The first step to establishing a facility such as ARC is understanding — in dollar terms — Africa’s weather-related food security risk. As the technical engine of the ARC risk pool, Africa RiskView is a core product of the ARC Agency. For its drought product, it combines existing operational rainfall-based early warning models on agricultural drought in Africa with data on vulnerable populations to form a standardized approach for estimating food insecurity response costs across the continent — information that is critical for financial preparedness for drought and for providing the basic infrastructure needed to establish and manage a parametric risk pool and trigger early disbursements.
In addition to supporting ARC, Africa RiskView provides decision-makers with expected and probable maximum costs of drought-related responses before an agricultural season begins and as the season progresses for every first-level administrative district in every country in sub-Saharan Africa. In addition to providing a financial early warning tool, identifying and quantifying risk in this objective way can also help countries and their partners direct appropriate drought response actions and target food security investments.
Africa RiskView is designed to interpret different types of weather data, including rainfall estimates, and information about crops, such as soil and cropping calendars. These data are then converted into meaningful indicators for agricultural production and pasture and applied to the vulnerable populations that depend on rainfall for crops and rangeland for their livelihoods. Africa RiskView then uses this information to estimate how many people may be directly affected (or have been affected) by drought or deficit rainfall in a given season. Using cost per affected person estimates, Africa RiskView estimates how much response costs to the observed drought event may be.
Population affected and response cost estimates can be calculated for the African continent, a region, a specific country or part of a country. Users can customise all aspects of the underlying Africa RiskView. Countries can also use the tool to select ARC risk transfer parameters to define how much of this modelled drought risk they wish to transfer to the ARC risk pool for each season. See the How ARC Works to read more about how these participation parameters are defined.
In addition to supporting ARC, the information produced by Africa RiskView has broader applications. It could help to target early food security assessments in specific geographic areas or help with contingency planning and emergency preparedness for future shocks in a country. It can also help to understand better the drivers and causes of food insecurity in an area and therefore inform the best investments or risk management strategies. The tool could also be helpful in guiding planning and investment decisions aimed at enhancing agricultural productivity or market development and be used to support micro insurance programmes. To date the tool focuses on drought, but work is ongoing to include risks due to river flooding and tropical cyclones.
Access to the Africa RiskView software is free and can be requested on the software webpage. Access is subject to the terms and conditions set forth in the Software License Agreement. Prospective users employed by national governments or organizations that have entered into a Software License Agreement benefit from the agreement and will be granted access to Africa RiskView. Prospective users who are not so employed will not be granted access to Africa RiskView. However, where appropriate, ARC and the relevant organizations can decide to enter in a Software License Agreement. All new registrations to the Africa RiskView are managed by the Africa RiskView Support Team.
The Africa RiskView models and software products are offered to countries working with the ARC Agency. Its models have been designed to be adapted and customised to work within national frameworks, strengthening existing systems and allowing governments and their partners to carry out their own risk analyses, as well as define their own risk management strategy and risk pool participation.
Africa RiskView uses several rainfall estimates (RFE) datasets, such as RFE 2.0 and African Rainfall Climatology v2 (ARC2) – both products of the Climate Prediction Centre (CPC) at the United States National Oceanic and Atmospheric Administration (NOAA) – but also others that continue to be added to as new dataset emerge. With these datasets, users can view actual rainfall estimates, calculate cumulative rainfall and analyse rainfall at several spatial scales. Historical rainfall datasets are available starting from the year 1983 to the present.
Africa RiskView uses the Water Requirements Satisfaction Index (WRSI) – an operational crop model originally developed by the United Nations Food and Agriculture Organization (FAO) and used widely throughout Africa and by many early warning institutions – as a meaningful indicator of how a shortage of rainfall may impact crop yields and the availability of pasture. The WRSI monitors water deficits throughout the growing season, and captures the impact of timing, amount and distribution of rainfall on staple annual rain-fed crops. In this way, Africa RiskView translates satellite-based rainfall information into a spatial drought index.
Africa RiskView then overlays this drought index on population vulnerability information to estimate drought-affected populations and response costs to assist them for past and current rainfall seasons. Vulnerable populations in ARC are determined by two factors — resiliency, which is indicated by a household’s distance from the national poverty line and exposure, which is the percentage of a household income that comes from agricultural activities (e.g. production, casual labor and livestock) that could be at-risk to drought. Household survey data reflecting these two dimensions are used to create a drought vulnerability profile of populations living in each administrative unit of a country. As countries customise the tool at the national level, the population vulnerability profiles are refined with the most up-to-date information as relevant national surveys and assessments become available.
In a final step populations affected are converted into responses costs based on a country’s response modalities and related budgeted contingency plans for ARC participation. These national modelled drought response costs are the underlying basis for ARC’s flagship parametric drought insurance products.
Every year since 2014 ARC has launched an insurance pool offering drought protection for its Member States for seasons starting in May and the year ahead based on Africa RiskView. It has been reviewed and customized by over ten countries to date and triggered four drought insurance payouts within the first two insurance pools.
