The North Korean Counter-Force Catalog
Or, What We Can Learn From Google Earth IMINT
The world once again finds itself on the brink of a North Korean crisis, and this time matters have come to a head - the DPRK has developed a capable solid-fueled ICBM and a tested thermonuclear warhead to place upon it. Theories have been overturned by the cold facts of capabilities, and now there is only a matter of months before these new weapons are fielded in significant numbers. The free world in general, and America in particular, now have very little time to weigh the consequences of action versus inaction.
Unfortunately, the academics and thinkers whom usually inform and enrich the public debate seem largely unable to grasp the nature and gravity of the current crisis, preferring retreat to the familiar bastion of Cold War era deterrence theory. As Crispin Rovere neatly summarized: "There are analysts discounting the possibility of war, but based on shallow reasoning: North Korea has nuclear weapons, nuclear war is unthinkable, therefore there will be no war." More and more frequently, the terrible specter of war is dismissed out of hand in the opening sentences of op-eds and think tank columns, as if ignoring the beast will compel it to leave. Recent rumblings from the Pentagon have made clear, however, that our military leaders cannot indulge these luxuries - they must confront the possibility, and thus the actual balance of capabilities that determine the outcome.
Unfortunately, public discussion is badly hindered because open-source information on the actual capabilities in play are painfully scarce. Even Barry R. Posen of the New York Times came up nearly empty, forced to extrapolate from published estimates of North Korean Transporter-Erector-Launcher (TEL) numbers and pragmatic worst-case assumptions (every TEL in its own hardened bunker) to work out the weapons required - and thus the scope of the military effort needed. Indeed, this dearth of information is so complete that Roger Cavazos's now well-known article debunking the myth of DPRK artillery leveling Seoul, "Mind the Gap Between Rhetoric and Reality," relied heavily on amateur Google Earth IMINT research done by an anonymous internet user known only as "Planeman!" And even this well is dry, the original internet discussion forum Planeman's work was hosted on having been shut down ages ago, and few others pursuing the same efforts. It's perhaps no surprise - at least one other talented amateur IMINT analyst received a job offer from Janes. (If you’re good at something, never do it for free, as they say.) Thus the public discussion of the potentials and perils of “the military option” in North Korea has been starved of information. The voting electorate are the final arbiters of power in American democracy, as they will eventually hold their congressional representatives - and their President - to account for whatever outcomes generated by their decisions now. This grants popular opinion great influence over events (and North Korea knows it,) yet those commentators seeking to sway the masses have almost no information or evidence to present them with.
Though unschooled and unworthy in the art of IMINT, I am an educated journalist, and can at least attempt the “footwork” to aid others. Fortunately, I can stand on the shoulders of giants in this effort - especially Jacob Bogle and his amazing work on “Access DPRK,” a colossal project to comprehensively map places and features in North Korea via free satellite imagery in Google Earth. Without Mr. Bogle's work - and all 53,722 Google Earth placemarks he generated - my efforts would be in vain. I also owe a debt to bookmarks compiled by “nkbypanda,” an anonymous amateur analyst, and even the original Google Earth .kmz bookmark file by “Planeman.” Building on their work, I've taken the next step and tried to classify hardened sites in North Korea, with an emphasis on identifying every hardened bunker or facility that might accommodate a TEL, in order to quantify the true scope of the challenge facing the Pentagon.
The Google Earth .kmz bookmarks file can be found here. Though still very much a work in progress, some general points of great significance are immediately apparent and are worth sharing now (especially with the ever-shorter timelines of potential crisis on the peninsula) and by publicizing I might invite the critique of the more informed.
Scope Of The Counter-Force Problem
Any preemptive strike on North Korea must assume worst-case scenarios - namely, that the regime will respond to any attack, no matter how limited, with full scale retaliation against Seoul, Tokyo and even the United States with a significant fraction of the weapons in their inventory, including conventional, chemical, biological and nuclear weapons. In other words, counter-force strike is the only feasible preemptive option. The conventional military balance on the peninsula being what it is, the equation is predominantly an “allied counter-force strike versus DPRK counter-value arsenal” problem. Even the oft-lamented artillery threat to Seoul is only notable when considered as delivery systems for chemical warheads.
