The Bohu Laser Facility, Part 2: Operations

Note, this is the second of 2 posts on this subject. You can find part 1 here.


In part 1, we presented new research regarding the Bohu laser facility. That research focused on the history and organisation of the site, demonstrating that it had been operating for longer than previously believed, and that it was subordinated to the Space Systems Department of the Strategic Support Force. This section will discuss the operations at Bohu, the main focus of our research. We present our hypothesis for how the site conducts laser operations, as well as demonstrate the steep rise in such operations over time. Finally, we’ll show how the Defense Intelligence Agency (DIA) has changed the language in their public reporting in a way that is correlated with our findings.

The Bohu laser site in 2022. Image courtesy of Planet, annotated in Google Earth.

Operational Hypothesis

Laser operations can be broadly classified into 3 categories, ranging, dazzling, and blinding. Ranging is the use of a low-powered laser to calculate a satellite’s orbit for tracking purposes. Dazzling is the use of a laser to temporarily “blind” a satellite’s optics, but without causing permanent damage. That means that when the satellite passes the laser, normal operations resume. Finally, blinding is the act of permanently damaging or destroying a satellite’s sensor with a laser. We believe Bohu uses two different systems for laser operations.

The lasers in Buildings A and C are mounted on pivoting platforms, located at the southern end of the building. This allows them to “fire” over the northern wall when the roof is open, and gives them an approximately 270-degree field-of-view. The roofs appear to have the ability to open and close relatively quickly; Building D was seen transitioning from closed to open in 4 minutes, while C did the same in 10 minutes. It’s likely that C can close faster than this, and that A has the same capability. We believe that these lasers are being used for ranging operations. Laser ranging is not in itself a concerning activity, the United States conducts similar operations at the Starfire Optical Range in New Mexico. At the same time, we note that Starfire was considered as a testbed for anti-satellite weapons.[1] However, we believe that China is utilising a vehicle-mounted laser system, or seeks to develop one, for more destructive types of laser operations. We base this on two assumptions:

  • The presence of unusual vehicles at the Bohu test site, whose configuration resembles a probable laser vehicle.
  • The utility of a mobile laser system.

The 2013 paper that described the laser test specified a vehicle-mounted system. The Bohu site has large concrete pads co-located near Buildings A and B, and each pad appears to have an underground cable running to it from the buildings. On the pads, unique trucks have been spotted. These trucks are significantly longer than others at the site and have an unusual white mounting in the centre. We’re not aware of similarly configured vehicles being seen elsewhere in China. We strongly believe that this white mounting represents a mobile laser system, and that it’s this element of the anti-satellite complex that may have a nascent destructive capacity.

A vehicle-mounted anti-satellite system makes considerable sense from a usability perspective, and would be in line with comparable systems like the Russian Peresvet.[2] Having to construct numerous fixed laser sites near strategic points, along with their infrastructure, is less valuable than being able to create a mobile anti-satellite system. It’s also unlikely that this vehicle is purely for blinding satellites, it almost certainly could scale its output (and therefore its effects) to engage in dazzling.

It’s important to stress that this is a hypothesis. There’s a lot we don’t know about Bohu. For instance, we don’t know why laser ranging operations take place when they do and it’s possible that we’ll never have a conclusive answer; imagery limitations and the building’s ability to open and shut on short notice makes intra-day collection challenging. A second question relates to the trucks. If these vehicles do represent a new anti-satellite weapon, why are they out in the open? We don’t have an answer in this regard. One possibility is that they’re being used to dazzle reconnaissance satellites, but there’s no specific evidence to support this.

The anomalous trucks on pads north and south of Building C. A subterranean cable is visible leading to the north pad, and one of the laser systems can be seen in Building C. Image courtesy of Maxar.


The discovery that satellites could be used to track roof movement at Bohu relied on freely available Sentinel-2 imagery, including what would eventually amount to ~600 days of coverage over a 5-year period. Following a request, Planet Labs provided additional imagery. From this, a database was created that included, (at the end of H1 2022) over 1700 days of imagery[3] from July 2015 to the present. Each image was coded based on whether buildings were open or not. This is a subjective exercise, but there are easily spotted differences with all buildings which reduce the room for error. In more recent Planet coverage, there are periods of time where multiple images captured at different times of the day will show roofs changing their state from open to closed, or vice versa. In these cases, roofs have been coded based on whether they were active at any point that day.


