The global toll of natural disasters, 10,000 to 20,000 lives lost annually, is tragically high, yet the more powerful story is one of progress. The data from EM-DAT shows that death rates from disasters have fallen by more than 90% over the last century, a remarkable achievement driven by early warning systems, sturdier infrastructure, and coordinated international response. This is not a reason for complacency, but it is a reason to study what works and to apply those lessons with the same rigor we bring to climate modeling and ocean observation. The world has become more resilient, and that resilience is a product of deliberate, measurable action.
For our readers, whether researchers, policymakers, or engaged citizens, the practical takeaway is clear: resilience is built on data, not instinct. The same integrated data ecosystems that track ocean currents and climate indicators are now essential for mapping vulnerability to disasters. We cannot protect communities we cannot see. As we noted in our coverage of From observation to action: measuring maturity in ocean practices, moving from passive data collection to active decision-making requires a calibrated understanding of what is being measured and why. The disaster database itself is a case in point: the apparent fivefold increase in disaster events since the 1970s is largely an artifact of improved record-keeping, not a sudden surge in nature's fury. We must be careful not to confuse better reporting with a worsening reality, just as we must avoid underestimating the gaps that remain.
Those gaps are significant. Heat-related deaths, for instance, are severely underreported, with over half of all recorded heat events coming from just nine countries, while sub-Saharan Africa is nearly invisible in the data. This is a critical blind spot. As the world warms, understanding the true toll of extreme temperatures will require the same kind of global collaborative effort that has improved earthquake and storm monitoring. The undersea cables that carry 99% of global data flows are a physical backbone for that collaboration, but as we explored in Undersea cable competition tests global data collaboration, those cables are also points of geopolitical tension. The infrastructure that enables real-time data sharing and early warning is itself vulnerable. Protecting it is as much a part of disaster resilience as building stronger seawalls.
What should we watch next? The economic costs of disasters remain severe and unevenly distributed, often hitting the same vulnerable populations year after year. While death rates have fallen, the number of people displaced or left homeless still exceeds 100 million annually in most years. The question is not whether we can reduce the death toll further, we have proven we can, but whether we can extend that same resilience to the millions who lose their homes, their livelihoods, and their stability. That will require not just better early warnings, but better longitudinal data on who is affected and how. The next breakthrough in disaster resilience may come not from a new sensor, but from a more complete, peer-reviewed picture of the people the sensors are meant to serve.