The Technical Centre for Agricultural and Rural Cooperation (CTA) shut down its activities in December 2020 at the end of its mandate. The administrative closure of the Centre was completed in November 2021.
Leading image

Weather Index-Based Insurance for the Pacific

Viti Levu, Fiji

Smallholder farmers in the Pacific have no access to weather index-based insurances, while flooding is a real threat for them. Preliminary research in the region suggests that weather and agricultural data, and the exact locations of farmers is weak in the region.

Catastrophe risk insurance can provide quick payouts in the wake of a major disaster. The Pacific Catastrophe Risk Insurance Pilot (PCRIP) was a pilot that tested a multicounty index-based risk pooling programme. It was established by Marshall Islands, Samoa, Tonga, and Vanuatu to increase their financial resilience against natural disasters, like tropical cyclones, earthquakes and tsunamis. The World Bank acted as an intermediary between those countries and a group of reinsurance companies. Payouts are triggered by specific physical parameters for the disasters (e.g. wind speed and earthquake ground motion) taken from the Joint Typhoon Warning Centre and the US Geological Services (USGS). 

The pilot made two payouts for an aggregate amount of US$3.2 million, in each case within 10 days of the disaster. Tonga received a payout of US$1.3 million following Tropical Cyclone Ian in January 2014. Vanuatu received a payout of US$1.9 million following Tropical Cyclone Pam in March 2015. The payouts were the first injections of cash received in the immediate aftermath of the disaster. These events demonstrated the pilot fulfilling its purpose: to provide governments with a quick, but limited, cash injection in the aftermath of a major disaster to finance immediate expenditures.

However, there is an urgent need to complement the catastrophe risk insurance with other financial solutions to cover more frequent, less severe events. For example, the small island states of the Pacific consist of a potential for weather index-based agricultural insurance. With as much as 80% of the population of some countries in the Pacific involved in agriculture, there is a potential for an affordable weather index-based agricultural insurance that could provide some form of coverage in the case of failure of crops. Given the low levels of income earned by farmers, traditional agriculture-based insurance products would be too expensive to meet these requirements. Until now, there are no existing index-based insurance schemes in the region. Fiji is the most advanced country in the region that works on developing an index-based insurance product. 

Risk of flooding

In a report by the Agriculture Ministry in Fiji, it is noted that the worst damage for farmers could come from a flooding submerging crops for 2 or 3 days. It also states that after such event root crop prices could increase 2-3% and vegetable prices could increase 5-80%. This represents a double impact to a farmer – they have less income due to the failure of their crops and they have to pay more for food due to lack of supply. The report also states that if a weather index based agricultural insurance was developed to mitigate the risk of such events, it would be based on the principle of “low premium, low payouts”, offering just enough coverage to provide resilience to adverse agricultural events at an affordable price. It is most likely that this product would be based on abnormally detrimental rainfall during critical growing periods during the year.

The country’s capacity to collect, analyse, and report on weather data in support of any weather-based index must come from the Fiji Meteorology Service (FMS), a department of the Ministry of Rural and Maritime Development and National Disaster Management. FMS maintains the national data base with historical data on Cyclone intensity and tracks, rainfall events and drought. In addition to Fiji Met there are regional bodies based in Fiji that maintain weather data including SOPAC at South Pacific Regional Environment Project (SPREP) and the University of the South Pacific (USP). The Fiji Meteorological Office also maintained 38 weather stations collecting rainfall and other data, 29 of which were manually operated and 14 of which were automated (9 stations had both). Stations report data daily, and information is stored electronically, but often there are stations that do not report as a result of the irregularity of collection of this data by caretakers, or due to malfunction among automated ones. 

