On August 15th, “The War Zone” published an article arguing that China was building a new airstrip on Triton Island in the South China Sea's Paracel Archipelago. This thesis was quickly picked up by major news sites and widely distributed. 

It was poor analysis then. It still is. Let us explain why.

Subsequent articles by one of the authors and the Asia Maritime Transparency Initiative cast immediate doubt on this theory, laying out the case that the structure in question was most likely an elevated roadway or tidal levy. 

Here at SeaLight, we then began a crowdsourced open-source analysis effort in conjunction with SkyFi to make sense of the evidence. Using SkyFi's commercial satellite imagery collected at intervals since the original article, our analysis has only further confirmed that the original article was wrong--there is almost certainly no runway under construction or even planned at Triton Island. 

(Want to contribute to our ongoing open-source analysis on Triton Island? Join our Discord discussion here.)

Why not an airfield?

It is highly unlikely that an airfield is being constructed on Triton Island. 

The earthwork originally misidentified as a runway is not aligned with the prevailing northeast winds at Triton, and is in fact oriented along the least prevalent wind axis (east-west). Moreover, when constrained by the recently built earthworks and construction sites, the current parts of Triton Island that are permanently dry cannot support a runway greater than 1,000 meters. 

Any future runway construction would likely occur on the eastern side of the island, oriented at a northeast-to-southwest axis. 

The eastern side the island possesses the largest amount of land area still capable of building a runway aligned with the prevailing northeasterly winds, should Beijing decide to eventually build one. 

Even this scenario seems very unlikely, however.

Indicators of new runway construction on Triton would include larger and more numerous dredging barges, expanded harbor and cement production facilities, and eventually the appearance of a long, straight earthwork aligned on a northeast-southwest axis.

None of these are in evidence.

The size and number of dredging vessels seen thus far are capable of only minor improvements vice large-scale reclamation, and the limited dredging to this point has only been used to add sand to parts of the island already above sea level.

With a small harbor lacking dedicated bulk cargo handling cranes and a temporary cement batch plant typical of much smaller construction jobs, projects wouldn't be of comparable scale to those undertaken at other PRC island-building projects over the past decade.

Even if Beijing did have plans for a runway correctly oriented in line with prevailing winds at Triton, several new facilities and the new levee have already been built across the site where such an airstrip would logically go. 

Want to know more? Our full analysis follows for all our OSINT (open-source intelligence) enthusiasts.

What is Triton Island?

Triton Island is the westernmost of the Paracel Islands, a disputed archipelago claimed by Vietnam, the People’s Republic of China (PRC), and Taiwan. The islands have been under de facto PRC control since a 1974 naval clash with South Vietnam, and since 2016 have seen a blitz of construction activity by the PRC, including land reclamation and the construction of new harbors, helipads, and other infrastructure. 

Woody Island, which lies approximately 100 miles northeast of Triton, hosts the only active airport in the archipelago and is the PRC's administrative center in the Paracels. 

An additional runway on Triton Island would be significant. It would add capabilities for the PRC to deploy additional bomber forces or fighters even closer to Vietnam, and to monitor southern approaches to the PLAN naval forces at the Yulin Naval Base on the island of Hainan. 

Understanding our analysis

Runway orientation

While the original earthwork constructed at Triton Island did appear in imagery as straight and long, our research of the prevailing wind conditions in the area indicates that this east-west oriented earthwork is not optimal for a runway. A properly oriented runway would be aligned in the same direction as the prevailing wind, minimizing crosswinds during takeoff and landing while providing aircraft extra lift when taking off into the wind.  

Using a “wind rose” diagram from nearby Woody Island (Figure 3), we can see that the prevailing winds in this area are out of the northeast in a typical northern hemisphere trade winds pattern prevalent at this latitude.

The wind rose for a straight east-west alignment shows very little wind from either direction, and any runway oriented on this axis would frequently subject aircraft to hazardous crosswinds. As comparison, the runway at Woody Island is well aligned at a “05/23” (050°/230°) northeast-southwest orientation that matches the local prevailing winds. 

Seawater intrusion

Based on the green marine growth pattern in Figure 6 (above), substantial amounts of seawater makes its way into the island at high tide. 

This seawater intrusion area was also detected by the National University of Singapore’s Center for International Law Satellite Research Project on Insular Geographic Features in the South China Sea. Using imagery from 2013 (Figure 7, below) this study delineated a large seawater intrusion area with a depth of up to 1 meter that comes from the north and covers the central portion of the island at high tide. 

