Papua New Guinea sits where four tectonic plates grind against each other - the Pacific, the Australian, the Caroline, and a series of smaller microplates whose motion complicates every earthquake catalog written about the region. At the Huon Peninsula, the Woodlark plate is moving north and sliding beneath the South Bismarck plate along the line of the Markham Valley. The result is a collision zone that crumples and tears the crust in patterns that geologists are still mapping. The Finisterre Range itself is geologically young - thrust faulting began building it just 3.7 million years ago, and it is still growing. The slopes are steep, the rainfall is intense, and the bedrock slides when provoked. Historical accounts date strong earthquakes here to 1876, and a 1922 quake caused major damage. The 1993 sequence was not an anomaly; it was a chapter in a long book.

For years, seismologists assumed the culprit under this part of the coast was the Ramu-Markham Fault - the shallow thrust that marks the visible boundary between the Finisterre Range and the Markham Valley floor. The 1993 earthquakes forced a revision. The rupture geometry did not match a shallow crustal fault; the depth and angle pointed instead to a detachment surface called a decollement lying beneath the Ramu-Markham Fault. When the 1993 sequence broke, surface measurements showed that both the decollement and the Ramu-Markham Fault had moved together, with the rupture ceasing just a few hundred meters below the ground surface. The geometry changed how the rest of the Papua New Guinea coast was understood. The visible fault lines had been hiding the real ones.

At least 58 villages were damaged badly enough that roughly 8,000 people had to be evacuated. Homes and schools built of wood fared surprisingly well - timber frames flex, and flexibility is a gift during shaking - but the gardens that feed highland villages were destroyed by the ground disruption. Food supply became the first emergency. In total, 1,224 homes were demolished and 24 schools closed. The initial death toll of four climbed as search teams reached remote areas. By the time operations concluded, at least 65 people were confirmed dead. Near the Leron River, one landslide buried 19 people along a steep hillside in a single slide. The Papua New Guinea government created an educational program for the displaced, teaching landslide awareness to communities that now lived in rebuilt or relocated villages. A month after the quakes, most residents went home.

Landslides did more than bury houses. When hillsides collapsed into valleys, they dammed the rivers running south from the range. Quake lakes formed behind the debris piles - temporary reservoirs filling with the monsoon rains that follow a highland October. Some held. Others breached, suddenly and catastrophically. A breach near the Lae-Madang highway sent a wave of debris tumbling down and destroyed two bridges. The Leron River was dammed repeatedly, its quake lakes filling and overtopping and failing in sequence. The aftershocks that continued for months triggered still more slides on already unstable ground. In April 1995, a follow-up earthquake breached another lake and destroyed one of the bridges that had been rebuilt after 1993. The Yonki Dam and hydropower station on the Ramu River shut down automatically as part of its fail-safe procedure; inspections found no serious damage, and operations resumed. The quakes were felt in Lae, Madang, and Goroka, where the damage was limited to cracked roofing sheets and water tanks.

Papua New Guinea's seismicity is not a hazard to be solved. It is a condition of the place - a permanent feature of the geology, like rainfall or altitude. The Finisterre villages that rebuilt in 1993 did not move to a safer country because no such country was within reach, and because the land around them had been home for generations. What changed was knowledge. Satellite imagery made visible the scale of the slope failures that would have otherwise been invisible outside the immediate disaster zone. Government programs began teaching landslide risk at the village level. Engineers reassessed the bridges on the Lae-Madang highway. The next major earthquake in this range will come - the only question is when - and the survivors of 1993 passed some part of their knowledge to the generation that would meet it.