Unobserved Fires and the Law of Armed Conflict


| Mar 3, 2023

Unobserved fires

The United States and Australia have both embraced the concept of “multi-domain strike,” which the United States describes, in a veritable missile salvo of buzzwords, as the “capability to strike in depth with lethal and non-lethal cross-domain effects … to creat[e] overmatch in operations against a peer adversary.” A critical component of this alleged revolution in warfighting includes precision long-range strike capability, otherwise known as the ability to attack targets from a long way away. The United States has developed doctrine and designed bespoke formations to undertake this mission, while Australia pursues “investment in long-range strike capabilities to maintain regional security, and deter or respond to aggression in the Indo-Pacific,” including collaborative development of the Precision Strike Missile (PrSM) and hypersonic weapons alongside the United States.

The authors of this post formed part of a multinational team of operations law attorneys, recently privileged to serve together on an Australian military exercise testing concepts for archipelagic large-scale combat. This exercise, and others like it, have tested the “ability to see, sense, communicate, shoot, and move at speed and scale, connecting all sensors with the best shooter and the right [command] node.” In this post, we interrogate what the law of war says about modern long-range fires, and explore the level of legal interoperability that undergirds our bourgeoning regional security partnership.

Indirect Fires and Historical Examples

Long-range fires confer the ability to strike targets in the enemy’s “deep area,” engaging at ranges previous generations of artillerists (and even early air power enthusiasts) would find ludicrous. The development of indirect fire – striking targets not in direct line of sight – is usually credited to the Napoleonic Wars. Later, artillery was the most prolific killer on the battlefield in both World Wars. The United States and Australia regularly employed unobserved fires during the Vietnam War. Artillery was fired for the purposes of “harassment and interdiction,” firing at locations to deny them to the enemy, and specified strike zones (commonly referred to as “free fire zones”) came to be a notorious feature of that conflict.

During the invasion of Iraq in 2003, the U.S. military prohibited unobserved fires unless for “immediate defense of friendly forces.” While no available unclassified source speaks directly to observation requirements for American or Australian indirect fires used in Afghanistan, the NATO International Security Assistance Force imposed a requirement that “[p]rior to the use of fires, the commander approving the strike must determine that no civilians are present,” a standard that seems to suggest a requirement for some form of observation of a target. One can also infer such a requirement from the investigation into an erroneous 2013 U.S. strike on a medical facility, which faulted the ground force commander because, among other errors, “[h]e could not observe any fires on the [hospital] Compound.”

Recently, Russian forces have been criticized for striking civilians during their unlawful invasion of Ukraine through firing blindly, as well as in situations where they were “tracking or ‘observing’ where they were landing, and [thus] Russian forces had information they could use to adjust targeting.”

Observed Fires in Doctrine

Fires doctrine tends to focus on the mechanics and procedures of observation, but is largely silent on the purposes (legal or technical) for observing fires, treating the reasons for observation as axiomatic. Australian doctrine defines indirect fires as “fire directed at a target that cannot be seen by the weapon system operator” (ADFP 3.1.1, para 5.6). While U.S. joint doctrine does not define indirect fires (referring instead to surface-to-surface capabilities), the U.S. Army includes “cannon, rocket, and missile systems as well as mortars organic to maneuver elements” under the banner of surface-to-surface indirect fires.

Australia does not define observed fire in doctrine, but mentions the term as an example of a restriction on the employment of fires (ADFP 3.1.1 Joint Fires and Effects Procedures, annex 2c, para 5). This obviously suggests that the concepts of observed and unobserved fires exist, even if not explicitly defined. U.S. Army doctrine defines observed fire as having a “point of impact or burst that can be seen by an observer, and can be adjusted on the basis of observation,” while helpfully explaining that “unobserved fire are projectile points of impact or burst not observed.” The U.S. Army Operational Law Handbook echoes ADF doctrine by discussing limitations on unobserved fires as an example of potentially restrictive rules of engagement.

Observation: In the Eye of the Beholder

Australian doctrine defines an observer as “[a]ny ADF or multinational personnel [who] request joint fires using the appropriate format.” (ADFP 3.1.1, para 3.41). In the context of operational assessment, ADF doctrine admits that observation need not always be visual, noting that “observation of secondary predicted effects (e.g. degradation of Internet protocol-enabled chat or slowdown of resupply) is required” to assess the performance and effectiveness of joint fires when direct, visual observation is not practical. ADDP 3.1 Joint Fires and Effects, para 6.7a. Although it does not directly refer to observation, ADF doctrine on the “track” stage of joint targeting acknowledges that “[t]o maintain continuous monitoring of the target, sensors must be directed onto the target” in order to assess “suitability for engagement” (ADDP 3.1, para 5.6). A further component of Australian targeting processes includes issuing targeting directives to, among other things, outline procedural steps and methodologies that enable compliance with the laws of war. One of those steps includes to “locate and observe the potential target” (ADDP 06.4, para 5.62 (emphasis added)).

