Anak Krakatau’s renewed activity has produced a serious volcanic-ash and aviation event across western Indonesia. The confirmed record does not, however, support every dramatic claim circulating online.
Indonesia’s National Disaster Management Agency, BNPB, said the current activity continued a phase that had intensified in July. An eruption at 14:05 Western Indonesian Time on 2 July produced an ash column about 200 metres above the summit. The alert was then raised from Level II, Waspada, to Level III, Siaga, at 16:30 that day. Level III remained in force in the latest official update.
The latest episode began at 23:07 on 4 September. PVMBG subsequently reported that a continuous eruption lasting about 25 hours had ended. That wording matters: the volcano remained active and capable of further eruptive events, but the continuous phase should not be presented as still under way after the agency said it had finished.
Trend table: what authorities had confirmed
| Indicator | Observation | Reference time or period | What it means — and does not mean |
|---|---|---|---|
| Alert level | Level III, Siaga | Latest PVMBG update accessed 7 September | A 3 km exclusion radius applied. It is not the highest Indonesian alert level. |
| Latest sequence began | 23:07 WIB | 4 September 2026 | BMKG used this as the start of the monitored eruption sequence. |
| Continuous episode | About 25 hours; subsequently ended | 5–6 September observation period | The end of continuous emission did not mean the volcanic system had returned to normal. |
| Seismic energy | RSAM rose on 5 September and declined on 6 September | PVMBG observation period | A measured decline over one day is not proof that the hazard has ended. |
| Ground deformation | Tanjung tiltmeter showed inflation since early August; Lava93 was comparatively flat | Latest PVMBG evaluation | Different instruments and locations can record different parts of the volcanic system. |
| Aviation ash | Ash from the surface to 15,000 ft, moving west at 15 knots | BMKG SIGMET valid 06:30–09:30 WIB, 7 September | This is an aviation observation, not the height of the volcano or a forecast of eruption size. |
| SIGMET documents | 39 issued | From first eruption to BMKG update on 7 September | The count records aviation hazard messages, not 39 separate eruptions. |
| Airports affected | Eight temporarily affected | BMKG morning update, 7 September | Operating status can change by NOTAM; passengers must check the airport or airline. |
| Sea-level anomaly | None indicating a tsunami detected | Monitoring through 07:30 WIB, 6 September | This describes the stated observation window; it is not a permanent guarantee. |
Ash, airports and the difference between a plume and an airport closure
BMKG used satellite, radar, seismic and sea-level systems to follow the event. Its satellite analysis at 08:00 WIB on 6 September identified two ash clusters across parts of Banten, Jakarta and West Java, with another reach towards Bengkulu. By the morning of 7 September, the agency said eight airports had been affected temporarily.
The latest SIGMET cited by BMKG for 06:30–09:30 WIB on 7 September placed volcanic ash between the surface and 15,000 feet, or about 4.57 kilometres, moving west at 15 knots. The affected area included parts of Banten, western and southern West Java, Lampung and Bengkulu, together with nearby waters.
An airport closure is an operational safety decision, normally implemented through a Notice to Airmen. It is not a direct measurement of eruption strength. Ash can damage engines, abrade windscreens and contaminate airport surfaces even when communities on the ground are outside the immediate exclusion zone.
Reports have offered different aggregate totals for cancelled flights and affected passengers. Nine145 is not publishing a composite number because the reports use different cut-off times and scopes. A changing operational total should not be converted into a single definitive statistic without a consolidated authority release.
Was there a tsunami?
BMKG’s multi-sensor Indonesia Tsunami Early Warning System found no sea-level anomaly indicating a tsunami around the Sunda Strait through 07:30 WIB on 6 September. This is the correct factual formulation. It does not erase the historical risk: the 22 December 2018 flank failure at Anak Krakatau generated a destructive tsunami without a large tectonic earthquake.
For the present episode, there is no verified basis in the cited monitoring record to claim that a tsunami occurred. Equally, “no anomaly through the stated time” must not be rewritten as an unlimited assurance about later conditions.
Public-safety boundary
PVMBG’s Level III recommendation barred residents, visitors, tourists and climbers from activity within 3 kilometres of the active centre. People beyond that radius still needed to follow official ash, health, maritime and aviation instructions. A surgical or cloth face covering may reduce inhalation of larger ash particles, but official health advice should guide protection where ashfall is present.
Data note
No confirmed deaths, injuries or residential displacement attributable to the September 2026 episode were identified in the official sources reviewed. That statement means “not reported in the cited record at the verification cut-off”; it must not be interpreted as a final disaster total.