Assessing post-earthquake patterns of violence in Myanmar

To facilitate rescue and recovery in the wake of the devastating earthquake that struck the country on 28 March, Myanmar’s National Unity Government (NUG), an opposition body formed after the 2021 coup, announced a temporary pause of its offensive military activities. This was followed by unilateral ceasefire declarations from other key conflict actors, including the Tatmadaw and a powerful bloc of ethnic armed organisations (EAOs).

But the ceasefires proved to be largely performative. The Tatmadaw continued with a ferocious air campaign against a wide range of targets, inflicting massive harm on civilians. It also pressed ahead with several counter-offensives, including in earthquake-affected areas. The regime’s violence, especially against civilians, provoked an enormous domestic and international backlash and obstructed any prospect of de-escalation.

Despite declaring unilateral ceasefires, groups such as the Arakan Army (AA) and Kachin Independence Army (KIA) did not in fact pause their offensives against strategic targets as the Tatmadaw scrambled to restore its command and control, and the full might of its air force, following the quake. Meanwhile, the NUG’s closest allies – including the Karen National Liberation Army (KNLA), Chin National Front (CNF) and Karenni Nationalities Defence Force (KNDF) – did not announce ceasefires, openly opting to fight on instead.

Airstrikes and violence against civilians before the quake

The Tatmadaw’s strength has steadily declined since the Brotherhood Alliance – which comprises the Myanmar National Democratic Alliance Army (MNDAA), Ta’ang National Liberation Army (TNLA) and AA – launched Operation 1027 against it in October 2023. In addition to significant losses of territory, especially in the borderlands, the army has suffered severe casualties to the point where, by mid-2024, it could no longer field its mobile strike divisions to launch meaningful counter-offensives. 1 The army’s depleted strength has affected the way it uses both airpower and other violence against civilians. An understanding of the dominant dynamics before the earthquake provides context for the violence that followed 28 March.

Until late 2023 the Tatmadaw had used airpower for three main purposes: in defence of embattled positions, as close air support to ground units carrying out offensive operations, or as a punitive response to opposition attacks and advances. The latter typically entailed intelligence-driven strikes on sensitive opposition assets such as headquarters, revenue sources, offices, or leaders of the unit the Tatmadaw deemed responsible. Although strikes on these targets often took place outside the immediate combat zone, they were directly linked to events along the frontline. During the initial phase of the war, civilian casualties caused by airstrikes were primarily collateral in nature, with ground forces responsible for the majority of direct attacks on civilians. Nonetheless, airstrikes caused the most severe incidents involving civilian casualties, such as the 2023 attack on a People’s Defence Force (PDF) command post and administrative office in Pa Zi Gyi that killed 165 people.

The Tatmadaw has a long and well-documented history of using violence against civilians as a counter-insurgency tactic, and this violence has followed a clear pattern. The attacks on civilians have almost always been a direct response to opposition activities undertaken in contested areas, with the intention of creating a feedback loop whereby civilians associate opposition activity, or even the mere presence of insurgents, with their own suffering. The tactic aims to force local populations to withdraw support from armed groups and pressure them to avoid confrontation or take up position elsewhere.

As Figure 1 shows, the nationwide total of armed clashes per month has continued to decrease, while the number of strikes by aircraft or drones has increased. This is because the Tatmadaw has been using air attacks in place of operations that would typically have involved infantry or artillery units, either in isolation or in tandem with air support. This partial decoupling of ground fighting from remote attacks carried out by aircraft, uninhabited aerial vehicles (UAVs) or, more recently, motorised paragliders is largely a result of the Tatmadaw’s diminished territorial control and personnel losses.

This same pattern is visible in the data for attacks against civilians. Until recently, most events involving Tatmadaw violence against civilians have been attributable to infantry or artillery. For example, in 2022 and 2023, Tatmadaw units frequently carried out raids across the central Dry Zone that included widespread killings of civilians (massacres in numerous instances) and the burning of villages. In the borderlands, indiscriminate artillery strikes on villages where EAOs operated were the most common form of collective punishment. As Figure 2 shows, remote attacks by aircraft and UAVs emerged as the dominant mode of Tatmadaw violence against civilians in the months prior to the earthquake. The overall monthly totals of Tatmadaw attacks on civilians, however, remain lower than during the peaks recorded in 2022 and 2023, although the data does not convey the severity of the incidents.

