Underground Intrigue: A Conjectural Assessment of Incursion Tunnels in a Future Nagorno-Karabakh Conflict

Overview and Background

2014 already appears to be as bloody, if not more so, than previous years along the front lines of the Nagorno Karabakh conflict. It seems that every week or two the so-called frozen conflict claims a new victim – typically a young Armenian or Azerbaijani soldier, but on occasion an innocent civilian who ended up in the wrong place at the wrong time. While the majority of the fatal ceasefire violations occur along the de facto Nagorno Karabakh Republic – Azerbaijani line of contact, they have also broken out along Armenia’s northeastern international border with Azerbaijan. More worrisome still, recent reports suggest that the violence may also be spreading to Armenia’s southwestern border with Azerbaijan’s Nakhchivan exclave.[i]

A plethora of documentaries and articles have sought to capture the tense atmosphere at the front – with commentators often comparing the situation to the type of trench warfare that occurred on the Western Front during World War I.[ii],[iii],[iv] Similar to Verdun, France one hundred years ago, two entrenched and generally static forces face off against each other along the Karabakh line of contact. Corroborated by a casual Google Earth investigation, reports suggest opposing trenches are, at places, separated by only a few dozen yards. Recent fatalities also indicate minefields are scattered throughout no-man’s-land, although it’s not entirely clear whether these are newly emplaced or remnants from the hot war that ended in 1994.[v] There are a number of acute differences between the Karabakh situation and prototypical trench warfare, however. Perhaps the most obvious is the fact that the Karabakh conflict no longer includes human wave attacks, although these suicidal charges did occur in the waning days of the war.

While major offensives no longer seem to be in the offing, at least in the near term, regional news reports reveal that ‘commando raids’ still occur on a frequent basis.[vi] It’s difficult to ascertain which side is responsible for the raids owing to mutual recriminations, but on the face of it, Azerbaijan seems to have the most to gain from such provocations considering it’s the least satisfied with the status quo. Nevertheless, it’s reasonable to assume the Armenian side also stages punitive attacks to demonstrate its resolve and military capability. The situation along the line of contact, therefore, can best be described as unstable, but with neither side willing to commit to a major offensive designed to capture territory. However, the purpose of this article is not to recap the current situation on the ground, but to focus on another aspect of the Karabakh conflict that doesn’t receive coverage – namely, what’s going on underground. In other words, this paper will explore the use, or non-use, of offensive incursion tunnels.

Armies have employed tunneling to bypass or undermine fixed defenses for hundreds if not thousands of years. The Ottoman Turks employed the technique in an attempt to capture Vienna in the 16th century only to be thwarted by the city’s defenders, who successfully dug what can best be described as counter-tunnels.[vii] During World War I, opposing armies hoped tunnels could help break the trench-warfare stalemate.[viii] Moreover, one should not mistake this tactic to be some obsolete historical relic, as four North Korean incursion tunnels, large enough to accommodate an invasion force of thousands, were discovered in the latter half of the 20th century; and commentators speculate that there may be dozens more yet to be discovered.[ix],[x] Even in our back yard, Mexican drug cartels successfully dig secret tunnels to bypass the heavily fortified and monitored southern U.S. border.[xi],[xii]

Given the evident utility of tunneling, it’s reasonable to ask why Azerbaijanis or Armenians aren’t employing the method to circumvent or bypass the heavily defended Karabakh line of contact. A more circumspect individual might caveat the question by further asking why neither side has reported the use of tunneling, since a fairly comprehensive, English-language, internet search yields almost no substantive documentation on the subject – hypothetical or factual (there admittedly are a few passing references in forums). After ruminating over the mystery a while, one would likely arrive at one or several of the following explanations:

  • (Hypothesis 1) Neither side is able to dig incursion tunnels for financial or technical reasons.
  • (Hypothesis 2) The opposing forces deliberately choose not to dig tunnels despite the fact that they are physically capable of doing so.
  • (Hypothesis 3) The sides have indeed built secret contingency/incursion tunnels, which will only be used if/when the conflict reverts to a hot-war.
  • (Hypothesis 4) Armenian and Azerbaijani forces actually use tunnels on a regular basis to stage commando raids, but choose not to release this information to the press.

