full size.]
and the hills around it, …

graves-hamhung-1500.jpg
and seemingly, every speck of unoccupied land that surrounds this city of 874,000 souls.

graves-hamhung-4000.jpg
graves-hamhung-21000.jpg
On some of the hills, the graves are crowded together so tightly that the density of death greatly exceeds the density of life in the city below.

graves-at-hamhung-2.jpg
There are other large, seemingly improvised cemeteries in North Korea. This one is just east of the capital, Pyongyang.

signs-graves-e-of-pyang-1.jpg
signs-graves-e-of-pyang-2.jpg
Hamhung, however, is the only city I know of where the combined surface area of the cemeteries rivals that of the city itself. Even in North Korea, Hamhung is the only place where I found it possible to look down on such an immense concentration of death. The fact that no other place even approaches the scale of what we see here suggests that we’re looking a horrible aberration. Every hill around the city is a burial site, though not always of equal size and density. On this rocky ridge west of the city, the forest is seeded with graves:

hamhung-forest-graves-and-ugf.jpg
The small dots in the forest throughout the image are graves. The placemarks in the lower left-hand corner of the image are higher density burial sites. Note the tunnel entrances in the upper right-hand corner of the image. The earthen berms in front of the entrances suggest that they’re probably for some kind of military use. Some kind of large facility is being built just to the east of this image, no doubt at great expense. I’ll venture no guess as to its purpose, except to infer for that it won’t be used for growing food.

It would be impossible to count the graves in these hills, but we can quickly and conservatively estimate that the number must be in the tens of thousands. The polygon in the image below is almost completely filled with graves, in an approximate density of 25 graves for every 400 square meters, for a density factor of 0.0625 graves per square meter. The polygon is 180 meters from north to south, and 980 meters — nearly a full kilometer long — from east to west.

hamhung-graves-polygon.jpg
That works out to more than 11,000 graves on just this one ridge. Starting from the same polygon, let’s rise to an altitude of 35,000 feet to put that polygon into the spatial context of the other burial sites in the hills around Hamhung. The arrows in the image below point only toward large, dense concentrations of graves.

hamhung-graves-polygon-with-placemarks.jpg
Even a cursory examination of these hills shows that the total number of graves probably exceeds 100,000. If other smaller and less dense burial areas are added in, it could plausibly exceed twice that. This estimate would obviously undercount any mass burials, such as those witnessed by Andrew Natsios from across the Chinese border, or the deaths of those who died by roadsides or in detention camps elsewhere. It would overcount mass mortality to the extent that some of these graves represent pre-existing or “natural” deaths.

Obviously, there is no way to determine the date or cause of death from a satellite photograph. At the same time, the devotion of so much space to the dead cannot be the natural state of things; after all, it isn’t the natural state of things anywhere else. We will never know for certain what killed these people as long as North Korea is ruled by such an obsessively secretive regime. All we know is that this city is surrounded by an unusually large number and concentration of poorly tended and haphazardly placed graves in a manner consistent with published reports filed from Hamhung during the famine. If these graves do in fact contain famine victims, and if our very rough estimate of 100,000 to 200,000 graves is accurate, it would generally corroborate the engineering student’s claim that 10% of the population of North Korea’s second-largest city starved to death during the Great Famine.

Stalin infamously observed that a million deaths is a statistic. That is so because of our own incomprehension at the very real significance of each spot on these pock-marked hills. Focus on just one of those graves long enough to imagine the child, the mother, the father, or the provider buried there, instead of seeing mere pixels in a ghastly panorama. It ought to be so even in a nation where for most, to be a pixel in the state’s panorama is the individual’s highest permissible aspiration.

It’s a sobering thing to realize, as a new administration asks itself how to approach this regime, that if these are indeed the victims of the Great Famine, none of them had to die. It wasn’t any disaster or crop failure that killed them. It was the ruthless decisions of one depraved man.

[You can see more Google Earth photo essays of North Korea here.]

