Mitigating basis risk in weather index-based crop insurance: harnessing models and big data to enable climate-resilient agriculture in India

发展中世界数百万小农的生计正受到干旱、洪水和热浪等极端天气事件的威胁，由于气候变化，预计未来几年风险将大幅增加。作物保险保护农民免受极端天气事件带来的金融风险，并被广泛倡导为帮助农户摆脱贫困陷阱并投资于气候智能高生产率农业的工具。然而，到目前为止，作物保险计划在整个发展中世界取得的成功和被接受的程度极其有限。在发展中国家推广作物保险的问题可以确定为几个原因。传统的基于赔偿的保险计划需要耗时地核实个别农民经历的实际损失，从而导致高昂的交易成本、索赔纠纷和延误，使农民不愿购买保险。为了应对这些问题，政府和保险公司寻求开发更具成本效益和更可靠的工具，以确定如果发生极端天气事件，保险应该在什么时候以及在什么水平上向农民赔付。参数保险，例如基于天气指数的保险，根据气象指数和作物产量之间预先建立的关系触发赔付，从而消除了昂贵的作物损失评估的需要。然而，当前天气指数保险面临的一个主要挑战是，赔付金额往往与农民的实际产量损失缺乏相关性，这一问题被称为“基本风险”，这给使用指数保险缓解气候风险造成了重大障碍。在这种背景下，科学家如何才能为设计更智能的指数保险产品做出贡献，以满足农民、保险公司和政府的需求？该项目的总体目标是通过使用最先进的环境建模和大数据集来改善目前基于指数的作物保险业绩不佳的状况，以降低基数风险，并更好地保护农民免受天气风险的影响。我们拟议的研究将开发基于天气指数的新型保险合同，通过结合作物生长模型、作物生长状况的卫星和智能手机图像以及对空间天气变异性的高分辨率网格估计，可靠而准确地预测与天气有关的作物产量损失。重要的是，我们的工作将产生新的工具和方法，以满足指数保险部门的两个明确需求：(I)通过设计准确反映生长季节期间对极端事件的产量敏感度变化的天气指数触发器来降低时基风险，以及(Ii)通过利用捕捉天气条件、作物发展、田野条件和管理实践中的空间异质性的数据集，将空间基础风险降至最低。与HDFC ERGO General Insurance合作，我们将应用这些方法设计和测试基于天气指数的新保险产品，以保护旁遮普邦和哈里亚纳邦(印度的粮仓)主要农业邦的农民免受极端温度和暴雨事件的综合生产风险。HDFC ERGO General Insurance是为印度各地的小农提供基于天气指数的保险的主要提供商。利用通过最近的IFPRI-HDFC基于图片的作物保险(PBI)项目收集的独特的田间数据，我们将进行农业经济影响评估，以量化基础风险的减少、农民福利的增加以及与当前指数保险产品和政府面积收益保险计划相关的新合同对保险的需求。我们的工作将直接有助于提高印度农民指数保险的质量，更广泛地说，将为基于天气指数的保险提供科学基础，以更有效地支持发展中国家的气候智能小农农业。

Livelihoods of millions of smallholder farmers across the developing world are under threat from extreme weather events, such as droughts, floods, and heatwaves, with risks projected to increase significantly in future years due to climate change. Crop insurance protects farmers against financial risks posed by extreme weather events, and has been widely advocated as a tool to help farmer households to escape poverty traps and invest in climate-smart high-productivity agriculture. Yet, to date, the success and uptake of crop insurance schemes across the developing world has been extremely limited. Several reasons can be identified for problems in scaling crop insurance in developing countries. Traditional indemnity-based insurance schemes require time-consuming verification of actual losses experienced by individual farmers resulting in high transaction costs, claims disputes and delays that deter farmers from purchasing insurance. To counteract these issues, governments and insurers seek to develop more cost-effective and reliable tools to determine when, and at what level, insurance should payout to farmers if an extreme weather event occurs. Parametric insurance, for example weather index-based insurance, triggers payouts based on pre-established relationships between meteorological indices and crop yields, removing the need for expensive crop loss assessments. However, a major challenge for current weather index-based insurance is that payouts often are poorly correlated with farmers' actual yield losses, a problem known as 'basis risk', creating a major barrier to use of index insurance for climate risk mitigation.In this context, how can scientists contribute to the design of smarter index insurance products that meet the needs of farmers, insurers, and governments? The overall aim of this project is to improve the current poor performance of index-based crop insurance by using state-of-the-art environmental modelling and big datasets to reduce basis risk and better protect farmers against weather risks. Our proposed research will develop novel weather index-based insurance contracts that reliably and accurately predict weather-related crop yield losses by combining crop growth modelling, satellite and smartphone imagery of crop growth status, and high-resolution gridded estimates of spatial weather variability. Importantly, our work will produce novel tools and approaches that address two stated needs of the index insurance sector: (i) to reduce temporal basis risk by designing weather index triggers that reflect accurately how yield sensitivity to extreme events varies during the growing season, and (ii) to minimise spatial basis risk by exploiting datasets that capture spatial heterogeneity in weather conditions, crop development, field conditions and management practices. Working in collaboration with HDFC ERGO General Insurance, a major provider of weather index-based insurance for smallholder farmers across India, we will apply these approaches to design and test new weather index-based insurance products to protect farmers in the major agricultural states of Punjab and Haryana - the breadbasket of India - against combined production risks from extreme temperature and heavy rainfall events. Leveraging unique field data collected through the recent IFPRI-HDFC Picture-Based Crop Insurance (PBI) Project, we will conduct agro-economic impact evaluations to quantify reductions in basis risk, increases in farmer welfare and demand for insurance from our new contracts relative to both current index insurance products and government area-yield insurance schemes. Our work will contribute directly to improvements in the quality of index insurance for farmers in India, and, more broadly, will provide the scientific foundations for weather index-based insurance to more effectively support climate-smart smallholder agriculture across the developing world.