In addition to drought, flood risk has also been identified as a high priority for many African countries. Developing flood insurance to help them respond more effectively to the needs of affected populations in the aftermath of a flood is a key ARC priority. The ARC River Flood Model (AFM-R) and ARC’s index-based river flood insurance product currently in development will be first of its kind in the world. With high resolution continent-wide flood extent depiction, the AFM-R model provides daily information on flood extent across Africa, with a focus on large-scale river flooding. Based on microwave data that can detect standing water in flooded areas, AFM-R has consistent historical data going back to 1992, is produced in near real-time and, when complete, will produce estimates of river flooding impact on countries and their vulnerable populations to underpin ARC’s parametric flood insurance product offering. The model, which will be integrated into Africa RiskView, is currently being tested and ground-truthed with ARC Member States.
Among ARC Member States, the island states of Madagascar, Mauritius and the Comoros are the most exposed to tropical cyclones, with the coastal areas of Mozambique also at risk. ARC’s Tropical Cyclone model covers the hazards of wind, storm surge and wave damage and converts these hazard maps into USD damage estimates per storm and per country in the South West Indian Ocean. Based on 70 years of historical cyclone records and a simulated track set for 1,500 years containing almost 10,000 cyclones, the model generates modelled loss calculations for these events – and for events that are being forecast as they develop – creating important information for early warning, disaster preparedness and planning purposes as well as for underpinning ARC’s parametric tropical cyclone insurance products. Already running and being used to track the 2016/17 tropical cyclone season in a stand-alone software tool, the tropical cyclone model will be fully integrated into Africa RiskView suite of products. Unlike in other regions where these models are deployed, the software allows users to customise some settings so that the model better fits their historic loss profiles.
Quality Assurance and Control
As ARC expands its operations, it continuously refines and develops the Africa RiskView software and underlying model to ensure that these tools continue performing at the highest standards. ARC continuously collects feedback and lessons learned from Member States to improve the functioning of Africa RiskViewin order to address country needs and better reflect on-the-ground realities.
To fortify its quality assurance, ARC consults external experts to ensure that Africa RiskView operates at the vanguard of climate risk modelling. The Africa RiskView Advisory Panel and the ARC Technical Review Forum are two advisory bodies that ARC convenes on an annual basis to evaluate Africa RiskView and provide insight on how to fine-tune the model and software. The Africa RiskView Advisory Panel and the ARC Technical Review Forum include leading experts in climate risk and drought modelling, insurance and reinsurance, information technology for development, and other relevant fields .
In order to minimize basis risk – the technical term used to describe the potential mismatch between ARC payouts and country needs – the Secretariat engages each potential participating country and its in-country partners in a year-long process to customise Africa RiskView to the respective national context, using local expert knowledge and information to adjust settings and assess the model’s performance and ability to capture the impact of drought events on vulnerable populations. The whole process will be carried out by national Technical Working Groups and supported by the Secretariat.
The ARC Technical Working Group perform customisations by adjusting settings and testing new settings by comparing the corresponding Africa RiskView outputs against available quantitative and qualitative datasets. This defines an improved set of parameters that lead to Africa RiskView results which perform well across a range of metrics defined by the ARC Technical Working Group. At the country level, many datasets that describe past drought events exist. These include:
- Quantitative datasets, such as yield data, statistics published by regular food security or needs assessments, or data on historical operations and beneficiaries assisted by government and its partners; and,
- Qualitative data on past weather and climate disaster events and other food security emergencies, sourced from the knowledge of the highly qualified ARC Technical Working Groups, which can serve as a practical balance to purely numerical datasets.
Given the range of available data on past crises, the ARC Technical Working Group has many ways to build a robust historical dataset of weather-related impact and need in the country, which therefore serves as a baseline against which to compare Africa RiskView outputs and justify its customisation choices. In addition to allowing countries and their partners to review the efficacy of Africa RiskView for their risk management needs, this process will also ensure that stakeholders understand how the model works – its inputs and its limitations – and how it can be used as the basis for triggering early payouts when required in severe weather-related emergencies.
Given the complex nature of food security in sub-Saharan Africa, there is chance that ARC payouts will not occur in years in which the food need is high if the main drivers of that food insecurity are not weather-related. One of the objectives of the year-long Africa RiskView customisation process is for countries to better determine the extent that exogenous factors like drought/flood/tropical cyclone are the predominant risk factor in their country and thus if participation in ARC will bring potential efficiency gains. The ARC team is confident that basis risk can be addressed as thoroughly and accurately as possible through Africa RiskView customisation processes and through open discussions on alternative risk management strategies and investments.
To further minimize basis risk, ARC established the Africa RiskView Customization Review Committee to provide an added measure of assessment for customisations carried out by ARC Member States. The Customization Review Committee examines member state customisations on national, regional, and continental levels, and the committee further provides recommendations to ARC Technical Working Groups on how customisations could be improved. This third-party evaluation strengthens the customisation process and helps to build confidence in Africa RiskView customisations and its outputs. A detailed explanation of the responsibilities of the Customization Review Committee can be found here.