Alert-level decision during volcanic unrest on La Palma's southwest rift
Since 03:00 on 12 October, the INVOLCAN network on La Palma has recorded 312 small earthquakes beneath the Cumbre Vieja southwest flank, clustered at 9–12 km depth, with a maximum magnitude of 2.9. Over the same 10-day period, two permanent GNSS stations show 4.5 cm of inflation, and the summit SO2 flux has risen from a baseline of ~5 t/day to 48 t/day. There is no volcanic tremor yet, and no visible deformation in InSAR. The area above the swarm contains the villages of El Paso and Todoque (about 2,500 residents within 6 km of the inferred dyke path), a banana packing plant, and a single road that is the only evacuation route. The Civil Protection director asks you, as the scientific advisor, to recommend an alert level on a four-step scale (green, yellow, orange, red) and whether to order a preventive evacuation. Tourism is important and a false alarm would be costly; the last eruption in the area began 11 days after a similar onset.
1. Interpret the monitoring dataset. What does each signal tell you about magma movement, and which signals are absent? Explain why the combination matters more than any single measurement.
Read the worked answer
The earthquake swarm at 9–12 km with M≤2.9 indicates brittle failure under increasing fluid or magma pressure, but the small magnitudes mean it is not yet a sign of imminent surface eruption. The 4.5 cm of inflation in 10 days is a clear departure from background and implies volume addition at depth; the lack of visible InSAR deformation is not contradictory, because 4.5 cm over a small area may be near detection limits or masked by atmospheric and vegetation effects. SO2 at 48 t/day, roughly ten times baseline, shows that magma-derived gas is reaching the surface and that the system is partly open. The missing signals are important: there is no tremor, no evidence that hypocentres are migrating shallower than 9 km, and no thermal anomaly. The combination of seismicity, deformation and gas is stronger than any single dataset because independent processes converge on a magmatic intrusion, but the absence of shallow seismicity and tremor means the dyke has not yet reached the surface. My interpretation is pre-eruptive unrest, not eruption.
2. Recommend an alert level and whether to order a preventive evacuation. Justify your choice with explicit reference to uncertainty, the precautionary principle and the cost of a false alarm.
Read the worked answer
I would raise the alert to orange, activate the emergency plan and order a preventive evacuation of the most exposed sector within the inferred dyke corridor, while stopping short of a full red-level evacuation of the whole municipality. The data are not conclusive: there is no tremor and no shallow seismicity, so eruption is not certain. But 4.5 cm of inflation, a tenfold SO2 increase and 312 events in 10 days are a coherent magmatic signal, and the historical precedent of an eruption beginning 11 days after a similar onset raises the conditional probability. The precautionary principle applies because the potential harm is irreversible and the exposed population is small enough to move along a single road if action starts early; waiting until tremor or shallow seismicity appears could compress evacuation into hours. The cost of a false alarm is real-lost tourism income, disruption and reduced future compliance-so I would make the evacuation zone narrow and explain the triggers for standing down. This is a judgement under uncertainty, not a certainty of eruption.
3. What additional observations would most change your decision, and why?
Read the worked answer
The most decisive change would be seismicity migrating shallower than about 5 km, because that would show dyke propagation toward the surface. Volcanic tremor would indicate sustained fluid movement and a much higher probability of eruption. A rapid increase in inflation, especially if it accelerates above about 10 cm, would imply pressurisation beyond the current trend. A jump in SO2 to hundreds of tonnes per day, or the first thermal anomaly, would suggest an open conduit. Conversely, a decline in seismicity, SO2 and deformation over several days would justify lowering the alert. I would also request wind forecasts for ash dispersal and rainfall forecasts for lahar potential, since rain on fresh ash can generate dangerous mudflows. Finally, social data-hospital capacity, school locations and road reliability-would affect how and when to evacuate. These observations change the decision because they discriminate between a stalled intrusion and one that is actually ascending.