Of the DPRK's known capabilities for delivering counter-value WMD attacks, three systems are predominant. In order of lethality, these are their SRBM arsenal, their 240mm Multiple Launch Rocket Systems, and their 170mm “Koksan” self-propelled heavy artillery guns.
This owes to several practical considerations. First, their solid-fueled, road-mobile IRBM and ICBM forces are still in their nascent phase, with their first successful tests conducted very recently, late last year at best, (in the case of the Pukkuksong-1.) Thus, at time of writing, North Korea simply hasn't had the time to produce and field them in significant numbers. This leaves their only operational MRBM the “Nodong,” a liquid-fueled weapon which is leashed to fixed bases and sizable support convoys, hindered by long (and fairly obvious) prep times, and above all is a 16 meter long weapon which makes it particularly ungainly to lug about the rugged North Korean terrain - especially on wheeled TELs.
The DPRK's SRBM arsenal (chiefly domestic variants of the 12-meter long, solid-fueled Scud) are far more mobile, compact, and quick to prep and launch. They can be dispersed faster, across more possible terrain, and hidden in more places for a longer time than the Nodong. Most crucially, their TELs - which determine how many they can fire at once - are available in numbers (published estimates vary widely, the median being 200 TEL vehicles or so.) Combined with their ability to deliver North Korea's first generation of fission weapons, (unlike MLRS or artillery,) the DPRK's Scud force is the most survivable counter-value asset they have, and - despite the bevy of land and sea based anti-missile systems now deployed to the ROK - the one most likely to survive in numbers sufficient to saturate defenses and strike Seoul. (Striking America or even Tokyo is unnecessary, as the potential devastation of even a single low-yield warhead striking Seoul proper is more than enough to constitute an effective deterrent - which is why North Korea still exists.)
Thus, quantifying the scope of the counter-force challenge depends on finding the Scuds, the MLRS, and the 170mm artillery. This is what I've found so far.
The Bunker Blitz - A Concerted Asset Dispersal Effort
The most telling insight so far has been a downright feverish effort by North Korea to build new reverse-slope bunkers, mostly near the DMZ opposite Seoul, beginning in 2009 but peaking between 2011 and 2013. Especially in the latter timeframe, new bunkers on reverse slopes appeared almost everywhere - the .kmz file has northwards of 300 so far. They come in three distinct styles, which I've dubbed “large,” “medium,” and “Koksan.” The lattermost one is distinctive, and sheds light on the goals of the entire reverse-slope building effort:
The curious shape, with its narrow room off the back, was ultimately explained by “nkbypanda's” past work, which provided a ground-level view of storage sheds for the 170mm Koksan gun and overhead imagery to reference. The presence of these storage sheds in close proximity - invariably with gun barrels visible, and occasionally an entire gun parked outside - confirmed the purpose of these bunkers. (The narrow room accommodates the long barrel of a Koksan gun.) They're almost always built in sets of six, with two sets placed as close together as terrain allows (for a full battalion of 12 guns.) The distinct shape of these bunkers (visible in post-construction surface scarring) and presence of Koksan gun sheds allows identification even when imagery of the construction phase is poorly lit, poorly timed, or even nonexistent. So far I've found nine Koksan battalions (about half within 60km of Seoul's center) and a few half-battalion (6 gun) or single-battery (4 gun) dispersal sites for use in wartime (no unit permanently stationed there.)
In addition to being direct evidence of very scarce reports and anecdotal evidence concerning reverse-slope basing of 170mm Koksan guns, this casts other bunker-building efforts in the same timeframe in a decisively offensive light. Half the battalions so far discovered are more than 60km from Seoul's city center (the maximum theoretical range of the Koksan if using Rocket Assisted Projectiles.) They are, however, well-positioned for deep fire on the Chorwon invasion corridor (see page 11.) This makes their other bunkers more suspicious - despite being much closer to the DMZ (and thus more vulnerable) than required to range Seoul, such aggressive forward basing also maximizes penetration chances (steeper, faster re-entry trajectories to avoid ABM defense) and their reach into the southern ROK. In other words, they’re not sited for strictly counter-value employment.