Risk modelling

In 2012, the Pacific Financial Inclusion Programme (PFIP) assessed the feasibility of weather-index insurance. It concluded that the weather station density, as well as the reliability of the weather information infrastructure, was not sufficient to develop detailed rainfall-indexed products. The Applied Geosciences and Technical Division of SPC (SOPAC) have advised that most of Fiji does not have any topographic flood plain data, and where there is data, the highest resolution is 20 meters. This resolution is probably not refined enough to use for developing an insurance model and the development of sophisticated flood risk modelling is time-consuming and expensive. As a guideline for developing weather index-based agricultural insurance, there needs to be at least 20 years of historical daily data and the missing data should not exceed 3% of the total daily data set. It is also advisable that the weather stations are automated. The data available in Fiji does not meet these requirements. 

The FMS can theoretically provide this data on a case by case basis for no charge to those who submit official requests; however, the process is time-consuming. This also represents two issues. Firstly, it is unclear whether the data available is of high enough quality and detail to develop an index-based insurance. Secondly, if a product was developed, the ability for an underwriter to get timely information from the FMS weather stations to trigger claims payments is doubtful. Timely payment of claims is essential for weather index-based insurances to be successful as farmers who suffer a loss need the payout as soon as possible.

Furthermore, demographic and agricultural data in Fiji are weak. PFIP could not locate any large-scale mapping showing concentrations of farming communities. Further, there are few examples of monoculture farming in Fiji, or farming on large scales. The average size per farm in Fiji is small at 3.9 hectares and farmers generally plant a mix of different crops. The existence of monoculture farming is a key requirement to development of weather index-based insurance schemes as different crops have different tolerances to weather. The mix of different crops planted by farmers in Fiji makes the designing of a policy with set loss triggers difficult. 

The one agricultural product that does have data is sugar cane. Fiji produced 2.3 million tons of sugar cane in 2008, on 50,907 hectares of land. Crushing and exportation of the cane is managed by the Fiji Sugar Corporation Limited (FSC), which oversees the 14,000 active members of the Sugar Cane Growers Council (SCGC). The SCGC has are interested to develop weather index-based insurances, but struggle with capacity constraints. 

Information in this article was mainly derived from the Pacific Financial Inclusion Programme’s Focus Notes publication “Weather Index-Based Insurance in Fiji: Brief on Initial Scoping” (March, 2012), authored by Barry Maher and Michael McCaffrey. Currently, CTA works on a new study that investigates the conditions, readiness and potential for weather index-based agricultural insurance in seven countries in the Pacific region.

Read More


Two innovative enterprises have integrated a weather service system within an agro advisory service for farmers in East Africa. The eProd handheld device collects the GPS locations and agronomic information such as soil type, seed variety and planting date. aWhere combines this information with their weather data so farmers can now be sent SMS weather forecasts, spray alerts, fertiliser advice and yield projections.


Providing added value services for smallholders using open weather data in developing countries is challenging. Therefore, on 21 and 22 November 2017 practitioners, policy-makers and academics gathered in The Hague, the Netherlands, to explore in two workshops the practical and strategic challenges they face to work with open weather data and how to address them.


Ensuring that data can be easily accessed, used and shared requires the use of data standards. If you are currently working on a data project you should take time to consider what standards might be available to you to help to achieve the goals of your project.


Climate warming affects the water cycle, which impacts negatively on agricultural production and derails the cyclical effects associated with weather predictions and agricultural seasonality. Making use of weather data could help farmers to mitigate to the circumstances and increase farm productivity. To succeed, pragmatic public-private multi-stakeholder partnerships are required.


Availability of and access to climate data and information products is critical to achieving climate resilient development. However, climate information is not widely used in Africa. Useful information is often not available or, if it does exist, is inaccessible to those that need it most. Efforts are being made to alleviate the problem of data availability and use.

Smallholder farmers in the Pacific have no access to weather index-based insurances, while flooding is a real threat for them. Preliminary research in the region suggests that weather and agricultural data, and the exact locations of farmers is weak in the region.

External links


Past issues

ICT Update N. 91

Next-generation ACP agriculture - innovations that work

ICT Update N. 90

Women and Digitalisation in Agriculture

ICT Update N. 89

Data4Ag: New opportunities for organised smallholder farmers

ICT Update N. 88

Unlocking the potential of blockchain for agriculture

View all