Attacks on civilians can vary in nature and reveal different conflict dynamics. Following the first and second parts of Operation 1027, EAOs had greater control of the country’s borderlands, where they moved to establish or expand their administrations. While the Tatmadaw’s losses drove an increase in punitive airstrikes, they also prompted a new pattern. Aiming to prevent a return to normality or the consolidation of opposition governance and control in areas beyond the reach of its infantry, the Tatmadaw stepped up airstrikes against any infrastructure or location connected with the opposition’s administrative capabilities or its provision of essential services, including offices, medical clinics and markets.

This aggression has contributed to the growing number of air or drone strikes against civilians, as classified by ACLED. However, there are some key distinctions that open-source data cannot reliably capture. In cases where airstrikes are classified as targeting civilians but take place far from the frontlines, it is probable that the Tatmadaw is targeting specific elements of the opposition and their capacity for governance and service provision. 2 Such attacks often hit individual buildings or locations, which indicates the deployment of precision-guided munitions and intelligence-based targeting – assets typically reserved for use against armed opponents. However, these attacks frequently target civilian members of opposition groups as well as administrative or service centres that ordinary civilians rely on, and the Tatmadaw makes no serious effort to avoid excessive collateral damage. Therefore, in airstrikes far from the frontlines, it is still the case that civilians pay the highest price in terms of casualties, disruption to daily life, and psychological trauma.

A different dynamic is apparent in contested areas much closer to the frontlines. There, airstrikes which ACLED classifies as targeting civilians commonly involve the indiscriminate bombardment of villages where opposition forces operate. In these cases it would appear that civilians are being targeted along with combatants, constituting a form of collective punishment. Again the intention seems to be to erode civilian support and force fighters to move elsewhere.

Less common are direct, targeted air attacks against ordinary civilians who the Tatmadaw regards as supporting opposition forces. Although infantry units have regularly carried out direct attacks – including dozens of massacres – against civilians for this reason, certain factors restrict the Tatmadaw’s ability to use airpower for this purpose. For a start, airstrikes are resource-intensive and far more expensive than mortar bombs or bullets, or the gasoline the Tatmadaw uses when setting fire to villages. 3 Moreover, given its current inability to deploy adequate reinforcements, close air support has become the Tatmadaw’s last line of defence against opposition attacks on critical assets. In other words, in deploying its limited air assets, the Tatmadaw must prioritise the more significant threats it faces, which are from armed groups rather than civilians.

As discussed below, airstrikes that inflict harm on civilians are one of the most significant obstacles to negotiated ceasefire in Myanmar.

Conflict in the earthquake zone

The overall rate of armed violence in the central Mandalay, Magway and Sagaing regions decreased slightly following the 28 March earthquake. According to ACLED, the three regions experienced a combined total of 134 armed clashes and 139 air or drone strikes in April, compared with 160 armed clashes and 153 air or drone strikes in March. The earthquake severely impacted transportation routes throughout central Myanmar and diverted combatants and their civilian supporters from the war effort.

In April, the fighting in central areas affected by the earthquake was most intense across four fronts.

1. To the north of Mandalay City, 26 airstrikes and 13 ground clashes were recorded along Highway 31 through the townships of Madaya, Singu and Thabeikkyin. The fighting here is the result of an ongoing Tatmadaw counter-offensive against the TNLA and its ally the Mandalay PDF, which resumed after a pause of several days following the earthquake. The Tatmadaw has also continued its counter-attack against the TNLA in Nawnghkio Township, Shan State, which began in February.

   The two-pronged operation has so far yielded moderate success in dislodging some TNLA and PDF positions, albeit at enormous cost to civilians. On 18 April the Tatmadaw began several days of bombing against the adjacent villages of Leik Kya and Yae Htwet in Thabeikkyin Township, resulting in the deaths of up to 40 civilians, including children. The area is home to a number of gold-mining businesses, and multiple armed groups, including the Mandalay PDF, operate there. 4

2. Airstrikes and clashes between ground forces continued near the banks of the Irrawaddy River in both Myingyan and Pakokku districts, albeit at a reduced tempo. ACLED recorded 31 armed clashes and 26 air or drone strikes across the two districts in April, compared with 50 armed clashes and 53 air or drone strikes in March. Since August 2024 this theatre has seen a loose coalition of PDFs wage a sustained insurgency against the Tatmadaw and a group of regime-aligned militias known as Pyu Saw Htee. Multiple PDFs continued to carry out assaults in the days after the quake, while the Tatmadaw launched counter-attacks and airstrikes against PDF camps and in support of its ground forces.