If Pyongyang harbors hopes of forceful unification still (after precluding American involvement with threats of ICBM strikes on American soil,) then they'll need the asymmetric advantage of WMD-equipped ballistic missiles to overcome the conventional forces imbalance on the peninsula, especially to strike more distant ROK bases, transportation chokepoints and other military targets. (The DPRKs recent pursuit of long-range precision conventional fires such as the KN-09 300mm MLRS, and local Tochka derivative, the KN-02 - both of which are rumored to have optical scene-matching terminal guidance - lends strength to this suspicion. If Pyongyang's motives are strictly self-defense deterrence, then pouring scarce money into new precision capabilities instead of more mature, legacy counter-value systems capable of carrying the DPRK's heavy first-gen fission weapons makes little sense.) With nuclear warheads, even Scuds are tactically and operationally relevant weapons, not just counter-value assets - which suggests that at least some of the new bunkers, outside of the 136 Koksan bunkers so far identified are meant for dispersal of Scud TELs.
Which Bunkers Hold What?
Discovering where TELs might be is easiest to do by ruling out where they can't be - i.e. where they cannot fit, or cannot access. The dimensions of bunkers are an obvious starting point.
Outside the Koksan bunkers, two other types predominate - a “medium” bunker (measuring roughly four and a half to five meters wide, and thirteen to fourteen meters long,) and a “large” bunker with a six or seven meter width and a 17 to 18 meter length. Even with many bunkers floor plans open and visible during construction, solid measurements are a bit difficult due to various factors - slant angle distortion, shadows hiding the bases of walls, and above all the half-meter resolution of the free DigitalGlobe imagery that predominates in the Google Earth database. When considering the impact of interior wall thickness on overall internal dimensions, the resolution limits impose obvious problems in calculations. The figures given are an overall guesstimate produced over the course of identifying, measuring and cataloging many bunkers.
Can a North Korean Scud TEL fit in these? That depends on the TEL, which complicates matters because various TELs have been shown off in DPRK parades, owing to their need to import and improvise with whatever they can spirit past the sanctions. The length of a Scud missile puts a hard limit on length, however - 12 meters. And a great number of their extant Soviet-delivered stock (as displayed in parades) are likely the Maz-543, which gives us a width. Determining a turn radius required some hunting, but a variety of sources confirmed 15 meters as a reasonable approximation. (Interestingly, half that of the American HEMTT owing to the unique two-axle articulation.) Thus informed, one can analyze road access:
Once one begins sliding around a to-scale rectangle of cardboard around their screen making motor sounds with their mouth, the lesser glamour associated with the IMINT disciplines becomes apparent. Nonetheless, it's effective, and once practicalities are accounted for (such as room to open vehicle doors, and the icy nature of severe Korean winters) certain conclusions become more or less probable. As usual, there's always another wrinkle, such as North Korea's recently revealed tracked Scud TEL, but given the presumption of a mostly wheeled legacy fleet, broad conclusions can still be drawn.
Observing Patterns and Environs
Overall, some patterns emerged:
Every “medium bunker” site had road access that could accommodate a MAZ-543 TEL (though some were tight fits.)
“Large bunker” sites invariably had good road access, but not always sized to comfortably accommodate a MAZ-543's 3.06m wide TEL, even when it would've been easy to do so (see second image above.)
“Medium bunkers” are often associated with large, vehicle-capable tunnels (with widths, slopes and turn radii that can accommodate a MAZ-543 TEL) boring into presumed large garages or bases beneath hills or even mountains, facilities that the oldest imagery shows predates the 2009-2013 construction blitz.
“Large bunkers” are often associated with large nearby military bases with a preponderance of barracks, usually favor areas with flatter, wider-open terrain and are more often found very close to the DMZ (as little as four kilometers on occasion.)
This may imply:
The “medium bunkers” are dispersal bunkers for wheeled TELs, and perhaps for 240mm MLRS, which can be difficult to conclusively separate from the HARTs built for the latter system (predominantly distinguished by presence of firing revetments and shrapnel-shield dirt mounts in front of the entrances, as seen here.)
The “large bunkers” are meant for housing APCs, IFVs and supply vehicles for forward-positioned troops who might be called upon to drive on Seoul in event of war.
There's many caveats to these conclusions - the “medium” bunkers would accommodate a pair of 240mm MLRS (with room for reloads and accommodation for the crew as well) much more comfortably than a MAZ-543 (which can just about squeeze in with room for the driver to wiggle out,) to say nothing of the road access at some sites as well. Additionally, the “large” bunkers, aside from being a more comfortable fit, could provide room for extended crew accommodation (cots, stove, basic maintenance equipment, etc.) which would be of obvious use when TELs are dispersed in times of high tension - which may last indefinitely. Given the well known and longstanding concentration of American ISR capabilities near the DMZ, moving TELs around is something the KPA will try to avoid.