Update, 8 October 2009: Recently, I learned that the noted human rights researcher David Hawk would be traveling to Seoul to interview North Korean refugees there.  I asked Mr. Hawk to seek out former residents of Hamhung to confirm that the objects I’d identified as graves really were graves.  One witness reported that he’d recently been to the location shown in the next three images below and reported that he saw only orchards, not graves.  On further examination, I suspect he may be right.  As a result, I am no longer confident that all of the images below show graves.  Given the challenges of revealing what North Korea doesn’t want us to discuss at all, I can only follow the best evidence available to me.  I will continue to seek out North Koreans to confirm what these images show and update this post as I learn more.

Update 2, 31 January 2014: I’ve meant to return to this post for some time, preferably after gathering some information from a defector from Hamhung. Unfortunately, not many people make it out of Hamhung, because of its distance from the border. Even so, I’ve reexamined the images carefully and I’ve changed my mind again. I don’t believe these objects are trees. The density, shapes, and shadows are all wrong. Furthermore, season-by-season, the objects don’t change color or density the way trees would; they change color like the grass. My best guess is that they’re graves. I’m not going to say that with confidence until I find witnesses to confirm it.

But there is this. You may want to expand this one to full size (just click).

Screen Shot 2014-01-31 at 6.06.57 PM

Those elongated, box-like objects resting on (or washing out of) the hillside are all between four and seven feet long. Draw your own conclusions.

17 comments

  1. Gray Hat says:

    Great work. This essay leaves me speechless.

  2. GI Korea says:

    I second that great work, I’m looking at Hamhung now on Google Earth and you are right, there is a disproportionate amount of graves there.

  3. KCJ says:

    This is only the respectfully buried dead. I would imagine that for every reverently buried North Korean there is another buried in a mass grave, cannibalized, left in the open, or unaccounted for.

    The DPRK has in the past transferred entire villages away from their ancestral homes in order to use graveyards as arable land for growing food. So those are missing as well. When you think about the traditional belief in han and the practice of filial piety and placating the spirits of the deceased, the level of misery in North Korea must be reaching heaven itself.

    [OFK: I suspect the same thing but don't have any direct evidence for it for this particular city. But there must have been many cases in which surviving family members were in too weakened a condition to climb hills, dig holes, and bury their dead. We also know that mortality was especially high in state institutions like orphanages, hospitals, and so-called 9/27 camps. Deaths under those circumstances would have been left for local governments to collect and bury, and local governments could have found that task too overwhelming to give every victim a proper burial. We have the account of survivor "Yomiko Chiba" to inform us how the state dealt with this in Sinuiju -- it tasked teachers and students to collect the dead from the streets at night for mass burial outside of town the next day. Many of those dead were people who wandered into the city from outside in search of food. The Daily NK corroborates the use of this method elsewhere:

    In the mid-1990s when the famine was at its height, the authorities put dead bodies on the streets for trucks to collect for mass burial.

    They called it as “sowing” at the time, because burying many bodies in the same place was like sowing seeds. That’s why the owner of this grave may be fortunate compared to those who died during the great famine. [Daily NK]

    In other words, those who died during the worst of the famine are less likely to have their own graves, and the absence of an anomalous number of burial mounds isn’t necessarily evidence that there were not mass deaths in a given city.]

  4. NKeconWatch says:

    It seems to me that (with the proper funding) it would be worthwhile to use satellite images of these locations from 1995-2003 to estimate the number of deaths from the famine (if they exist). Although not authoritative on its own, this test could help us validate or reject previously estimated statistics.

    Dozens of these locations are marked all over North Korea on my Google Earth project which can be downloaded here: http://www.nkeconwatch.com/north-korea-uncovered-google-earth/

  5. Irene says:

    Excellent post as heartbreaking as it is – there is no “duality” anymore with respect to the DPRK – I, for one refuse to harbor these “inconvenient truths” anymore…Joshua, your work in this area proves without a doubt, the true nature of the DPRK regime.