Hypothesis 1

In evaluating these four options, it seems the first scenario can be safely discarded for a number of reasons. Azerbaijan frequently boasts that its annual defense budget is larger than Armenia’s entire state budget, because of its burgeoning oil industry. This fact alone indicates the country is financially capable of digging tunnels only a few hundred yards in length, especially considering previous instances in which less-endowed antagonists were able to afford similar undertakings, e.g. North Korea. While the Armenian/Karabakh side is admittedly poorer than its rival, Armenia’s annual defense budget is still ~400 million dollars per year. This figure is more than a hundred times the amount purportedly needed to build a sophisticated narcotics tunnel beneath the Mexican-US border.[xiii],[xiv] In other words, Armenia would only need to devote one or two percent of its military budget to carry out this type of activity, which is hardly an insufferable expense.

Technical challenges also fail to satisfactorily explain the apparent lack underground excavation. While the topography of the line of contact is indeed rugged and daunting at places, it nevertheless remains surmountable from an engineering standpoint. An anecdotal example that reflects this reality is the strategic auto-tunnel that lies to the south of Kelbajar.[xv] From a more macro-scale/geological perspective, much of the line of contact is situated on ground that is composed of flood plain, alluvial deposits.[xvi] This loose soil/gravel/stony mixture usually doesn’t require dynamite or heavy equipment to move, which seems evident enough after viewing photos of existing trenches on the frontline.[xvii] There is one noteworthy exception to this general rule, however; namely, the east-west running Murovdag/Mrav mountain range, which consists of harder basalt, sandstone and limestone. In any case, if the opposing sides wanted to engage in tunneling, it appears they would be able to do so – at least for most sections of the front.

Hypothesis 2

The second explanatory hypothesis considers the possibility that there is a de facto – if not de jure – policy against underground excavation. There are a number of reasons why a leader might deliberately choose to forgo the advantages derived from tunneling. For instance, they may fear a scenario in which the adversary showcases a discovered tunnel to prove that the other is acting in bad faith vis-à-vis the ceasefire regime. This theory breaks down under closer scrutiny, however, because the antagonists, Azerbaijan in particular, do not appear to be overly fearful of making truculent statements and carrying out provocative actions. This is best exemplified by Baku’s 2012 decision to free and publically venerate Ramil Safarov, who brutally murdered an Armenian serviceman during a NATO exercise in Hungary. This inflammatory action generated nearly universal western condemnation, yet Azerbaijan adamantly maintains that it was acting well within its rights. In a similar vein, Baku frequently claims it is developing and/or possesses the military capability to forcibly resolve the Karabakh conflict in the event that the ongoing Minsk Group break down, which could antagonize European and American interlocutors who repeatedly proclaim that there is no military solution to the Karabakh issue. Given this behavior, it seems unlikely that Azerbaijan would shun subterranean activity solely to avoid hypothetical Western criticism. And even if Armenian forces were to claim that they located an Azerbaijani incursion tunnel, Baku could always deny responsibly or could call such accusations “lies,” thereby ensuring that true and accurate revelations get lost in the broader propaganda war currently being waged between the two countries.

Another factor that may deter political or military commanders from ordering underground operations is the prospect that such a decision could needlessly and dangerously destabilize the line of contact, and, by extension, the status quo. While the current situation on the front is bloody, it is nevertheless predictable, if not well managed. After one or two soldiers are killed on one side, several troops on the opposing side will typically end up getting shot over the course of the next several days; this cycle then repeats itself on a periodic basis. Introducing tunneling into the equation, however, could jeopardize this set pattern by allowing one or both sides to engage in asymmetrical retaliation. Provided the antagonists are able to covertly enter enemy territory via incursion tunnels, one side could swiftly kill a dozen rivals to avenge the death of one of their own. This in turn may precipitate an escalating response from the enemy (e.g. heavy artillery), which would set in motion an unexpected and undesired chain of events that could easily spiral out of control. Thus, since neither side is interested in an accidental war, it’s possible they have arrived at some sort of de facto or gentlemen’s agreement to avoid underground warfare altogether.