3. The Tatmadaw launched a string of airstrikes in Kawlin District from 9 April, the day after forces led by the KIA had captured the strategic town of Indaw, not far to the north. The airstrikes were concentrated along the two roads that stretch south from Indaw, and could have been either retribution for the loss of the town or an attempt to halt the southward advance of the KIA and its allies. One of the airstrikes targeting a PDF checkpoint in the village of Nan Khan resulted in the deaths of at least 20 civilians, including children. Prior to these incidents the Kawlin area had seen little combat since November 2024.

4. A cluster of 26 airstrikes was recorded in April along the Chindwin River in Yinmarbin District – not far from Monywa Town, where the regime’s Northwestern Regional Military Command is located. These attacks, many carried out by motorised paraglider, did not coincide with any major fighting on the ground. They included indiscriminate strikes on villages and appeared to be a harassment campaign designed to keep pressure on opposition forces and civilians while the local Tatmadaw garrison appeared to stay in its barracks. The area is a stronghold of the Burma National Revolutionary Army led by Bo Nagar and is also home to Letpadaung Copper Mine, a contentious Chinese investment project.

Violence and conflict hotspots in the earthquake zone, April 2025

Loading map ...

Custom Layers

Click to toggle on/off

Violent events

ACLED

Conflict hotspots

MIMU boundaries

Earthquake intensity

USGS ShakeMap

Attack/armed clash

Air/drone strike

Offensives and counter-offensives outside the earthquake area

Outside central Myanmar and the immediate earthquake zone, multiple EAOs continued their offensives or launched attacks in new directions in the days and weeks following 28 March, resulting in several notable advances. Besides its attacks on the TNLA, the Tatmadaw carried out two other counter-offensives.

Perhaps the most significant opposition advance was the AA’s seizure of the Nyaung Kyoe camp on 2 April, the day after the Brotherhood Alliance’s ceasefire declaration. The camp lies along the mountain road leading from Taungup, Rakhine, to the town of Pyay, Bago. Its capture opens the door for AA incursions into lower Myanmar, including the area where many of the Tatmadaw’s weapons factories are located. The Tatmadaw has sent several divisions into the area to repel the AA, but the current situation is unclear because reporting on the counter-offensive is almost non-existent.

Within days of the quake, forces supported by the AA launched attacks on the regime in Salin Township, about 55 kilometres north of Magway Town. The assaults are part of an escalating proxy war the AA is waging outside Rakhine State, its main area of operation and control. Clashes and airstrikes were also recorded in Kyaukphyu, Rakhine, which is the location of the beginning of the China–Myanmar pipeline and one of the few remaining areas in the state where the AA is fighting to expel the regime.

A significant escalation was recorded in Kachin State, with 56 armed clashes and 17 air or drone strikes recorded in April, compared with only 21 armed clashes and 12 air or drone strikes in March. Contributing to this tally was uninterrupted urban combat in the town of Bhamo, which the KIA has been besieging since December. In the days after the quake the KIA launched constant assaults on the remaining Tatmadaw positions in the town. 5 The ongoing battle has cost the KIA significant resources and hundreds of casualties.

Across the state border in Sagaing, the KIA and allied PDFs won a major victory on 9 April when they seized the strategic town of Indaw following an eight-month battle. The attackers captured the commander of the Tatmadaw’s 77th Light Infantry Division along with stockpiles of weapons and ammunition. The town sits at a crossroads guarding one of the main corridors between central Myanmar and Kachin State, and is vital for both sides. As of late May, the Tatmadaw had not attempted to retake the town.

