Measuring progress in reducing disaster risk is perilously hard, catastrophe risk management group RMS has warned.

Chief Research Officer Robert Muir-Wood told the United Nations World Conference on Disaster Risk Reduction in Sendai, Japan, the nature of natural catastrophes makes it impossible to reach sensible conclusions by analysing historical data.

He used Haiti as an example: between 1900 and 2010 fewer than 10 people were killed by earthquakes, but in one afternoon in January 2010, 200,000 people died.

“You cannot simply take a decade of data and see what the underlying risk is,” he said.

“It will give you a completely erroneous sense.”

He says the only answer is to use catastrophe modelling to create a data set going back tens of thousands of years.

The insurance industry has used such models for 25 years, and Dr Muir-Wood believes the disaster resilience community must do the same to gauge progress in reducing risk.

“The future of managing disaster risk is going to be by making models and using those models not only to measure your progress but also to decide what actions you need to take to get you to where you want to go.”