Hypothesis 3

While the preceding theories explain why incursion tunnels haven’t been constructed along the line of contact, it’s entirely conceivable that we (the public) are unaware of existing tunnels simply because regional media outlets haven’t reported their existence. One would actually expect to hear very little on the issue from traditional news sources, because incursion tunnels are most useful when they remain unknown to the enemy; otherwise, the opposing side would carry out disruptive countermeasures, (e.g. the abovementioned counter-tunnel-tunnels dug by the Viennese against the Ottoman Turks). As such, it stands to reason that the Karabakh combatants wouldn’t inform their own national media about their covert/clandestine underground activities. It’s also quite possible that government or military officials have explicitly instructed, or else pressured their country’s reporters to refrain from reporting on the issue, citing national security concerns. This theory, however, fails to explain why Azerbaijan doesn’t publicize Armenia’s tunnels and vice-versa. To account for this reality, an observer would likely arrive at one of the following conclusions: the sides simply aren’t tunneling (see Hypotheses 1 and 2); neither side knows if, or where the enemy has engaged in tunneling operations; or lastly, the combatants are keeping their knowledge of enemy digging operations secret, in order to preserve the element of strategic surprise in a contingency situation, i.e. depriving the enemy of an asset, which the latter had integrated into its war plans. The latter two explanations, therefore, indicate there are at least a few realistic scenarios in which unpublicized tunneling might occur.

If Azerbaijan and Armenia are indeed engaged in underground tunneling, one would naturally be interested in determining the nature and scale of these operations. To accomplish this goal, it would be prudent to refer to antecedent case studies in order to draw appropriate lessons and rules of thumb. One North Korean incursion tunnel, for instance, is reported to be a mile long, two meters across, and two meters high. Reports also indicate it’s large enough to allow 30,000 soldiers to pass through every hour.[xviii],[xix] Given these dimensions, it’s clear the tunnel was designed with strategic objectives and missions in mind, i.e. a full-blown invasion of the south. Tunnels of a more tactical nature, however, will likely be smaller in scale and length. Therefore, in drawing parallels to the Karabakh conflict, one should first try to understand the antagonists’ doctrine and end-goals, as well as their manpower and resources.

With respect to the latter, Azerbaijan frequently demonstrates its growing offensive capability by way of huge arms purchases, ballooning defense budgets, and technological advances; the country also claims its military is capable of liberating all disputed territory within 10 days.[xx] Azerbaijan’s official, parliament-sanctioned doctrine is also readily available on the internet, and it (not surprisingly) establishes the country’s right to use military force to “liberate” Nagorno Karabakh.[xxi] Given this information, there is little reason to doubt that Baku may have actual plans to eventually breach the line of contact and invade Karabakh, even if its near term goal in making these belligerent proclamations is merely to intimidate the Armenian side into sacrificing additional concessions in the ongoing Minsk Group negotiations. Whether these plans entail the conquest of all of Karabakh or are of a more limited nature, however, is a matter of a great deal of speculation.[xxii],[xxiii] Moreover, one should also consider the fact that plans can quickly change or become obsolete depending on circumstances on the ground, and in this respect, Azerbaijan’s objectives may shift as it succeeds or fails to attain near-term objectives. Regardless, any major offensive would likely require Azerbaijani forces to quickly penetrate multiple layers of defense.

North Korean-style, strategic tunnels, therefore, should prove appealing to Azerbaijan (Hypothesis 3), since they theoretically would spare its forces from having to engage in high-casualty human wave attacks. This latter point is of prime importance, since traditional military theory asserts that the attacking side should have at least a three to one numerical superiority over a defender in order to have a reasonable chance of success.[xxiv] According to open source reporting, Azerbaijan is presently assessed to have ~95K personnel (plus reservists), while Armenia and Karabakh together are thought have a cumulative of ~67K personnel (plus reservists).[xxv],[xxvi] Consequently, Azerbaijan may only have a numerical superiority of three to two, although local disparities and concentrations can vary to some extent. Regardless, Baku would likely seek ways to bridge any personnel shortages via force multipliers (e.g. superior equipment or command & communication infrastructure), and force preservation measures (e.g. incursion tunnels).