  6. jsternsp says:

    I have been thinking about this story for hours now. I don’t think the number of graves around this city is necessarily any indication of the number of famine deaths, or that the haphazard manner of the arrangement of the graves is any indication there was a crisis when those people were buried. Start thinking like a North Korean with no resources.

    You compare the haphazardly arranged North Korean grave mounds to the neatly arranged national cemetery in Seoul, South Korea. Well, Hamhung is not in South Korea. I know you point out the small area where there seems to be a neatly arranged cemetery. I have looked around other North Korean cities and found the same haphazard grave mound placement. Can you give an example of a large, neatly arranged, North Korean cemetery? Anywhere? Ok, there is probably a military one somewhere.

    Why would the North Korean cemeteries not be neatly arranged? Well, unlike South Korea, they do not have easy access to vehicles. They might be very poor. So much easier to drag grandma up the nearest convenient hill and bury her there, rather than carting her body miles away by hand for burial in a “proper” cemetery. Even if one had access to a vehicle, do you believe they would waste fuel to cart grandma miles away when there is that handy dandy hill right out back? Over the years, the convenient local hill becomes chock full of mounds, as I will demonstrate later, so who cares about arranging them in straight lines?

    No one has any way of knowing how old these graves are unless, as someone else has suggested, previous satellite images were to be compared. We could be looking burial mounds that cover decades worth of burials.

    If you look at your estimate of Hamhung’s population (approx. 874,000) and the current crude death rate in North Korea (approx. 10.5 deaths per 1,000 persons), one could expect more than 9,000 deaths in a city of Hamhung’s size per year, and this is without a famine (crude death rate for 2009 used.) Over 20 years, that would be more than 180,000 burials. I am assuming there is no tradition for cremation in Korean culture and everyone would be buried. I am also assuming a stable population and death rate over 20 years, which would not be the case. I’m talking rough estimates here.

    So if you could expect 180,000 deaths in 20 years in a city the size of Hamhung, and the population does not have the resources to transport bodies for burial elsewhere, no wonder every square inch of non-arable land within a reasonable distance of the city is covered in graves. I am assuming Hamhung has been a populated city for more than 20 years, so you can see how this problem would be compounded over time. The Korean culture is very old. I would assume one would find burial mounds thousands of years old in a well established city with a long history.

    I believe the question is not why there are so many graves on the hills, but why are there not more? The only way to draw any conclusions from these current satellite images is to compare them to pre-famine images of the hills around the city.

    One of the logical flaws in the article is to assume the small neatly arranged cemetary is the “starting point.” One can’t assume that unless you have older satellite images to compare it to. Just because it is neatly arranged, does not mean it is the oldest burial site. There are no more than 300 grave mounds in that portion of the cemetery (a very rough estimate, I didn’t count them.) That could be the burial area for one prominent family, or favored government officials, or a monument to the local war heroes from the Korean War. Considering you would expect 9000 deaths per year, that small neatly arranged burial plot is insignificant.

  7. jsternsp says:

    Also, I just checked out the Google Earth image of the Seoul National Cemetary. No mounds in the photograph. Marble grave markers and lots of flat green lawn. Looks much like Arlington. You can’t compare that cemetery to anything a North Korean would have the resources to accomplish.

  8. 1. I think I was clear in the post that definitive conclusions about the mortality shown in this picture are impossible without the confirmation of witnesses. My point is that the sheer number of graves here is (a) an aberration, compared to other cities, and (b) consistent with contemporary published reports.

    2. There is a tradition of cremation in Korean culture.

    3. If your math explains the number of graves, then where are the cemeteries of similar size around other North Korean or Chinese cities of comparable population? You’re assuming that graves last for decades, when that just might not be the case.

    4. I’ll recheck the Seoul cemetery imagery later on GE, but you’re right about one thing — a 16,000 foot view isn’t much use in making my point. I’ll put up something better later.

