Challenge Fund: Round 3

Round 3 centered on disaster risk financing for early action, machine learning and big data for disaster risk financing, and disaster risk financing mechanisms to manage food insecurity.

Towards Impact-Based Forecasting: Upgrading InaSAFE and GeoSAFE to Enable Forecast-Based Action

Institutes

Description

Implementing FbF is often quite difficult, because when a hydrometeorological forecast arrives, it is not clear to humanitarians what kind of impacts to expect (Houses destroyed? Roads closed? Where?). Without information about potential impact, humanitarians do not know what early actions to take and where to implement them. To answer these questions, the Red Cross Red Crescent Climate Centre and partners have adopted the concept of Impact-based Forecasting (IbF), an approach that combines the understanding of forecast skill, impact-hazard curves, and risk analysis to generate an intervention map that will inform when and where funds for early action should be deployed.

More information

SMART: A Statistical Machine Learning Framework for Parametric Risk Transfer

Institutes

Description

Despite the advantages of simplicity, transparency, and rapid payouts, there is a lack of confidence and reluctance to use parametric insurance because of basis risk – the misclassification of events due to false positives and false negatives. Basis risk leads to inefficient transactions and higher product costs, reducing their appeal to end-users and investors. This project will make novel and expert use of appropriate machine learning and statistical concepts to address these two issues and develop a new framework that is of general relevance to diverse hazards and communities.

More information

Forecast-Based Financing for Food Security

Institutes

Description

Despite advances in humanitarian aid and the growing understanding that pre-arranged disaster risk financing can be more cost-effective than post-disaster expenditures, associated uncertainties in forecast systems and cost-effectiveness remain large. As a result, forecast information is not routinely used as a basis for financing early action for food insecurity risk. Forecast-based Financing for Food Security (F4S) proposes to address this by: developing an impact-based forecasting model using machine learning on food insecurity drivers; collecting evidence on food insecurity triggers; evaluating the cost-effectiveness of different cash transfer mechanisms; and exploring potential channels to disseminate knowledge and make first steps towards operationalization.

More information

Challenge Fund Rounds