With respect to Armenian doctrine and capabilities, it is often assumed that the country maintains an exclusively defensive posture, because it already occupies most of the territory of the former Nagorno Karabakh Autonomous Oblast – meaning, it’s relatively satisfied with the status quo. Nevertheless, senior Armenian military leaders occasionally indicate that they too reserve the option of carrying out an preemptive strike against Azerbaijan if the latter felt it necessary to secure Karabakh.[xxvii],[xxviii] As with Azerbaijan, it’s difficult to determine if this rhetoric is simply propaganda or if it reflects genuine policy per se, although Armenia’s acquisition of short range (300 km) surface-to-surface missiles suggests there is at least some desire in Yerevan to develop strategic offensive capabilities.[xxix] Therefore, it’s not outside the realm of possibilities to assess that Armenia may also seek and/or possesses strategic-scale incursion tunnels bisecting the line-of-contact.

Whether they’re Azerbaijani or Armenian, these hypothesized invasion tunnels would most likely be located in areas that possess some combination of the following attributes: 1) favorable geology and topography; 2) adversary deemed to be particularly vulnerable; 3) a lighting offensive would surprise the enemy. For instance, the geology near Agdam may be conducive to tunnel digging (supporting factor 1), but many analysts have speculated that an Azerbaijani offensive in this area is among the most likely scenarios (partially undermining factor 3).[xxx] Strategists in Baku, therefore, would have to weigh these respective advantages and disadvantages accordingly. Planners must also consider the risk that the enemy will discover the location of their strategic tunnels, as this would result wasted resources as well as ruined operational plans. This fear, in turn, would likely preclude either side from using strategic tunnels for minor punitive attacks. In other words, once built, these tunnels will most likely remain unused.

Hypothesis 4

In contrast with strategic tunnels, combatants may tempted to actually use smaller, tactical tunnels in the current environment, because their discovery would not necessarily ruin any invasion-scenario war plans. Such tunnels, in fact, could theoretically explain how commandos sometimes succeed in sneaking across a no-mans-land, which is ostensibly littered with minefields and sniper nests.[xxxi] However, this theory appears to be partly undercut by reporting of alleged civilians and conscripts inadvertently wondering across the line-of-contact, suggesting the front is not entirely impervious to overland excursions.[xxxii],[xxxiii],[xxxiv]

Another reason this hypothesis falls short is because regional English-language media have not reported the existence of enemy expendable/one-time-use tunnels, even though (as mentioned above), such a disclosure would not affect either side’s war plans (in contrast with strategic tunnels). Indeed, there is no obvious reason for Azerbaijani or Armenian officials to withhold from their respective media outlets the fact that such tunnels were discovered in post commando-raid investigations. Advocates of the fourth hypothesis may also postulate that the combatants are simply unable to locate enemy tunnels – even after they’ve been used in a punitive attack. This explanation, however, seems fairly implausible, given the forensic/investigative technology likely in the possession of both sides; but this author concedes that there may be a few situations where post-use tunnels remain undetected (e.g. extremely complex terrain). In any event, the fourth hypothesis appears less likely than the third.

Locating Hidden Tunnels

While any of the above mentioned hypotheses is possible, it would be wise for the warring sides to anticipate a worst case scenario, which in this case, stipulates that one or both combatants have constructed secret, strategic tunnels to be used in a future, large-scale offensive (hypothesis 3). With this threat in mind, one might ask what the Armenian side is doing in order to detect dormant Azerbaijani incursion tunnels – and vice-versa. To answer these rhetorical questions, one should first consider past precedent.

As mentioned earlier, U.S. border patrol officers routinely locate narcotics tunnels that bisect the Mexican border. According to open source reporting, many of these were discovered through the use of informants, or else were chanced upon by officers out on patrol.[xxxv] Similarly, one of the aforementioned Korean incursion tunnels was supposedly disclosed via a North Korean defector, and another was located after South Korean soldiers observed steam suspiciously emanating from the ground.[xxxvi],[xxxvii] Given the above, it appears human intelligence (HUMINT) is a fairly effective method for discovering clandestine underground activity. This is especially applicable in the Karabakh conflict, considering recent reports, which indicate that Armenian and Azerbaijani intelligence services regularly task recruited sources with collecting intelligence on the military activities and capabilities of the enemy.[xxxviii],[xxxix] Thus, one should expect sources to be specifically charged with determining the existence and location of tunnels along the front.