  9. jsternsp says:

    Joshua,

    I’m sure you are right. You have much more experience at this. (I was stationed at Camp Casey in South Korea in ’87-’88, but that does not make me an expert by any stretch of the imagination.) I have the annoying habit of trying to apply my powers of deductive reasoning to everything. I work in the legal field.

    Your response begs the question…what do they do with their dead under “normal” circumstances? Have you ever seen a neatly arranged cemetery in North Korea? If this burial arrangement is atypical, what is typical? I guess I will have to look around on Google Earth to see if I can find out.

    I actually rechecked the Seoul cemetery again, and there are small areas of the cemetery where you can see burial mounds, but it is a very small portion of the cemetery. Most of the graves seem to be a more western style marble marker sort of thing. There is one user picture posted in one of the larger burial areas and that is how I knew they used marble markers.

    I would think, under normal circumstances, although the famine was anything but normal, a grave would last at least a few decades until at least living family members who knew the deceased are dead. We have graves in this country that last for centuries, and most cultures have a strong taboo against disturbing the bones of their ancestors.

    One other thing I don’t know is how large these mounds actually are. I could not find a photograph of the burial mounds in the national cemetery in Seoul. I Googled “Korean burial mounds” but the images seemed to be of mounds that were quite large.

    Is it possible there is more than one burial in each mound? That is something else I failed to take into account in my analysis. I suppose it would make sense if one were dying of starvation and more than one family member died within a short period to bury them together. It is possible there are multiple remains in each mound.

  10. jsternsp says:

    Here is a link to a cemetery in North Korea near the DMZ posted by Reuters:

    http://www.reuters.com/news/pictures/articleslideshow?articleId=USSEO739320080911&channelName=worldNews#a=1

    The mounds seem small enough to indicate a single body, but I think it is quite possible there are multiple bodies under those mounds in Hamhung.

  11. jsternsp says:

    Shoot! I’m sorry. It’s a cemetery in South Korea at the DMZ, but the mounds are probably typical of Korean burial mounds.

  12. jsternsp says:

    Here is a link to a photo of an actual North Korean cemetery:

    http://www.travel-images.com/photo-korean2.html

    The graves are not burial mounds, but they are on a hillside on terraced levels. I posted another comment about the abandoned city in North Korea, where I wondered why there were graves on hillsides that seemed to be terraced for agriculture. Maybe it was not terraced for agriculture. Maybe that was an actual example of a “typical” cemetary. I will have to look at that again.

  13. jsternsp says:

    42.4220N, 129.4137E = What appears to be a “typical” cemetery in Chongjin, North Korea’s third largest city. The burial mounds are neatly placed on terraces on a hillside. It is more difficult to see because the pictures look like they were taken in summer when there was a lot of vegetation. I have seen very little evidence of haphazard burial mound placement around this city. The coordinates I have used are actually what looks like the newer part of the cemetery. The cemetery appears to be quite large. The older portions of the cemetery are south and east of the coordinates I give. I think I have answered the question about how to tell whether the burial mounds in Hamhung are old. You can clearly see in the older portions of the cemetery, the burial mounds are much less distinct, and the terracing less evident due to erosion.

    39.5933N, 127.3457E = A neatly arranged cemetery in Hamhung. Most of the graves seem more distinct and newer than the “famine” graves Joshua used in this article. When I look at it more widely, there seem to be other sections of cemetery nearby.
    This whole analysis is complicated by the fact that trees and bomb craters also look like mounds on satellite photos. I have what I believe to be bomb craters in the hills above Hamhung and Chongjin. If you don’t think bomb craters from the Korean War would still be visible, check out Pointe du Hoc in Normandy, France. They left it just the way it was on D Day and you can still clearly see the bomb craters from WWII. I have also read posts on this site that claim to see people when it is obvious they are just shrubs.
    Not to piss in anyone’s Cheerios…I think the images of the graves Joshua uses are from the city’s older cemetery. I think the terracing eroded somewhat. The graves do seem to follow the contour of the hill, and are not really haphazardly placed. If you look at the northern slope of the hillside where there are graves in the area Joshua marked with a yellow box, you can clearly see the shadows cast by terracing. One does not go to the trouble of terracing the hillside when one is starving and overwhelmed by the number of bodies to bury, I think.