Imagery shows that activity at the 3 laser buildings (A, B, and D) has steadily risen, albeit at different rates, culminating in a near constant active state since late 2020. Importantly, this has occurred independently from the rise in data availability, making it more likely that this is a development occurring over the last 2 years, as opposed to something that previously noticed solely due to a lack of imagery. Even when coverage became more consistent (e.g 1H 2020), roof movements for Buildings A and C didn’t increase. It wasn’t until 2H 2020 that the number of movements rose.

Time PeriodNumber of Days with Images
2H 201520
1H 201629
2H 201657
1H 201762
2H 2017137
1H 2018155
2H 2018150
1H 2019146
2H 2019150
1H 2020158
2H 2020156
1H 2021165
2H 2021165
1H 2022167

Table A: Number of days with images per 6 months since 2H 2015.

TimeDays with Usable Imagery
2H 201513
1H 201622
2H 201638
1H 201748
2H 2017116
1H 2018117
2H 2018131
1H 2019123
2H 2019118
1H 2020107
2H 2020126
1H 2021126
2H 2021134
1H 2022135

Table B: Number of non-weather affected days with images per 6 months since 2H 2015

While it’s true that data from 2H 2015 through to the end of 1H 2017 is sparse, the data from 2H 2017 onwards is consistent enough in volume to be a usable basis for analysis. 2H 2015 – 1H 2017 has been included for context.

Time PeriodBuilding ABuilding CBuilding D
2H 2015171
1H 2016001
2H 20162017
1H 20170010
2H 201714617
1H 2018002
2H 20180037
1H 20193358
2H 2019142060
1H 20204536
2H 2020536498
1H 202174114112
2H 2021105118105
1H 2022125128125

Table C: Roof movements of each building per 6 months from 2H 2015.

As Table C shows, there was a major shift in roof movements beginning in 2H 2020. What’s particularly notable is the change at Buildings A and C; previous peaks had been the result of activity that appears to have been seasonal in nature.[4] It’s important to recognise the shift in operations from a primarily seasonal model to a sustained upwards trend. 90% of Building A’s movement occurred in or after 2H 2020, the equivalent figure for C is 91%. Even Building D, which was the most active of the 3 prior to 2H 2020, has seen 65% of its activity in or after that date.

Graphic A: The number of days where roofs at each building have been observed as open, compared with the number of usable images.
Graphic B: The number of days where roofs at each building have been observed as open relative to the number of days with usable imagery.

Two periods, 2H 2015 and 2H 2016, present anomalies. Building C in 2H 2015 is a case of the lack of data distorting the results, and doesn’t threaten the overall thesis. Data collection in 2H 2015, as Tables B and C demonstrate, were limited. Building D in 2H 2016 is more challenging. There’s no obvious explanation for the sudden spike in activity, apart from the possibility that this was simply an extreme seasonal jump. However, the fact that this spike was contained to the 2nd half of 2016, as well as to Building D, suggests that it was caused by factors specific to that building.

It is worth noting that since the end of the 1H 2022 reporting period, Bohu operations underwent another shift. From late-July until the 30th of October, no movement at Building D was detected. However, we do not believe this threatens the overall thesis, as movement at Buildings A and C remained stable over the period. Additionally, roof movements at D have now resumed, strengthening our conviction that it does not threaten the overall thesis.

DIA’s Changing Language

DIA produces a semi-regular report, Challenges to Security in Space.[5] Since 2019, DIA has changed its language relating to China’s directed energy programs. The 2019 report assessed that it was likely China would have a laser by 2020 at the latest. For reference, the precise quote is contained below.

“China likely is pursuing laser weapons to disrupt, degrade, or damage satellites and their sensors and possibly already has a limited capability to employ laser systems against satellite sensors. China likely will field a ground-based laser weapon that can counter low-orbit space-based sensors by 2020, and by the mid-to-late 2020s, it may field higher power systems that extend the threat to the structures of non-optical satellites.”[6]

To be clear, this does not say that China will field a ground-based laser weapon in 2020, merely that it was likely to. It stated that it was possible, but unconfirmed, that a “limited capability” had been achieved. This year, the Security in Space report received an update, using information up to March 2022.[7] In that update, DIA felt it necessary to change its language. The relevant quote is contained below (emphasis added).

