2018 Sulawesi earthquake and tsunami

The night the bay answered the earth

It was a tropical evening when the earth began to move. At 18:02 local time on September 28, 2018, buildings in Palu—hotels, shops, apartment blocks—shuddered and rolled. People who later recounted the moments described furniture sliding, glass rattling and a deep rumble that never quite stopped. The quake was centered offshore along the Palu–Koro fault, a major left‑lateral strike‑slip fault that runs along the narrow neck of Sulawesi. By the time the shaking eased, the worst was not yet visible: the sea had a different answer to the ground’s violence.

Palu sits at the head of a long, narrow bay—an amphitheater of water surrounded by steep undersea slopes and a shoreline of soft, recent sediments. That geometry turned out to be a trap. Within minutes of the earthquake, waves rushed into the bay with a speed and force many did not expect. Where people thought safety might be only a short walk away, the water arrived faster than warning systems could respond.

The fault that woke up a whole city

Sulawesi is a jigsaw of colliding plates and microplates. The Palu–Koro fault is one of the region’s cancerous seams: capable of sudden, large slips and surface rupture. The 2018 event registered at about magnitude 7.5 and had a shallow focus—conditions that produced intense shaking close to shore. Networks across Indonesia recorded the quake; power and mobile networks faltered in its wake.

Strike‑slip earthquakes, which slide blocks past one another, do not always push large volumes of seafloor up or down. Yet the Palu quake did something unexpected: in the confined geometry of Palu Bay, seafloor disturbance and, crucially, submarine slope failures combined to feed a tsunami far larger than many specialists had anticipated for a strike‑slip event. The science behind that combination would become clearer only after days, weeks and months of study.

When the water came faster than a warning

Tsunami warnings are notoriously difficult to use for near‑field events—those whose source lies close to the coast—because seconds and minutes matter. BMKG, Indonesia’s meteorological and geophysical agency, issued warnings, but the warning window in Palu was measured in minutes. In many neighborhoods sirens failed to sound, mobile networks were already spotty, and people were left to decide using what they felt and saw.

Eyewitness video and testimony show the first waves arriving rapidly along the shore: cars floating off coastal roads, boats smashed against piers, and waves washing through neighborhoods. In Palu Bay the shape of the coastline amplified incoming energy; the bay funneled the waves, increasing their height and reach. In places water crested several meters above the normal tide line and poured inland, dragging debris, vehicles and buildings in its grip.

Streets that flowed like rivers — the liquefaction that ate neighborhoods

While the tsunami felled parts of the coastline, something else was happening on land that would horrify engineers and residents alike. Neighborhoods built on young, waterlogged alluvial deposits—areas where rivers once spread out and left loose sediment—turned into moving ground. In Petobo and Balaroa, entire blocks flowed like wet clay. Houses were not simply collapsed; they were carried away or buried in chaotic, hummocky deposits. Roads buckled, trees leaned at impossible angles, and the ordinary geometry of streets dissolved.

This is liquefaction: saturated soils lose strength under intense shaking and behave like a fluid. The results were catastrophic here because homes and communities had been built on those vulnerable sediments. Satellite imagery later showed the scoured patterns and the dislocated parcels of land—easy to see from above, impossible to ignore on the ground.

Rescue under broken wires and damaged roads

When daylight came, the scale of the destruction was visible. Hospitals strained under the influx of the injured; makeshift wards and triage points appeared in schoolyards and open fields. Indonesian national agencies—BNPB (the national disaster management authority), BASARNAS (search and rescue), the military, police and scores of community volunteers—poured into the affected zones. International teams offered assistance; rescue dogs and heavy equipment were flown in.

The operation was hampered from the outset. Communications were patchy, roads were impassable in many places, bridges sustained damage, and aftershocks continued to rattle nerves and structures alike. In some neighborhoods, rescue teams had to work through thick mud and unstable ground, complicating search efforts and increasing the danger to responders. For weeks the cadence was the same: dig, search, pull survivors and bodies from the wreckage, and move on to the next site.

Counting the cost: lives, homes, livelihoods

Official tallies eventually placed the death toll at around 4,340 people, with more than 10,000 injured and tens of thousands displaced. Those numbers tell a partial story. Entire families vanished under the flowing ground; communities lost elders, children, and breadwinners in a matter of minutes. Thousands of houses were destroyed or made uninhabitable. Ports and small businesses along the coast were smashed. Fisheries that had fed generations were disrupted. The economic hit was enormous: estimates of direct losses and reconstruction needs ran into the hundreds of millions and, by some calculations, approached and exceeded a billion US dollars.

