A communication gap between scientists and decision-makers was revealed during the Horn of Africa's 2011 drought.
Flickr/ Oxfam International
Stronger links between scientific tools and the environment in which they operate can improve the effectiveness of early warning.
To say there is a lot riding on disaster management is an understatement. The lives and livelihoods lost, as well as extensive damage, underscore the human, environmental, social and economic cost of extreme events that affect millions every year. 
Though these events differ (for example, some strike rapidly while others develop slowly), they all highlight the vulnerability of poor communities around the world — and variable success for science and technology (S&T) in disaster planning and response.
Thousands of lives were lost in the 2004 tsunami, exposing the need for a coordinated early-warning system in the region. In Africa, while scientists warned of a serious drought, failure to heed these predictions revealed the gap in communication about the risk with decision-makers.
Information and communication technologies (ICTs) such as social media and mapping, used after the Haiti earthquake, did help aid workers quickly assess damage and people's needs. Even here, the benefit of technology is limited without integration with existing response systems.
There is a range of preparedness and response activities that can help reduce risk from natural hazards. But the space between early warning and action is a narrow window of opportunity where more effective use of scientific tools for alerting communities can make a difference.
Spotlight on early warning
This week, we publish a set of articles that explores the barriers to early action and how early warning tools can be used to their best potential.
An overview article looks at how early warning systems have developed and what they can realistically achieve. Lucy Pearson of the Humanitarian Futures Programme at King's College London and the Asian Disaster Preparedness Center in Thailand, our editorial consultant for the project, identifies the factors that account for the gap between early warning and action, and highlights the role of knowledge held by local communities.
In a news feature, Smriti Mallapaty explores the benefits of integrating indigenous knowledge with modern science to improve early warning, even though traditional practices may not always keep pace with social and environmental change.
Andrew Collins, director of the Disaster and Development Centre at Northumbria University, United Kingdom, makes the case for embedding warning systems in local communities, and argues for better understanding of the factors — from behaviour to poverty — that stand in the way of early action.
Scientific uncertainty is a key reason why communities and decision-makers may not heed early warnings. Emma Visman of the Humanitarian Futures Programme, Benedict Dempsey of Save the Children, London, and S. H. M. Fakhruddin of the Regional Integrated Multi-Hazard Early Warning System in Thailand, argue that dialogue can improve how people engage with complex information.
The critical moment — for rapid-onset emergencies in particular — comes when a warning is issued. Veronica Grasso of the UN Development Programme in Switzerland and José Rubiera from Cuba's national forecast centre say this is where coordination and political will determine success or failure.
S&T is only a part
Cuba shows that early warning can be successful even without sophisticated technology. But coordination and political will are elusive goals, and are much easier to achieve where governance is sound and where people are at the centre of a country's priorities.
Just as problematic is the fact that S&T rarely offer a solution alone, not least because alerts need to be translated for action.
Meanwhile, the increasing number and sophistication of early warning systems — and the 'better safe than sorry' approach to using them — means that false alarms are rising. The recent conviction of Italian scientists for not issuing a warning about the L'Aquila earthquake can only add to an over-cautious approach.
The fact remains that early warning is one of the most powerful tools for reducing disaster risk. The use of S&T has become more organised. Digital tools such as social media and mapping are now seen by some as the future for organising humanitarian work.
And while better remote sensing, forecasts and mobile communications are essential, so are developments in the socioeconomic frameworks that shape our ideas of how early warning can become more effective.
Poverty and development links
For instance, the concept of risk reduction shows that exposure to natural hazards is not all that matters. Vulnerability — and conversely, resilience — can make all the difference. The risk is highest where people are least able to protect themselves. Disaster risk and poverty are locked in a cycle where natural hazards undermine economic development in the countries least well prepared to cope.
As the contributions to this Spotlight suggest, risk reduction must include factors linked to poverty reduction, such as ecological and social stability. And to be more effective, new tools need reinforcement from old processes including traditional knowledge, dialogue and coordination.
Progress in science is incremental, and there is a sense of urgency about disaster risk. The threats to poor countries will only grow and become more complex as climate change takes hold and resources are depleted.
So how can we make best use of that precious window of opportunity offered by early warning?
These articles highlight the need to build a mutual relationship between early warning science and the environment in which it operates. This means not only implementing technology but also empowering communities to reduce risk in a way that aligns with wider development priorities.
 UN. The future we want: Disaster resilience. Rio+20 Fact Sheet (2012)
All SciDev.Net material is free to reproduce providing that the source and author are appropriately credited. For further details see Creative Commons.