On the morning of 31 March, KIA fighters attacked and dislodged a combined force of Tatmadaw and militia soldiers along the Ledo Road near the town of Kamaing, Kachin State, sparking a two-week battle. By 23 April, a two-pronged counter-offensive by the Tatmadaw and powerful ethnic militias opposed to the KIA had taken shape. This operation seeks to regain control over Hpakant, a lucrative jade-mining area, with Tatmadaw and militia forces pressing north from Indawgyi Lake and west from Kamaing. The offensive could also aim to draw the KIA away from Sagaing at a moment when many KIA fighters are bogged down in Bhamo.

Another notable escalation took place in Kayin State, with 51 armed clashes and 48 air or drone strikes recorded in April, compared to just 33 clashes and 12 air or drone strikes in March. This trend was driven by a fresh offensive led by the KNLA against the regime-held towns of Kawkareik and Kyondoe. In March, clashes and airstrikes were confined to the western foothills of the Dawna mountains, just outside Kawkareik Town. In the first week of April, KNLA forces moved to encircle the town, prompting the Tatmadaw to respond with shelling at various locations. By mid-April a major battle along the Asian Highway 1 was under way.

In eastern Bago Region, fighting escalated to the east of the Yangon–Mandalay highway. This area, which the earthquake affected, has been the site of an unrelenting campaign waged by the KNLA and allied PDFs since late 2023. At least 40 armed clashes and 31 air or drone strikes were recorded here in April, compared to 30 clashes and 20 air or drone strikes in March. Opposition forces briefly captured a battalion headquarters in Shwegyin Township on 8 May but were later repelled.

Armed clashes and air or drone strikes (all actors) in April 2025. Source: ACLED

Uninterrupted violence

The earthquake and subsequent calls for a humanitarian pause to the fighting constituted the first real opportunity for a nationwide de-escalation since the 2021 coup. However, most key actors continued to fight on for various reasons, with the specific dynamics varying across theatres. By and large, in areas where conflict actors held the military initiative before the quake, they have continued to press their advantage.

Some opposition forces and their supporters feared that a humanitarian pause after the earthquake could derail the momentum of their armed struggle, and that international relief funnelled through the junta would appear to legitimise its rule. On the other hand, the disaster provided an ideal strategic opportunity to strike the regime. The earthquake crippled the junta’s administration, disrupted command and control between Naypyidaw and frontline units, restricted the army’s mobility, and possibly had some impact on supply lines and military-industrial production. Opposition forces also benefited from an initial disruption to several of the regime’s air bases, although this did not last long. The AA, KIA, KNLA, CNF, and KNDF all made territorial gains shortly after the earthquake.

As the ACLED data shows, in the areas around Monywa and Kawlin, the Tatmadaw compensated for its weakened force posture by escalating air attacks in the absence of ground clashes. In April it waged an almost constant counter-offensive against the TNLA and Mandalay PDF in the three townships north of Mandalay, with airstrikes taking place at double the frequency of armed clashes in the area. It also launched a new offensive against the KIA in Kachin.

Of the 333 air or drone strikes the Tatmadaw carried out in April, 137 (41%) were recorded in the earthquake-affected regions of Mandalay, Magway and Sagaing, although these three regions saw only 30% of the armed clashes recorded across the country in April. Moreover, the earthquake did not lead to a reduction in airstrikes, despite a marginal decrease in armed clashes between March and April, both in earthquake-affected areas and at the national level. These dynamics followed the pre-earthquake trend where the incidence of armed clashes fell as the number of air and drone strikes rose. The finding makes sense from an operational perspective, as the earthquake hampered the Tatmadaw’s mobility on the ground and drew away resources, prompting it to partially compensate with airpower despite the dire humanitarian emergency following the disaster.

Analysis of conflict hotspots shows that airstrikes recorded after the quake followed the usual patterns. For example, the Tatmadaw launched airstrikes in defence of its embattled positions or in support of counter-attacks. These dynamics were most apparent in Kayin State, Rakhine State, Myingyan and Pakokku districts, Bhamo, and the battlefront north of Mandalay City.