Although HUMINT may be a time tested collection method, the two Karabakh combatants may also be relying upon more modern, high-tech means to detect nefarious underground activity. Among the more interesting methods is acoustic seismology, which essentially entails sending compression/acoustic waves through the ground and recording the subsequent reverberations. Based on the varying densities of underground material, different (i.e. time delayed) acoustic reflections will be recorded by a sensor on the surface. One should, therefore, be able to detect a tunnel, since the air that occupies it is significantly less dense than the surrounding rock.[xl] The United States itself has explored the possibility of using this method to find narcotics tunnels, but technological challenges remain an impediment (e.g. those pertaining to irregular tunnel shape).[xli] Thus, one could speculate that Azerbaijan or Armenia likely find acoustic seismology equally unreliable.

As far as high-tech gadgetry is concerned, a more appealing option may be ground penetrating radar. Indeed, a number of manufacturers specifically tout their systems’ military applications, including their ability to detect concealed tunnels.[xlii] As one may infer, this method relies on electromagnetic (vice acoustic, transmissions) and could theoretically allow one to detect underground anomalies up to 50 feet deep. Ground penetrating radar, nevertheless, has its own limitations, including degraded performance when operating over ground that is heterogeneous in composition.[xliii] Therefore, if the soil near the Karabakh line-of-contact is too stony, this method could prove ineffective.

While there may be countless additional ways for Armenia or Azerbaijan to locate hidden tunnels, for the sake of brevity this paper will focus on just one more – namely, cutting edge infrared cameras. As mentioned earlier, South Korean soldiers were able to find an incursion tunnel after observing a suspicious steam vent during a late-November patrol in the 1970s.[xliv] Operating on the same concept, modern infrared cameras are capable of detecting minute temperature changes that occur as a result of underground tunneling operations. In fact, this method has already been field-tested along the U.S. border with varying success, and a number of private firms claim they are able to detect hidden tunnels by conducting infrared pattern analysis, i.e. comparing before-and-after infrared images of the same area.[xlv],[xlvi] Nevertheless, the technique still appears to be in the developmental stage, which may dissuade the Karabakh factions from following suit.


Considering the above analysis, what are the implications for the Karabakh conflict? Given their historical use and potential effectiveness, it’s possible both sides are furiously digging tunnels in preparation for a major invasion or preemptive attack. By extrapolating from the Korea precedent, these theorized tunnels have the potential to significantly alter the course of events should a major war re-erupt. Tens of thousands of troops could simply bypass heavily fortified trenches – thereby sparing them from mass casualties, while also allowing them to wreak havoc in the enemy’s rear areas. In essence, tunnels could be a game changer.

Conversely, there’s a decent chance that the above fear is sheer speculation and that Azerbaijan and Armenia are not engaged any underground excavation whatsoever. Either way, the public is in the dark (no pun intended), since the absence of evidence does not necessarily signify evidence of absence. Moreover, since military leaders from both sides are faced with a similar conundrum, it would be logical for them to dig tunnels, because they would (or should) assume that their foes are doing the exact same thing. Operating according to this prisoner’s dilemma-type logic ensures that neither side relinquishes some theorized strategic advantage. Prudent commanders will also likely try to shed light on the matter by seeking to confirm or deny the existence enemy tunnels by using some or all of the means discussed in the section directly above. Thus, a lot may be going on along the Karabakh line-of-contact, which hasn’t yet come to light (pun intended).

[i] http://armenianow.com/karabakh/55432/armenia_karabakh_soldiers_killed

[ii] http://news.bbc.co.uk/2/hi/programmes/this_world/one_day_of_war/3622431.stm

[iii] http://www.opendemocracy.net/od-russia/thomas-de-waal/nagorno-karabakh-crimea-doppelganger-azerbaijan-armenia

[iv] http://mondediplo.com/2012/12/12karabakh

[v] http://news.am/eng/news/201998.html

[vi] http://armenianow.com/karabakh/54744/armenia_karabakh_azerbaijan_ceasefire_violation_inflitration_attempt

[vii] http://www.freerepublic.com/focus/news/2424922/posts

[viii] http://www.channel5.com/shows/ww1s-tunnels-of-death-the-big-dig/episodes/earthquaking-the-ridge