    There are other areas of Hamhung where there are much smaller areas of graves that look more haphazard. It is impossible to know the rate at which a traditional Koran burial mound erodes. I think the older ones are less distinct. I think I have seen newer graves in Chongjin and Hamhung in well arranged cemeteries that are very similar to what you state are haphazard graves.

    My final analysis is that it is a typical North Korean cemetery with typical burials.

  14. jsternsp says:

    42.4220N, 129.4137E = What appears to be a “typical” cemetery in Chongjin, North Korea’s third largest city. The burial mounds are neatly placed on terraces on a hillside. It is more difficult to see because the pictures look like they were taken in summer when there was a lot of vegetation. I have seen very little evidence of haphazard burial mound placement around this city. The coordinates I have used are actually what looks like the newer part of the cemetery. The cemetery appears to be quite large. The older portions of the cemetery are south and east of the coordinates I give. I think I have answered the question about how to tell whether the burial mounds in Hamhung are old. You can clearly see in the older portions of the cemetery, the burial mounds are much less distinct, and the terracing less evident due to erosion.

    39.5933N, 127.3457E = A neatly arranged cemetery in Hamhung. Most of the graves seem more distinct and newer than the “famine” graves Joshua used in this article. When I look at it more widely, there seem to be other sections of cemetery nearby.
    This whole analysis is complicated by the fact that trees and bomb craters also look like mounds on satellite photos. I have what I believe to be bomb craters in the hills above Hamhung and Chongjin. If you don’t think bomb craters from the Korean War would still be visible, check out Pointe du Hoc in Normandy, France. They left it just the way it was on D Day and you can still clearly see the bomb craters from WWII. I have also read posts on this site that claim to see people when it is obvious they are just shrubs.

    Not to pp in anyone’s Cheerios…I think the images of the graves Joshua uses are from the city’s older cemetery. I think the terracing eroded somewhat. The graves do seem to follow the contour of the hill, and are not really haphazardly placed. If you look at the northern slope of the hillside where there are graves in the area Joshua marked with a yellow box, you can clearly see the shadows cast by terracing. One does not go to the trouble of terracing the hillside when one is starving and overwhelmed by the number of bodies to bury, I think.

    There are other areas of Hamhung where there are much smaller areas of graves that look more haphazard. It is impossible to know the rate at which a traditional Koran burial mound erodes. I think the older ones are less distinct. I think I have seen newer graves in Chongjin and Hamhung in well arranged cemeteries that are very similar to what you state are haphazard graves.

    My final analysis is that it is a typical North Korean cemetery with typical burials.

  15. Daniel says:

    I think what they left out were the concentration camps[Yodok].

  16. Leonard says:

    This is how Koreans traditionally bury their dead, it is the same in South Korea. Don’t look at official government run cemetaries as true examples in South Korea Traditionally mountains hold a spiritual meaning for Koreans so they bury them on moutaintops and sides. South Korea has trees on its hills so you can’t see the grave mounds on them, but they are there. North Korea has stripped its land of trees, mountaintops and sides are visible and thus there are more grave mounds visible. Because North Korea needs all the land it can get, grave mounds are concentrated so they won’t take up much room. Plus, in South Korea the growing trend is for cremation, not burial. I don’t see this as evidence of anything other than Koreans die and Koreans bury them in mounds on hills.

  17. P Elisabeth says:

    42.894444,130.215915
    More mass graves, close to the Chinese border, the graves are still open…

Leave a Reply

Your email address will not be published. Required fields are marked *