The Tatmadaw also carried out intelligence-driven strikes in opposition-controlled areas as a punitive response to EAO advances. On the evening of 30 March, for example, there was a UAV strike on Laiza, the KIA’s nominal headquarters. Since 2022, attacks on Laiza have been a common response to KIA operations further afield. On the evening of 31 March, a Tatmadaw airstrike killed 36 PDF cadets at a training ground in Mohnyin Township, where the KIA’s 8th Brigade operates. The 8th Brigade is known to train and support PDFs that it sends to fight in Sagaing, and was probably involved in the assault on Indaw. The Tatmadaw carried out similar punitive attacks in northern Rakhine throughout April and May, inflicting numerous civilian casualties. Sources said at least some of these attacks – such as the 13 May strike on Rathedaung that killed at least 12 civilians – targeted AA administrative offices in residential areas.

Additionally, the Tatmadaw carried out almost continuous airstrikes against civilian targets. Of the 333 air or drone strikes in April, ACLED classified 108 (32%) as targeting civilians. These strikes generally fall into one of three categories (although the data does not distinguish between them in this way): attacks targeting civilian members of opposition groups, indiscriminate attacks on civilians in contested areas, or attacks on unreported military targets that inflicted collateral harm on civilians.

One major incident that did not fit any of those categories was the 12 May attack on a school in Tabayin, Sagaing, that killed more than 20 children and injured dozens more. IISS researchers examined available satellite imagery along with photos and videos (some unpublished) of the aftermath. Notably, the blast destroyed only part of the school’s roof, causing it to cave inwards, and all the walls remained standing. There were no burn marks on either the building or the victims, while several of the children suffered severe head trauma although their bodies were fully intact. This probably indicates the use of a precision-guided airburst munition that detonated above the roof of the school, raining fragments down onto the children below. Although the Tatmadaw denied carrying out the airstrike, it is the only actor in Myanmar possessing such an advanced weapon. The attack took place away from the frontline and it is not clear why that specific school was targeted in such a manner.

Airstrikes as a primary obstacle to de-escalation

Airstrikes against civilians are among the most significant obstacles to near-term conflict de-escalation in Myanmar. They inflict grievous physical harm on their victims and have severe psychological consequences for communities that experience them. Air attacks also highlight the asymmetry of the conflict, as opposition forces have no way to defend against them, and provoke intense reactions from the public. Many domestic and foreign media outlets, for example, framed their reports on the earthquake around the continued airstrikes. 6 The Tatmadaw’s air campaign, especially against civilians, broadly precludes conflict de-escalation by stoking outrage among the Myanmar public and making it generally unwilling to support negotiations with the regime.

But there is also a more technical way in which airstrikes are an impediment to negotiated conflict de-escalation. Given its loss of territorial control and shortage of personnel, airstrikes remain one of the Tatmadaw’s most potent forms of leverage. Meanwhile, continued attacks on, or control over, critical Tatmadaw assets like cities, command headquarters and critical infrastructure constitute the opposition’s biggest bargaining chip. In any negotiation, the Tatmadaw would probably demand a halt to major opposition attacks or the return of some captured areas in exchange for a pause in airstrikes.

However, this proposition would remain anathema to opposition forces, who reject negotiations in which Tatmadaw air attacks that frequently inflict civilian casualties are equated with their own battlefield advances. The difficulty is that, despite their illegality and brutal nature, air attacks against civilians and the services they rely on remain a fundamental part of the Tatmadaw’s counter-insurgency doctrine. From a technical negotiation perspective, there is no way to separate this instrument of power from the wider military context in which opposition forces otherwise retain the upper hand. In other words, because airstrikes and attacks on civilians are two of the Tatmadaw’s principal bargaining chips, opposition forces would need to concede elements of their own military advantage to get them to stop. In the aftermath of the earthquake, unilateral ceasefire declarations could not close this gap.

With that said, the recent ceasefire and deal between the MNDAA and Tatmadaw show that this dynamic can be overcome when conflict parties engage in mediated talks. After the fall of Lashio, the Tatmadaw continued aerial attacks on sensitive MNDAA assets and civilian infrastructure in the city. Under the ceasefire deal brokered by China in January 2025, the MNDAA has effectively limited its support for allies and proxy forces and handed back Lashio, which it had paid a significant price to capture. In exchange, the Tatmadaw halted air attacks and granted de facto recognition to the MNDAA’s authority over the rest of its conquered territory.

This article is Part 1 of an extended analysis on the earthquake’s impact on conflict dynamics in Myanmar. In the aftermath of the earthquake, Part 2 will consider the implications of ongoing violence for the future of the conflict.