This seawater intrusion area explains why this earthwork is oriented east/west and very likely represents Chinese efforts at tidal control along the island's southern edge.

Future construction options limited

The remaining dry land on which a runway aligned with prevailing winds could be constructed has been cut significantly due to the newly constructed earthworks and structures. Under the assumption that a NE-SW runway would cross neither newly constructed earthworks nor the area now obstructed by new structures, the longest runway Triton could support at this orientation is approximately 810 meters.  

Of course shallow areas around the island could be reclaimed to extend this area, similar to other Chinese bases built up around the South China Sea. Assuming such efforts could be supported by the island's geology, it appears that a properly oriented runway could be built with a maximum length of approximately 2,400 meters after reclamation. 

As shown in Figure 8 (below), this maximum runway length uses only undeveloped dry land (green) and potentially reclaimed land (yellow). This represents only the runway and does not include other required airfield features such as taxiways and aprons.  

(The full 2,700 meter runway from Woody Island is included in red for reference, but that extends beyond the potentially reclaimable area.) 

We chose the runway location for this analysis to maximize length while remaining outside of the exclusion areas shown. The first exclusion area, labeled “Base Area” in Figure 8, represents that which is protected by the newly constructed levees. Any runway within this area would have to contend with existing structures that would obstruct takeoff and landing, occupy valuable dry space for other facilities, and require that the runway travel through the recently constructed levees--thus compromising their purpose to prevent seawater intrusion. 

The second exclusion area, labeled “Seawater Intrusion”, represents the portion of the island that is currently subjected to seawater inundation at the highest tides. The water in this intrusion area would prevent normal runway operations. 

The resulting hypothetical runway lengths have been included in Figure 9 (below), comparing them to existing runways of other Chinese airfields in the South China Sea. 

All of this begs the question of why China would build its facilities across the diagonal path of its most logical runway siting if a runway was ever contemplated in first place.

Dredging activity

While the PRC has shown itself capable of the kind of large-scale land reclamation required to build a military runway, all analyzed imagery at Triton Island has shown the presence of only one dredger. At approximately 20 meters long, this dredger is much smaller than those used in the construction of its other islands with large runways. Moreover, it has only been observed sending sand to construction sites and earthworks on existing dry land, and it almost certainly lacks the capacity to reclaim significant amounts of land on its own.  

The size of this dredger stands in stark contrast to the ones seen at other PRC island reclamation projects, which featured several dredging vessels of 100 meters or larger operating at the same time.

Given that it is the only dredger currently observed at the island, it is likely that the amount of construction currently planned at Triton will only require a moderate amount of dredged material--certainly nowhere near enough to support land reclamation for a military runway.

Construction and infrastructure

Cement production facilities and port infrastructure at Triton Island are significantly less robust when compared to that on Subi Reef during reclamation efforts there in 2015. The port at Triton Island does not possess any dedicated bulk cargo cranes and is likely relying on excavators and other self-propelled construction equipment to unload the imported aggregates required for on-site cement production. 

Furthermore, the navigation channel to the harbor measures approximately 33 meters wide, limiting the size of ships that can access the harbor. While the depth of the channel is unknown, it is almost certainly another limiting factor.

The cement batch plant on the island can be identified by its proximity to piles of aggregate material used in the production of cement stockpiled nearby, as well as observed shadows that reveal the presence of vertical objects typical of batch plant towers (Figure 15). The area occupied by the batch plant measures just over 2,100 square meters with a dedicated aggregate pile storage area of 3,600 square meters, totaling 5,700 square feet of the island dedicated towards cement production.  

This number represents just 20% of the 27,500 square meters that was dedicated towards cement production across two cement batch plants at Subi Reef during April 2015. Additionally, during the Subi construction efforts we observed the presence of dedicated bulk cargo unloading cranes and deep draft bulk carriers. The lack of similarly robust construction capacity at Triton Island indicates a relatively small scale project.

Analysis of alternatives

Alternate exclusion zones

One of the assumptions we made when determining possible runway locations was that any potential runway construction would not extend through any recently constructed levees. This was based on the idea that existing and future structures would block the operation of aircraft and compromise the ability of the levees to prevent seawater intrusion from inundating the new facilities. 

Since this area was excluded from our original analysis, we reanalyzed the potential runway locations with this constraint removed in order to find the longest possible runway location considering only the prevailing winds and lack of existing permanent structures.

