Skip to main content

How does ACLED code air- and drone strike events?

Frequently asked questions about air- and drone strikes, including how ACLED aggregates strikes and what this means for interpreting event counts in the dataset

24 June 2026

What is an air-/drone strike event?

ACLED’s Codebook defines the “Air/drone strike” sub-event type (nested under the Explosions/remote violence event type) as any air-to-ground attack by either aircraft or drones, occurring in the absence of any other engagement. Where an air- or drone strike happens alongside other tactics within the same event, ACLED codes the event according to the highest event type on a hierarchy

For example, see below how ACLED codes an event containing multiple tactics, according to the hierarchy:

An exchange of gunfire breaks out between state forces and an armed group in a town on the same day. During this firefight, a grenade is thrown, and property is looted

⟶ This example contains three of the six event types that ACLED codes. BattlesExplosions/remote violence, and Strategic developments.

ACLED codes to the highest event type on the hierarchy, as Battles → Armed clash. Information on the use of grenades and looting is available in the Notes column.

Why is the number of events different from the total number of strikes?

There are two key reasons why ACLED’s number of Air-/drone strike events may differ from the total number of reported strikes: How ACLED aggregates events and how ACLED disaggregates locations.

Event aggregation

ACLED records a single Air/drone strike event for all air-to-ground strikes (individual missiles, drones, or munitions) that occur on the same date and in the same location, involving the same actors

Strikes that differ on any of these criteria are recorded as unique events. This method of disaggregating events applies to all types of political violence ACLED records. (For more, see Event aggregation in the Codebook)

For example, if military forces hit the same target with four airstrikes on the same day, ACLED codes one single event. However, if military forces hit four different locations with airstrikes on the same day, ACLED codes four separate events.

Example of airstrike aggregation

This means that after a period of intensive conflict, the gap between the number of strikes cited in official statements and the number of strike events ACLED records may be significant. 

Location disaggregation

ACLED disaggregates locations based on multiple factors including geographic size, event volume, population, and relevance for trends analysis. Some cities are considered a single location, while others are disaggregated into smaller location units, such as individual neighborhoods. Disaggregated locations are noted in the location column of the dataset, following the format of “city - neighborhood”. For example, “New York City - Manhattan”. 

This variance has implications for aggregated event counts.  If City A is treated as one location in our dataset, then all strikes in a single day on the whole city involving the same actors will be recorded as one event. If City B is disaggregated into neighborhoods, then all strikes in a single day in each location unit involving the same actors will be recorded as one event. That means if five different location units in City B are hit with strikes in one day, we will record five Air/drone strike events for City B. 

ACLED continues to improve location granularity and any newly disaggregated locations are announced in ACLED’s Update Logs on the Knowledge Base

Where can I find the total number of strikes in an air-/drone strike event?

If you are looking for total strike or munition counts, these may be recorded in the Notes column of individual events in the dataset. For guidance on how to extract information from the “Notes” column, see How to Analyze ACLED’s Notes Column using Natural Language Processing (NLP). 

How can I filter for specific drone information?

Information specific to drones, including the type used (for example, fiber-optic drones) may sometimes be recorded in the Notes column where reported by sources. As with strike and munition counts, this is not systematically reported across the data due to differences in information available during different conflicts, so it is more useful for identifying types of munitions in individual events rather than for trends analysis. 

To extract events with reported drone use from other types of airstrikes, the Notes can be filtered for keywords, such as “drone.” For guidance on how to extract information from the Notes column, see “How to Analyze ACLED’s Notes Column using Natural Language Processing (NLP).” 

How do Shelling/artillery/missile attack events differ from Air-/drone strike events?

The “Shelling/artillery/missile attack” sub-event type captures the use of long-range artillery, missile systems, or other heavy weapons platforms in the absence of any other engagement. This covers the use of shelling, artillery (stand-alone or tank-based), mortars, guided missiles, rocket-propelled grenades (RPGs) and other shoulder-fired missiles, heavy machine guns, and automated grenade launchers. The key distinction from Air/drone strike events is the delivery method — ground- or sea-based weapons systems versus air-to ground attacks by aircraft or drones. 

How can I find data on intercepted air-/drone strikes?

The sub-event type "Disrupted weapons use” is used when an air-/drone strike is prevented from occurring. For more, see Strategic developments in the ACLED Codebook. To find interception events, filter for “Disrupted weapons use” sub-event type, and search the Notes column for “Interception,” and the keyword of the tactic of interest, e.g., “Drone” or “airstrike.” 

Note: Interceptions of reconnaissance drones, rather than armed projectiles, are coded under the “Looting/property destruction” sub-event type, since Disrupted weapons use is specific to weapons and weapon systems.

Air- and drone strikes that cause damage or casualties, even where an interception was attempted, are coded as Air/drone strike and not Disrupted weapons use.

How can I find a full picture of strike activity under all event types?

Apply all three of the following filters in the dataset:

  • “Air/drone strike” sub-event type 
  • “Disrupted weapons use” sub-event type filtered for “Interception” in the Notes column 
  • “Battle” event type filtered for “airstrike” and/or “drone” in the Notes column
    Region
    Global
    Share on

    Related content