FESA Micro-insurance: Satellite Indices Break-through

Andries Rosema, EARS

Weather index insurance is considered an important risk-sharing mechanism to assist farmers in resisting the vagaries of climate, to increase their production and escape from poverty. But the introduction of index insurance is hampered by lack of data. This situation, however, is rapidly changing. Dutch remote sensing company EARS has developed a 30 year climatic database on the basis of Meteosat, which covers the entire African continent. Using its relative evapotranspiration(RE) drought index, EARS  is successfully cooperating with a range of partners and developing  drought and excessive precipitation insurance across the African continent.

Introduction

FARMD (March 2012) | Farmers can raise their production and income considerably by applying improved tillage techniques, better seeds, fertilizer and pesticides. For this purpose most farmers need a loan. However, financial institutions are reluctant to offer credit. Because of climatic disaster, in particular drought, famers may lose their crop and would not be able to pay back the loan. Micro-insurance is the solution.

There are millions of smallholders with only a few acres of land and a production of say 1-2 tons of millet or maize annually. The loan to be insured may be in the order of $200 and the insurance premium some $20. Thus micro-insurance must be low cost. At the same time the index should closely represent the reduced crop production due to drought.  (Figure 1: Meteosat relative evapotranspiration (RE) used for drought insurance)

In the past, drought insurance has been based on precipitation. Indemnification of farmers would then take place if precipitation during the growing season would not meet certain pre-defined levels. In Africa, however, there are few rainfall stations. A very dense and costly network would be required to adequately represent the spatial variability. Moreover, adding rain gauges would not provide for a long precipitation history that is required to assess the drought risk and to price the policy.

Another limitation is that rainfall is not a good measure of actual crop water use. A considerable part may run-off or may percolate into the subsoil. It is also possible that rainwater is stored in the soil for considerable time and used by the crop with months of delay. Therefore, using rainfall data for micro-insurance involves considerable basis risk, both in terms of co-location of the data and representativeness for crop growth.

Climate data from space

After the declassification of remote sensing at the end of the 1990’s, a Dutch remote sensing school developed, specializing in the physical description and mathematical modeling of “object-sensor interaction”. EARS is a product of this school. The company is using visual and thermal infrared satellite data in mapping temperature, radiation, evapotranspiration and precipitation from space. Systems and services are provided for satellite based drought monitoring, crop yield forecasting and river flow forecasting. Operational systems have been or are being  implemented in China (4x), Mongolia and Niger. EARS is producing regional crop yield forecasts for Europe and Africa. 

In 2009 the company received a grant from the Minister of Development Cooperation to carry out the FESA Micro-insurance project, as a Netherlands contribution to the UN Millennium goals. The objective is to develop satellite based micro-insurance that reaches every farmer in Africa. In the framework of this project a 30 year retrospective data base of Meteosat hourly visual and thermal infrared data has been compiled and has been processed to daily and 10-daily climatic data fields. The methodology has been reported and extensive validation of the climatic data has taken place (Rosema et al. 2010).

FESA is not using precipitation for drought insurance, but the relative evapotranspiration (RE). Since evapotranspiration is hardly measured on the ground, this index is not yet well known. But it is the best possible agricultural drought indicator. RE is proportional to crop yield (Stewart 1973, Doorenbos and Kassam 1979).

The RE data developed in the FESA project covers the entire African continent at 3 km resolution. 30 year data series can be extracted for any location on the African continent. Thus FESA based insurance can truly reach every farmer.

Maize insurance pilot and scaling up

In early 2010 EARS completed it first drought insurance design for Maize growers in Burkina Faso and Mali. The project is carried out in cooperation with micro-insurance broker Planet Guarantee. The index was approved and priced by Swiss Re. The corresponding pilot project was carried during the 2011 growing season. EARS monitored the index and each 10 days a data report was provided with information on the start and progress of the season and the temporal development of the index.

During this pilot a remarkable event took place. Due to fragmented rainfall, some locations had a very timely start of the season, while others, not far away, were 40 to 50 days later. This analysis was confirmed by rain gauges that for the purpose of validation had been placed on the ground.  Based on this successful first pilot this insurance initiative is now being scaled up to more than 800 locations, covering farmers in large parts of Burkina Faso, Mali and Benin. 

Figure 2: RE 30 year data series at Dande, Burkina Faso 

Expanding activities

Cooperation with micro-insurance brokers is also expanding. Partnerships are developing with Syngenta Foundation, MicroEnsure, FSD-Kenya, Cardano and RMS. In cooperation with Syngenta Foundation a new challenge was met. In the foothills of the Aberdare range farmers grow French beans using water flowing from the mountains. It was shown that the RE index reflected well the overall water availability. Variations in RE were, with some phase shift, correlated to production. In this area, the FESA project also entered a new path: excessive precipitation insurance was developed, based on Meteosat derived cold cloud duration (CCD). The CCD is measure of the dwelling time of Cumulonimbus clouds (rainstorms). It is a good proxy of precipitation. Both the drought and the excessive precipitation insurance were approved and priced by re-insurer Swiss Re.

Partner MicroEnsure is carrying out drought insurance activities for the Tanzania Cotton Board. An inventory of existing rainfall data in the Bunda district showed these to be insufficient in quantity and quality.  EARS was requested to develop an RE based drought insurance structure. Since December 2011 the pilot is running and the RE index is monitored. MicroEnsure regional managers were trained in understanding the RE-index. As a spin-off, MicroEnsure also requested EARS to develop drought insurance for maize and rice growers in Rwanda. A second cotton insurance project has started in Kenya with Planet Guarantee.

Also in Kenya, EARS joined FSD-Kenya and World Bank ARMT to develop drought insurance for wheat growers in the district Narok. This concerns mechanized agriculture with property sizes of tenths to hundreds of hectare. A most profound design process took place with the objective to properly reproduce the drought events as witnessed by local farmers in recent years. The final RE based insurance structure did well reproduce the crop losses in recent drought years. The pilot is to start soon.

Current FESA activities also tend to extend further southward on the continent. RMS, from Newark, California, is studying the use of the RE index data for large scale maize insurance in Mozambique. In cooperation with risk manager Cardano from Rotterdam, drought insurance for rice growers in Madagascar is explored. 

Outlook

FESA micro-insurance is growing fast. It is expected that millions of African farmers may be insured in about 5 years. After due introduction in Africa, the technology may also be extended to other parts of the world, in particular to eastern Asia, where Japanese and Chinese geostationary meteorological satellites can provide 20-25 years of data. Geostationary meteorological satellites serve a large meteorological community and belong to the most reliable and stable satellites in the world. For Meteoat there is a permanent back-up satellite in orbit. Therefore reception of the data is assured. Consequently there is a stable basis for further deploying the FESA micro-insurance technology.  We are proud that, after many years of development, this innovative satellite remote sensing technology can play an enabling role in the development of large scale, affordable micro-insurance.

References

Stewart J L, and Hagan R M (1973) “Functions to predict effects of crop water deficits”, ASCE J. Irrigation and Drainage Division 99, p. 421-430.

Doorenbos J and Kassam A H (1979) “Yield Response to Water”, FAO Irrigation and Drainage Paper 33, FAO Rome, pp 193.

Rosema A, DeWeirdt M, Foppes S, Wilczok C (2010) “FESA Micro-insurance: methodology, validation, contract design”, Milennium project no 38, DG International Cooperation, Ministry of Foreign Affairs, the Netherlands. Publication EARS Earth Environment Monitoring BV, Delft, the Netherlands, pp 95.

Contact

Andries Rosema ([email protected]) or Jolien van Huystee ([email protected])