Pacific cables and the cost of single-thread connectivity

Submarine cables carry essentially all of the international internet traffic in the Pacific Islands. Satellite capacity has grown with the arrival of low Earth orbit constellations, but for most countries cables remain the primary path. Many island states depend on a single international cable, and several depend on a single domestic cable linking outer islands to the gateway. This is the connectivity baseline against which resilience needs to be planned.

The 2022 Hunga Tonga eruption and tsunami severed the only international cable serving Tonga and the domestic cable linking the outer islands. Restoration of the international link took approximately five weeks, with limited satellite backup providing constrained capacity in the interim. Financial services, government operations, humanitarian coordination and family communications were all degraded for the duration. The event is the clearest case study available of single-cable risk in a Pacific state, and its lessons have shaped donor and operator planning since.

Other recent events reinforced the same pattern. Cable faults affecting Vanuatu, repeated disruption to the Coral Sea Cable System serving Papua New Guinea and the Solomon Islands, and incidents in Samoa and elsewhere have each tested the ability of governments and operators to maintain critical services on reduced connectivity. The faults themselves were not unusual. Cable breaks happen routinely worldwide. The difference is that in densely connected markets the traffic reroutes invisibly, and in single-cable markets it does not.

Cyclone exposure adds a second dimension. The South Pacific cyclone season regularly produces Category 4 and 5 systems, and the long-term trend points to more intense events even where total counts remain steady. Cyclones disrupt power, telecommunications, port operations and aviation simultaneously, and recovery times in remote islands depend on the arrival of external support that itself depends on the same disrupted infrastructure. Organisations planning resilience for Pacific operations need to plan for events that affect every input at once, not for events that affect one input at a time.

Energy fragility compounds the picture. Many Pacific utilities depend on imported diesel for a significant share of generation, with limited inter-island grid connection and modest reserve margins. A single fuel shipment delay, a cyclone strike on a generation site or a fault in a key substation can produce extended outages. Renewable investment is changing the mix, but the transition is uneven and the underlying concentration remains.

Practical resilience for organisations operating in or serving the Pacific Islands tends to follow a few principles. Build for offline operation as a default, not as a fallback, for any service that customers or staff rely on during disruption. Pre-position satellite capacity, including LEO services, with the contractual terms and operational readiness required to bring it up within hours rather than days. Maintain relationships with national operators, regulators and disaster management offices before an event, because the response window is too short to build them during one. And accept that recovery time objectives written for densely connected markets do not translate directly to single-cable contexts.

There is a broader policy point as well. Cable diversification across the Pacific is progressing through projects such as additional spurs and new routes, supported by a mix of commercial investment and donor finance from Australia, New Zealand, Japan, the United States and others. Each additional cable that lands meaningfully improves the resilience baseline for the country it serves. Until that work is complete, the resilience burden sits with the organisations operating in the region, and the planning assumptions need to be honest about that.