However, the preponderance of evidence seen across multiple sites impels me towards the above conclusions - especially in light of what the sudden bunker-building effort implies.
Deliberate Dispersal of High-Value Assets Away From Legacy Bases
Tangible information on the famous North Korean Hardened Artillery Shelter - outside of a single Nautilus publication from the late 80s and the occasional KCNA propaganda clip - is hard to come by. Looking at the sites themselves, however, reveals a multitude of types with varying protection - and that HARTs for self-propelled guns, including the Koksan, never provided overhead protection for guns while they were actually firing. Indeed, some of the new Koksan bunker sites were built at already-extant Koksan battalion bases, with the gun sheds visible pre-2009 and the old firing revetments nearby still visible today. The bunker-building is almost certainly a reaction to the drastic shift in artillery effectiveness enabled by modern computer fire control and fast proliferation of cheap GPS guided shells with impressive accuracy, which render revetment protection mostly useless.
This strongly implies the other bunkers were also built as intelligent adaptations to changing ROK/US capabilities as well. Almost no information is available on North Korean missile bases (outside of a handful of badly outdated anecdotes from defectors and vague military press releases,) leaving us with naught but ludicrous claims of missile bases deeper than 1,000 feet (despite defector's testimony about the DPRK's inability to deal with the water table) that are nigh impregnable to all attack. Impregnable though the bunker may be, the exits are not so blessed:
As long as the bunker-busting bomb goes off on the right side of the blast door, the occupants will be unharmed (the image illustrates how the GBU-57 MOP subverts the bunker-builder's expectations in dramatic fashion.) But intact or not, a few tons of rock in your tunnel entrance complicates vehicular traffic some. A TEL that cannot sortie is a TEL that cannot fire. If DPRK missile bases are indeed buried deep under mountains, the last decade of rapid advances in precision standoff guidance weaponry and ever-more advanced and effective bunker-busting warhead design can be expected to compel rapid KPA efforts away from these newly vulnerable bottlenecks.
Indeed, this seems to be the case. The facility I dubbed the “Kaesong Missile Base” features three generously-sized tunnels that lead under a sizable hill, and in historical imagery a 16-meter long vehicle is visible - almost certainly a Nodong missile on its TEL. Though the number of recently-dug bunkers in close proximity to this base are far more vulnerable than the under-mountain base, each must be attacked individually - a far higher burden on any would-be attacker than simply sealing every missile underground by hitting just three tunnel entrances. Given the short distances of the entire Korean theater, and conventional Allied military strength, by the time base occupants dig themselves out, they'll be in occupied territory. Thus the mass of the first retaliatory salvo is paramount. In light of modern weapons, this strongly favors dispersal of SRBM assets.
In light of this logic and the frequent observed association of pre-existing, TEL-accessible bases beneath hills and the “medium” sized bunkers, I find it highly plausible they're meant for dispersing SRBM launchers.
Conclusions
This document is informative; detailing the research and rationale that underlie the guesses offered within the .kmz file itself. I advance no argument from this data as-is, especially as the research is incomplete (in MLRS HARTs and hardened airbase facilities especially) and I must expose it to the criticism of the more experienced and informed before trusting it with such weight. However, a few simple conclusions are self-evident:
The DPRK is an alert, adaptive and responsive enemy who is keenly aware of allied counter-force abilities. The enemy is never idle.
From their decision to keep investing great money and effort into dispersal bunkers, despite concurrent investments in TEL mobility and the obviousness of construction efforts (for instance, amateurs can easily find them on Google Earth's low-res imagery) one may conclude they rate their chances of evading modern ROK/US ISTAR assets in a compact and terrain-constrained theater to be too low to ensure force survival alone.
Once the tracked TELs revealed this year are put into mass production, the counter-force targets will go from two-hundred odd wheeled SRBM TELs to more mobile, nimble and survivable TELs - and their already-mobile Scuds will be an order of magnitude more elusive.
In sum: time is short.
Commentary and questions best directed to my twitter, or use the anonymous “ask” box on this blog.