“During the past two decades, Chinese defense research has proposed the development of several reversible and nonreversible counterspace DEWs for reversible dazzling of electro-optical sensors and even potentially destroying satellite components. China has multiple ground-based laser weapons of varying power levels to disrupt, degrade, or damage satellites that include a current limited capability to employ laser systems against satellite sensors. By the mid- to late-2020s, China may field higher power systems that extend the threat to the structures of nonoptical satellites.”[8]

China now, according to the DIA, is in possession of multiple weapons of “varying power levels” to “disrupt, degrade, or damage satellites.” In addition, DIA has no longer hedged its language by saying that China “possibly” has a limited capacity to employ laser systems; these qualifiers were removed in the 2022 report. Interestingly, though perhaps unsurprisingly, the DIA report does provide a citation for its claim. Specifically, it cites what appears to have been a speech delivered by John Ratcliffe at the eighth meeting of the National Space Council. However, his remarks as released publicly do not contain any specific mention of laser developments in China. Notably, the 2022 China Military Power Report (CMPR) changed its assessment on laser technology to align itself with the earlier DIA assessment, reproducing the above excerpt in full.[9] This comes after the 2020 and 2021 CMPRs used identical language to describe the laser program and sought to describe it as developmental only.[10]


Significant unknowns surrounding Bohu remain. While Planet’s ability to take numerous medium-resolution images over a single day has been useful, the passes are still limited to a narrow time period. Secondly, the effectiveness of an anti-satellite laser remains unknown. However, the DIA report suggests that China hasn’t demonstrated a destructive capability but does possess the capacity for it.

Perhaps most importantly, we have little context in which to place this information. The role of anti-satellite lasers as opposed to ASAT technologies is unclear. China may find that alternative technologies are more cost-effective, or that lasers aren’t achieving the desired performance, calling into question future development. With this in mind, we interpret “limited capability” as meaning a capability that is operational from a technical perspective but isn’t in widespread use. Finally, it’s not certain that Bohu represents that “limited capability” (though open-source reporting hasn’t revealed a more likely candidate). The increase in activity correlates with the DIA’s estimates, but it doesn’t confirm it. There’s a possibility that the data is real and the interpretation incorrect.

Conclusion and References

The Bohu directed-energy site has seen a significant rise in activity since 2020, at the same time as the DIA has assessed that China has achieved a “limited capability” in anti-satellite directed-energy technology. Bohu is believed to represent at least one element of China’s program for developing anti-satellite capabilities, and the most likely element to which the DIA is referring.

[1] William J. Broad, ‘Administration Researches Laser Weapon’, The New York Times, 3 May 2006,

[2] Bart Hendrickx, ‘Peresvet: A Russian Mobile Laser System to Dazzle Enemy Satellites’, The Space Review, 15 June 2020,

[3] This figure includes days where imagery was not available due to inclement weather. 21% of imagery was affected by inclement weather, making the number of usable images 1,354 over 5 years.

[4] Before 2H 2020, 81.5% of A’s and 80% of C’s roof movements had occurred in the 2nd half of the year.

[5] Defense Intelligence Agency, ‘Challenges to Security in Space (2019)’ (Defense Intelligence Agency, 2019),

[6] Defense Intelligence Agency, ‘Security in Space (2019)’, 20.

[7] Defense Intelligence Agency, ‘Challenges to Security in Space (2022)’ (Defense Intelligence Agency, 2022),

[8] Defense Intelligence Agency, ‘Security in Space (2022)’, 18.

[9] Department of Defense, ‘2022 Report on Military and Security Developments Involving the People’s Republic of China’, 93, November 2022,

[10] See Department of Defense, ‘2020 Report on Military and Security Developments Involving the People’s Republic of China’, pp. 65, September 2020, Additionally, see Department of Defense, ‘2021 Report on Military and Security Developments Involving the People’s Republic of China’, 65, November 2021,


  1. P (History)

    Thank you for sharing your analysis here!

    I’m afraid I see very little I can usefully add. But I’m going to throw some comments here, in the hope you’ll find some of them useful.

    – In the first part, you announced you were going to call this facility “Bohu”. In the second part, you sometimes call it “Bohu”, sometimes “Korla”. I’d recommend calling the same way everywhere; but that’s clearly a minor nitpick.

    – The symbol on the triangular lawn at the entrance appears to be a plasma ball.

    – It’s funny to see how people walk from and to some buildings through the grass. Mainly from and to Buildings E, F and H on Google maps, but also to other buildings in your more recent image. These are probably just footpaths: some look too narrow for vehicles, and one of them even crosses what looks like ground-level piping (on Google maps again). Maybe the evolution of these footpaths can be used as a proxy for the activity here?

    – I don’t know what building O is for, but there is no true road going to it, only footpaths and a preexisting dirt road. It looks unimportant.

    – I didn’t figure out where this facility was getting its electricity from. But maybe I just missed it on Google maps.

    – The constructions immediately to the east-northeast in Google maps leave me perplexed: some kind of helipads or airship servicing facilities?