In contrast, U.S. doctrine takes a clear position that observation need not be visual. The Army’s recent operations doctrine envisions a future battlefield that features “constant observation and the threat’s ability to gain and maintain contact in all domains, wherever they are located.” The U.S. Joint Publication covering fire support authorizes “observation and analysis of target characteristics, including visual recognition; [electronic support] systems; non-cooperative target recognition techniques; identification, friend or foe systems; other physics-based identification techniques; or human identity-based biometric data collection devices.” U.S. Army doctrine explicitly allows that an “observer can call for and adjust fire utilizing an Unmanned Aircraft System (UAS) … with a remote viewing system,” and notes that sometimes an “observer may be able to make a spotting even though the round impact may not be able to be seen,” such as when the impact of a round can be heard. Thus, both nations have implicitly endorsed the idea that observation, at least in some situations, need not be strictly visual, albeit to varying degrees and in different terms.

Watching the Watchers: Observation in ROE and Policy

The Newport and San Remo model rules of engagement (ROE) define observed indirect fire as “indirect fire for which the point of impact or burst can be seen by an observer. The fire can be controlled and adjusted on the basis of observation.” Even in these model policy documents, underpinned by compliance with LOAC, observation is discussed in terms of technical capability, as opposed to the capability to comport with legal obligations. Interestingly, both handbooks recommend that observed fire should be “permitted unless restricted by a rule,” but that “[u]nobserved indirect fire is not permitted unless authorized by a rule.” Thus, the scholars and practitioners behind the premier unclassified resource on ROE seem to evince a policy preference for observed indirect fires.

During our exercise, we determined that predispositions toward observed fires, coupled with narrow views of what constitutes observation, may generate challenges in operational environments. This is likely to require policymakers and operational commanders to choose between authorizing unobserved fires, or interpreting observation to incorporate non-traditional sensors, such as electronic emissions, radio traffic analysis, or synthetic aperture radar. The longer the range of weapons systems and the more contested the battlespace, the more salient this issue is likely to become. Whether unobserved or non-traditionally observed, fires must still comply with the law of war.

LOAC Targeting Regime

Assuming they are employed to attack the enemy, fires implicate the full LOAC targeting regime, in particular the rules governing distinction between lawful targets and civilian objects; obligations to take feasible precautions during an attack; and the principle of proportionality. The United States and Australia agree on the applicability of the LOAC targeting regime to long-range strike, but differing international obligations, especially Australia’s ratification of the 1977 Additional Protocol I (AP I) to the 1949 Geneva Conventions, and divergent national interpretations of customary international law may create legal friction during multinational operations.


Distinction, the practice of restricting attacks to lawful military objectives, is a bedrock principle of the LOAC. The Australian view of distinction is influenced by AP I, and expressly includes “a requirement … on all parties to a conflict to establish and maintain the distinction” (emphasis added). The United States Law of War Manual notes that “persons using force must discriminate between legitimate and illegitimate objects of attack in good faith based on the information available to them at the time” (emphasis added). Australia’s reservation to articles 51-58 of AP I, speaks in very similar terms by articulating the understanding that commanders would “necessarily have to reach their decisions on the basis of their assessment of the information from all sources, which is available to them at the relevant time” (emphasis added).

Both nations accept that target identification must be established to a customary international law standard, articulated in the commentary to AP I as being a standard of “common sense and good faith.” However, it is important not to conflate the requirement for target identification with a requirement for observation. Although Australia’s explicit requirement to maintain distinction may contribute to a policy preference for observed fires, it is far from settled law that maintaining distinction demands visual observation, versus merely requiring a good faith decision based on the information available at the time.

A threshold criticism of unobserved fires is that it is an indiscriminate means of warfare, but this ignores the fact that common sense and good faith decisions about a target’s identity can be made on the basis of information that is not derived from visual (or video) observation. For example, there are numerous pieces of solely military equipment (military objectives by nature) that emit a distinctive, detectable electromagnetic signature. There are numerous sensors that can acquire these signatures and multiple long-range weapons that “seek” those signatures with varying degrees of autonomy. The concept of linking “shooter to sensor” may be technically complex, but it is not legally significant – if the commander, through “information available to them at the time” from whatever source is able to identify a target as a valid military objective, in a common sense and good faith manner, the principle of distinction is satisfied.


Under the principle of proportionality and its supporting rules, attacks are prohibited if the expected loss of civilian life, injury to civilians, and damage to civilian objects incidental to the attack would be excessive in relation to the concrete and direct military advantage expected. Many practitioners reflexively assert that observation is required to allow commanders to make decisions about the proportionality of an attack, but that is not necessarily correct. This view is a mix of intuition and a hangover from recent state practice, in which observation (“sensor soak”) of targets prior to an attack was the norm.

But the law of war does not require visual observation to inform expectations of civilian harm. A commander can consider other factors to determine what harm is expected. For example, a commander could make a good faith determination that there would likely be minimal or no harm to civilians or civilian objects from an attack against an emitting enemy fire control radar located in a military cantonment, or counter-battery fire against enemy artillery located in the ruins of a city that intelligence reports say has been abandoned by the civilian population, even if the commander could not visually observe the impact of her fires.