Under these new constraints, we identified a new maximum-length runway location of just over 3,600 meters with land reclamation and just under 1,000 meters without. The first figure assumes the potential for future reclamation beyond the cement production and dredging capacity currently observed.  While these potential sites are longer than that found in the original analysis, each one possesses significant drawbacks.  

The 998-meter runway that could be constructed without reclamation is located hazardously close to relatively tall new construction. 

Moreover, a new structure (besides the levee) has already been built in a location that significantly curtails what would be the longest possible properly aligned runway location on unreclaimed land (“Runway Max #1” + “Additional Length to Max #1 - Blocked” in Figure 18). 

The fact that the PRC chose to place this structure and the levee through this potential runway site further indicates that they have no such plans.

The other possibility we considered in this alternative analysis represents the longest possible runway location with the levee exclusion zone constraint removed. At 3,606 meters long it is the longest runway that could still be built when accounting for existing construction sites and including potential land reclamation. 

Any attempt to build this runway would require reclamation of at least 1,700 meters, for which we see no indication in the near or even mid term, given the lack of local dredging or cement production capacity. 

Finally, as shown above in Figure 17 such a runway would also come hazardously close to an existing construction site. 

Wind constraint removed

Another alternate analysis involved the removal of the constraint of an optimally oriented ("05/23") runway aligned with the prevailing winds. However, not all runways on islands in the South China Sea are optimally aligned due to space and construction constraints. 

For example, as shown in Figure 19 (below), while the same NE-SW winds at Triton also prevail at Pag-asa (Thitu) Island, the 1,300 meter long runway there is constructed along an east-west axis. The length of this runway and its alignment relative to winds limits it to STOL (short takeoff and landing) aircraft and may be limited by weather conditions.

Removing the wind constraint on Triton Island and using the same alternate exclusion area constraints in order to obtain the longest runway possible, our analysis yielded a 1,248 meter runway oriented at 035°/215° degrees. This is coincidentally only 15° off of the 050°/230° oriented runway from Woody Island and nearly aligned with the prevailing winds. 

However, this length is still shorter than the 1,257 meter long runway site that is perfectly aligned with the prevailing winds that also went through the levees but was blocked by new construction. All this makes it unlikely that this site would have been selected when a better site nearby is now blocked by structures that could have been sited elsewhere.

Intelligence gaps

Supporting infrastructure required for an airfield

The overall footprint of additional facilities required to support a military airfield (hangars, aprons, etc.) was not used in this location analysis. We restricted our analysis to the runway alone in order to deal directly with the hypothesis presented by the original article in "The War Zone".

Elevation data

While our analysis was able to generally determine where low lying areas of the island were from the visible signs of seawater intrusion and from shadows cast earthworks construction, we do not possess high quality elevation data for a more exhaustive study. Even so, the island appears generally flat based on ground and aerial photos and we assumed it to be such for this analysis.

Suitability of island geology

While China has undertaken several land reclamation projects in this area, we do not know the suitability of Triton Island’s geology and environment for the type of large-scale land reclamation efforts seen at these other locations. For this analysis we assumed that the type of reclamation seen elsewhere would be constrained only by the logistical capacity for such an effort as observed in imagery.

Summary of findings

When all the possible runway construction sites are analyzed and compared against runways from other PRC airfields in the South China Sea (Figure 21), it becomes clear that any potential construction on Triton would require a large-scale land reclamation effort. Based on the observed dredging and cement production capacity alone, it's unlikely that China is planning to do any reclamation at at this time.

Based on its orientation to the prevailing winds, its length and its construction in an observed seawater intrusion area, the “runway” originally identified by “The War Zone” is almost certainly not a runway and is very likely a tidal levee meant to prevent seawater inundating the area south of the levee at high tide. 

Furthermore, as shown in Figure 22 (below), any runway construction on Triton Island would likely occur on the eastern side of the island, oriented at a NE-SW axis. Any future observed work in this area along this orientation would be a significant positive indicator towards the hypothesis of runway construction.

However, lack of remaining sites that could realistically support a runway--while considering existing structures and logistical constraints--indicates that China probably has no plans to build an airfield on Triton. 

While the potential technically still exists for a 1,250 meter long runway that would be comparable to the other shorter runways found in the South China Sea, this would require building a runway though a newly created levee. Any other possible runway would either be significantly shorter than any other South China Sea airfield or require major reclamation efforts not yet in evidence.