[ix] http://en.wikipedia.org/wiki/North_Korean_infiltration_tunnels#Incursion_tunnels

[x] http://fortunascorner.com/2013/04/22/north-korea-may-have-13-undiscovered-invasion-tunnels/

[xi] http://www.cnn.com/2013/10/31/us/mexico-drug-tunnel/index.html

[xii] http://www.cnn.com/2010/CRIME/11/03/san.diego.border.tunnel/

[xiii] http://uk.reuters.com/article/2012/09/11/uk-azerbaijan-armenia-conflict-military-idUKBRE88A0F720120911

[xiv] http://www.usatoday.com/story/news/world/2014/01/14/drug-tunnels-us-mexico-border/4478073/

[xv] http://en.wikipedia.org/wiki/Battle_of_Kelbajar

[xvi] http://4.bp.blogspot.com/-Phs_pAuJtv0/UGw7AvET8hI/AAAAAAAABPs/Oa4ofXf_1KU/s1600/16.jpg

[xvii] http://hetq.am/eng/multimedia/slideshows/59

[xviii] http://en.wikipedia.org/wiki/Third_Tunnel_of_Aggression

[xix] http://blogs.telegraph.co.uk/news/malcolmmoore/9902611/Inside_North_Koreas_Third_Tunnel_of_Aggression/

[xx] http://hetq.am/eng/news/15312/azerbaijani-press-review-azerbaijani-army-needs-10-days-to-liberate-occupied-lands.html

[xxi] http://biweekly.ada.edu.az/vol_3_no_12/Newly_approved_military_doctrine_Goble.htm

[xxii] http://asbarez.com/78940/assessing-azerbaijan%E2%80%99s-military-threat-to-retake-karabakh/

[xxiii] http://www.armenianweekly.com/2014/01/05/artsakhs-prospects-for-the-future/

[xxiv] http://smallwarsjournal.com/blog/journal/docs-temp/647-thiel.pdf

[xxv] http://www.globalsecurity.org/military/world/azerbaijan/intro.htm

[xxvi] http://en.wikipedia.org/wiki/Armenian_military

[xxvii] http://www.azatutyun.am/content/article/24981715.html

[xxviii] http://www.armenianow.com/news/9132/offense_taken_aliyev_talks_tough_i

[xxix] http://www.azatutyun.am/content/article/24335604.html

[xxx] http://www.armenianweekly.com/2014/01/05/artsakhs-prospects-for-the-future/

[xxxi] http://www.rferl.org/content/azerbaijan-karabakh-soldier-killed/25235874.html

[xxxii] http://asbarez.com/120184/armenian-villager-captured-by-azeris-returns-home/

[xxxiii] http://www.armradio.am/en/2014/03/08/another-armenian-villager-captured-by-azeris/

[xxxiv] http://news.am/eng/news/114859.html

[xxxv] http://www.usatoday.com/story/news/world/2014/01/14/drug-tunnels-us-mexico-border/4478073/

[xxxvi] http://en.wikipedia.org/wiki/Third_Tunnel_of_Aggression

[xxxvii] http://en.wikipedia.org/wiki/Korean_Demilitarized_Zone#Incursion_tunnels

[xxxviii] http://www.lragir.am/index/eng/0/right/wv5k/31865

[xxxix] http://asbarez.com/72188/armenia-arrests-two-on-charges-of-spying-for-azerbaijan/

[xl] http://en.wikipedia.org/wiki/Reflection_seismology

[xli] http://ncpa.olemiss.edu/seismic-tunnel-detection/

[xlii] http://www.geophysical.com/militarysecurity.htm

[xliii] http://en.wikipedia.org/wiki/Ground-penetrating_radar

[xliv] http://en.wikipedia.org/wiki/Korean_DMZ#Incursion_tunnels

[xlv] http://www.entechworld.com/services/subterranean_tunnel_detection#tab-1

[xlvi] http://www.insightcrime.org/news-analysis/tunnel-detection-technology-at-mexico-border-worth-the-effort


3 thoughts on “Underground Intrigue: A Conjectural Assessment of Incursion Tunnels in a Future Nagorno-Karabakh Conflict

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s