Feasible Precautions

Australia adheres to the AP I, Article 57 injunction to take all feasible precautions in choice of means and methods of attack with a view to avoiding, and in any event to minimizing, incidental loss of civilian life, injury to civilians and damage to civilian objects (ADDP 3.14.2, Targeting Procedures, para 6.19c(1)). The United States has not ratified AP I, but recognizes a customary international law obligation to “take feasible precautions to reduce the risk of harm to the civilian population and other protected persons and objects” (U.S. Law of War Manual, § 5.2.3). Australia defines feasible precautions as those “practicable or practically possible taking into account all circumstances ruling at the time” (ADDP 06.4). The United States defines feasible precautions in the same terms, while also taking the position that “[t]he standard for what precautions must be taken is one of due regard or diligence, not an absolute requirement to do everything possible” (U.S. Law of War Manual, § Ultimately, these positions are reconcilable, if not identical. The United States is merely more explicit in the position that feasible means reasonable as opposed to possible.

Visual observation of indirect fires could be a precaution, although, as a matter of capability, the location of impact of many modern munitions cannot be altered after firing. As such, observing a munition’s actual point of impact is of limited utility as a precaution in attack, at least so far as a first munition is concerned, vice following munitions for which the aim point may be adjusted.

Even if visual observation (which is clearly most useful before weapons release) may assist in mitigating civilian harm, the feasibility of visual observation for long-range strike in all circumstances is debatable. The future likely augurs a move away from bespoke strike cells conducting boutique manhunting towards large-scale combat operations employing artificial intelligence and autonomous long-range weapons. This is particularly so as countries continue to seek and develop capabilities to hold adversary targets at risk from increasingly longer distances. Both improved technology and a conflict’s context will alter the feasibility of precautions. The luxury of “soaking” every target with visual surveillance is unlikely to be practicable in a State-on-State war for national survival where airspace is contested and bandwidth is constrained. This starkly illustrates Gail Curley and Paul Golden’s observation that “many relevant policies and the concrete [counterterrorism] experiences most service members are laden with are based on assumptions that do not exist in a highly contested environment, such as conflict with a peer or near peer enemy.”

National Divergence

Legal interoperability is defined as “achievement of shared understanding of respective authorities, permissions, restrictions, obligations, and interpretations of international and domestic law and policy that enables the Combined Force to act together lawfully, coherently, effectively and efficiently to achieve tactical, operational and strategic objectives.”

In the context of observed fires, the above analysis points to a divergence that needs to be clearly understood. Put plainly, the United States does not have a requirement for fires to be observed, and U.S. doctrine reflects comfort with non-visual means of observation. In contrast, Australian doctrine is substantially less clear, even if it does not expressly mandate observation.

There are, however, sufficient inferences to encourage a preference for observation as a default, which is exacerbated by a recent history that is replete with experience in counter-insurgency conflict in which visual or video observation was the norm. Assuming this predisposition is given effect through operational framework documents (such as rules of engagement and targeting directives), commanders may be impeded from taking decisive action in situations where moments and minutes (not tens of minutes or longer) matter. It would also mean that Australia’s approach to employing long-range fires could depart significantly from one of its closest allies. Such divergences in practice between coalition partners are not unheard of, or even uncommon. However, the wisdom of this divergence might reasonably be questioned in light of the availability of information through various means (other than visual or video) to enable commanders to comply with their law of war obligations while using long-range fires.


Both Australia and the United States recognize that an era of accelerating geopolitical competition has brought with it increasing focus on strategically important military capabilities such as multi-domain strike and long-range fires. Military professionals must be trained and ready to avoid potentially self-imposed limitations arising from the legacy of counter-insurgency conflicts – whether in doctrine, thinking, or practice. From a law of war perspective, long-range fires are subject to the same targeting regime as other forms of fire. While visual or video observation may have been a clear precaution of choice in counter-insurgency environments, its feasibility will always be circumstance dependant. Importantly, information sources of any description can be used to satisfy the requirements of distinction, proportionality, and feasible precautions.

A review of extant doctrine shows that the United States (which does not mandate an observation requirement) is well placed to adapt to situations in which visual or video observation is not feasible. While the situation is less clear for Australia, the lack of specificity in Australian doctrine (which refers to “observation” but does not define what that means) generates opportunity. It is possible for Australia to adopt a position that observation of fires need not be visual or video provided that the laws of war are satisfied through other means. Such an approach would enhance decision agility and contribute positively to interoperability while remaining aligned with the rules-based global order to which both countries ascribe.


Wing Commander Janine Fetchik is a legal officer in the Royal Australian Air Force. 

Major Matt Montazzoli is an active duty Army judge advocate, currently assigned as Operational Law Attorney (Military Personnel Exchange Program), 1st Division / Deployable Joint Force Headquarters, Australian Army at Gallipoli Barracks, Brisbane.



Photo credit: Australian Armed Forces