Although human civilisation is built on farming, through the course of ages farming has remained a risky business. Bad harvests are particularly harmful to small farmers in developing countries and have even resulted in famine in some regions. That is why international financial institutions and developing agencies have sought poverty alleviation mechanisms that reduce the risk of farming and that go beyond relief operations and hand-outs. Indexed insurance is one of the possible solutions.
The idea is simple: one chooses one or more parametres, determines a variety of thresholds (or just one) and if the value of the parametre goes over that threshold (or stays under it, depending on the definition), the farmer gets a pay-out, regardless of whether his crop has actually been damaged or not. This approach has the advantage that (after the pilot phase) virtually no field checks or other control operations are required, thus reducing costs and making the whole undertaking commercially viable.
EARS, a Dutch company, in cooperation with a number of (re-)insurance companies has developed a system for indexed agricultural insurance that is based on evapotranspiration. Evapotranspiration is assessed by analysing Meteosat Second Generation (MSG) satellite images. These images have the advantage of a very high temporal resolution and complete coverage of the target area; thus providing all the information that is needed. Supported by the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT), EARS has access to a 32-year database of hourly visual and thermal Meteosat imagery. This enables the use of historical time-series to assess the correct establishment of the evaporation parametre that is the threshold for pay-out. After all, if this is not done correctly, either the insurance is not attractive to farmers or the insurance company goes broke. One of the basic concerns of farmers is that they can at least recover enough funding to buy the inputs for next year’s crop. After all, the tragedy of famine is prolonged and aggravated when the affected population is forced to consume all they have and no seeds are left for the next cropping season.
Evapotranspiration is chosen as index, because it has certain advantages over alternative solutions. A possible alternative is the Normalised Difference Vegetation Index (NDVI) that gives an estimate of the (development of) total biomass. It measures total vegetation cover and not land under cultivation. Rainfall is another alternative index that is commonly used, but intensities and amounts may vary considerably over a small distance. As with NDVI, the relation with crop growth is not very clear and intensive campaigns on the ground are needed for calibration, taking into account that rain gauge data are not very reliable in Africa. Evapotranspiration represents actual crop water use and is therefore more closely related to crop yield than precipitation. Processing satellite data with the Energy and Water Balance Monitoring System (EWBMS) is therefore the most promising solution. The system is developed by EARS since the early 1980’s. It generates temperature, radiation, evapotranspiration, cloudiness and rainfall data fields, covering the African continent at 3-km resolution.
Overview of the energy and water balance monitoring system
Two insurance indices are derived from the Meteosat data. One, dekad relative evapotranspiration (RE), as an agricultural drought index and second, dekad cold cloud duration (DDE), as an excessive precipitation index.
RE is the actual evapotranspiration (the amount of water actually used by plants and evaporated by the soil) divided by the potential evapotranspiration (the amount of water that would be used by plants (and soil) under optimum growing conditions). If a crop is close to or at potential evaporation during the different stages of the growing season, this will lead to increased yields. Indirectly, relative evapotranspiration is therefore also an indicator of crop health and the effect of (or potential for) fertiliser use, which rainfall is not.
Cold cloud duration (CCD) is a measure of cumulonimbus dwelling time. In the tropics, almost all rainfall is from very high clouds, making it a suitable index for the total amount of rainfall. CCD is expressed as a fraction of time and can be used as indicator for excessive precipitation insurance.
To translate the physical model into insurance products, additional information is needed on the growing season phasing (EARS applies three phases), the timing of the growing season and the start of the growing season window. A trigger of a relative evapotranspiration of 65% works well in practice to determine the start of the growing season. The payout is ideally based on the relative yield that is lost, because of drought. There are different modalities of insurance and calculation methods, but relative evaporation serves as a good proxy for yield loss. The most basic form of insurance is where a farmer recovers the costs of his original input, which enables the investment needed for the next growing season. The insurance company indicates the level of risk coverage. EARS subsequently develops the index insurance, including design parameters such as strike (threshold where payout starts), exit (threshold where payout stops, or maximum payout), burning costs and pure risk premium (average payout) to arrive at a product that is viable for both the farmer and the company. Thanks to the historical archive, these parameters can be substantiated in a reliable way.
The indices provide a relatively low intrinsic and spatial basis risk. The data are uniform, objective and abundant. They can be produced economically, which is important, because the paying capacity of the potential clients is limited. There is no need for extensive ground measurement networks and the approach can be scaled up easily. Pilot projects have been carried out for maize and cotton in Mali, Burkina Faso and Benin (with Planet Guarantee), for French beans in Kenya (with Syngenta Foundation), for cotton in Tanzania, for maize and rice in Rwanda (both with MicroEnsure), for maize in Malawi and Mozambique (with COINRE), for cotton and sorghum in Kenya (with Planet Guarantee and others) and for wheat also in Kenya (with FSD Kenya and the World Bank).
Andries Rosema, EARS Director, confirms that the results of the pilots were very positive: more than 20,000 farmers were insured and the programme is expanding rapidly. The availability of reliable measurements and historical data, the cost of data and the possibility of scaling operations are all limiting factors when using rainfall as an index but are no limitations to evapotranspiration from MSG images at all. There are some demand-side issues that still need to be resolved: the ease of understanding and trust in the concept (by the target groups) do not come as naturally with evapotranspiration as with rainfall. Taking this into account, the application of index-based micro-insurance for agriculture is expanding rapidly: already in 2015 more than 300,000 farmers are expected to be covered, albeit mostly still in the form of projects.
Published under the title 'Crop Insurance for Every Farmer in Africa' in Geospatial World Magazine April 2014