Improving the Role of Information Systems in Anticipatory Disaster Risk Reduction (IRIS)

Lead Research Organisation: London School of Economics & Pol Sci
Department Name: Centre for the Analysis of Time Series


Humanitarian agencies are able to use weather (and other) forecasts to act in anticipation of humanitarian crises. For example, when a heatwave or hurricane is forecast, supplies can be moved into position early and emergency supplies positioned or pre-distributed. This reduces the overall impact and the cost of responding to the disaster. However, financing in advance of a disaster requires a high level of confidence in the forecast, to avoid the possibility of misallocated or wasted resources. Many forecasts are currently available but not all are accompanied by an assessment of the forecast quality. For example, it may be that the forecast is over-confident, predicting an event more times than it is actually observed, or it could be under confident, failing to predict events which do then occur.

We propose to develop and demonstrate a general method of measuring and displaying the information content of forecasts, using a novel idea which is based on existing research and freely available data. This will allow humanitarian agencies to act confidently in anticipation of humanitarian crises when there is sufficient information in the forecast, and to implement forecast-based financing schemes such as insurance or anticipatory funding allocation only when there is known to be confidence that the scheme will be effective.

Planned Impact

The direct impact of this work will be to improve the robustness and confidence of anticipatory disaster risk reduction mechanisms. This will have benefits for the following stakeholders:

1. Third sector disaster risk reduction organisations will benefit from more transparent and objective use of model forecasts in their decision making procedures. This not only improves efficiency, saving costs and potentially lives, but also mitigates to some extent the reputation risks associated with having opaque decision-making procedures or being seen to make "wrong" decisions or misallocating resources. This will be achieved by keeping the project open and transparent, working directly with our Start network partners throughout the duration of the project, and regularly soliciting open feedback from humanitarian and other organisations. As we near the end of the project we will seek to identify individuals within the Forewarn group who might take on development of the methodology and insights, and/or make further funding applications to develop the tool in new directions if further productive avenues are identified.

2. People caught up in the types of disasters we have studied will have earlier and better-targeted access to humanitarian assistance. Many of the alerts in the case studies we have chosen to focus on affect thousands or tens of thousands of people; some even more. Cost effective humanitarian interventions at an early stage can be critical in minimising the immediate and ongoing negative impacts of these events: for example, choosing between longer-term actions like constructing reservoirs and digging wells versus short-term distribution of water (and the many options in between).

3. The Government of Botswana faces challenges in making decisions about appropriate uptake of disaster financing mechanisms with respect to regional drought events, which have recently been severe. Our work with the Botswana Institute of Technology, Research and Innovation will help to strengthen the evidence basis for these types of investment decisions and allow an informed choice between the different options on offer, including the fallback option of investment in local infrastructure and resilience rather than in financial instruments.

4. The insurance industry including our partner Lloyd's of London will benefit from the increased robustness of the evidence basis for the use of model forecasts to produce disaster risk financing mechanisms such as parametric or forecast-based insurance. This will expand their potential markets into relatively undeveloped geographical regions, which are also attractive from the point of view of being relatively uncorrelated with many other major risks taken on by insurers and reinsurers. To ensure the knowledge transfer in this direction we plan to hold a workshop for insurance and humanitarian stakeholders about halfway through the project, and will solicit feedback and insights from the community. Our project partner at Lloyd's will also offer commentary and insights on a more regular basis on the outputs of the project, and will help to connect us with other relevant contacts in the sector.


10 25 50
Description The overall aim of this research programme is to develop a framework for making the best use of weather forecast information to support decision-making in advance of forecast humanitarian crises. We have begun the work with reference to a case study of heatwave in Pakistan, for which there are reasonably good weather forecasts and the possibility of action, as well as a strong presence of Start Network member NGOs on the ground. This has shown that the principle of looking at the timescales of predictability and action is both academically interesting and important in practice. Forecasts on short timescales can be confident, but there is little opportunity to take action. Where forecasts are available on longer timescales, there is more opportunity to take action but less confidence in the predicted outcome. Our research looks at the history of past similar weather events to define the useful timescales. In the case of heatwave in Pakistan, this showed that confident action could be taken with a lead time of over a week.
Exploitation Route The briefings and tools we produce will be used by Start Network members and could also be used more widely by other humanitarian agencies, by national or local government agencies, or by medical services.
Sectors Communities and Social Services/Policy,Environment,Healthcare,Government, Democracy and Justice,Other

Description The findings have been used to support internal decision-making procedures at the Start Network, in responding to a forecast humanitarian crisis due to heatwave in Pakistan. Dr Thompson produced a brief guide for NGO representatives explaining the outcomes of her work identifying suitable timescales for forecast-based action. Following a webinar and feedback (April 2018), this was put to the test when a heatwave was forecast for Sindh Province, Pakistan, in May/June 2018. Dr Thompson provided some additional real-time support for members as they used the information to improve decision-making, resulting in the timely release of £70,000 to support on the ground activities to reduce the impact of the crisis, saving lives and money.
First Year Of Impact 2018
Sector Communities and Social Services/Policy,Environment,Healthcare,Government, Democracy and Justice,Other
Impact Types Societal,Economic,Policy & public services