For survivors the challenges would not end with reconstruction budgets. Trauma, questions about where to resettle, livelihoods lost to rubble and saltwater, and the slow, grinding process of finding and identifying the missing left wounds that rebuilding money could not quickly heal.

The quiet work of understanding why it was so bad

In the weeks and months after the disaster, scientists and engineers turned to data—seismic records, tide gauges, satellite photos, and field surveys—to untangle why a strike‑slip earthquake had produced such a devastating tsunami and why ground failure had been so extensive.

The emerging picture was layered. First, the bay’s narrow funnel and steep bathymetry were primed to amplify wave energy. Second, rapid submarine landslides triggered by the shaking moved large masses of sediment downslope; these mass movements acted like pistons, generating local, high‑amplitude waves. Third, local fault displacement and seafloor disturbances contributed additional complexity to the wavefield. When all of these factors aligned in the confined geometry of Palu Bay, the result was a tsunami of unusual size and reach.

Meanwhile, surveys confirmed that liquefaction and flow failures were concentrated where young, saturated alluvium lay thick. The damage patterns at Petobo and Balaroa became case studies in how modern coastal development can be vulnerable to such failure modes.

Voices that demanded change

Policymakers and communities took notice. Criticisms of the warning and communication systems were blunt: too many people did not get timely, clear instructions; sirens were absent or did not reach all neighborhoods; some official messages were delayed or inconsistent. The proximate nature of the hazard—in many places the tsunami arrived before centralized systems could fully react—meant that the best warning in a near‑field event is often what people feel: strong shaking.

Indonesia responded on multiple fronts. Investments increased in tide gauges, early detection networks and modelling for near‑field tsunamis. Local siren systems and public alert infrastructure were expanded in some high‑risk zones. Recovery and reconstruction plans emphasized hazard‑informed land use: moving settlements out of the most vulnerable coastal strips and liquefaction‑prone plains, building evacuation routes and vertical safe havens, and strengthening building codes in critical areas. Community preparedness became a sharper priority—teach people to run uphill immediately after strong shaking, and to trust their senses even when official word is delayed.

Recovery that stretched across years

The work of rebuilding Palu and surrounding regencies was not a single campaign but a marathon. Temporary camps and shelters housed displaced families while reconstruction plans argued with the realities of land availability, funding, and community attachment to place. Some people relocated to higher ground; others tried to return and rebuild in damaged neighborhoods. International aid and Indonesian programs provided tents, medical care and reconstruction funds, but processes for permanent relocation and full infrastructure replacement moved slowly and unevenly.

Identification of the missing, clearing of debris, rebuilding of emergency services and schools—these tasks continued for years. In many places the physical scars remained visible: sinks in the ground where houses had been dragged away, lines of new mounded earth for proposed relocations, and memorials that marked where homes once stood.

What Palu taught the world

The Palu disaster has become a frequent point of reference in tsunami science and disaster policy. It highlighted that near‑field tsunamis are governed by more than the obvious uplift of the seafloor: submarine landslides, local fault motion and coastal geometry can produce locally catastrophic waves even in events dominated by strike‑slip motion. It underlined the limits of centralized warning systems for hazards that can arrive in minutes, and it reframed thinking about coastal development on soft sediments.

Practically, the legacy has been a mixture of technical upgrades—more gauges, better modelling, louder and more widespread sirens—and cultural shifts in preparedness. Engineers, planners and communities now pay more attention to soil conditions, evacuation paths and the hard lesson that feeling an earthquake can be the most immediate sign to flee to higher ground.

The slow work of memory and resilience

Disasters change places and people in ways that are not easily quantified. In Palu there are rebuilt schools and new evacuation routes; there are also families who carry the absence of loved ones and towns that still measure time by the before and after of September 28. Scientists continue to study the event; policy makers continue to adjust plans. For residents, resilience has meant both practical adaptation—relocating houses, rebuilding livelihoods—and carrying forward the memory of a night when the bay and the ground conspired to uproot lives.

The images from Palu are stark but instructive: a ruined coastal road stained with mud and debris, a collapsed pier listing into the bay, tents where homes once stood. They are also reminders that in many coastal places around the world, a strong nearby earthquake should prompt immediate movement to higher ground. Palu’s lessons are technical and human, hard‑won and still relevant. They ask a difficult question: how do we live with hazards whose speed outraces our systems? The answers are part engineering, part community culture—and they are still being written.