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

发展中国家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，预计未来几年气候变化将使风险显著增加。农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和投资于气候智能型高生产力农业的工具。然而，迄今为止，发展中国家农作物保险计划的成功和推广极为有限。发展中国家扩大农作物保险规模的问题有几个原因。传统的基于责任的保险计划需要耗费时间核实农民个人的实际损失，导致交易成本高、索赔纠纷和拖延，使农民不敢购买保险。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可靠的工具，以确定如果发生极端天气事件，保险应在何时以及在何种程度上向农民支付。参数保险，例如基于天气指数的保险，根据预先确定的气象指数和作物产量之间的关系触发赔付，从而消除了昂贵的作物损失评估的需要。然而，当前基于天气指数的保险面临的一个主要挑战是，赔付额往往与农民的实际产量损失相关性很差，这一问题被称为“基础风险”，这给使用指数保险减轻气候风险造成了重大障碍。在这种情况下，科学家如何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减少基础风险并更好地保护农民免受天气风险，从而改善目前基于指数的作物保险的不良表现。我们拟议的研究将开发新的基于天气指数的保险合同，通过结合作物生长模型，作物生长状态的卫星和智能手机图像以及空间天气变化的高分辨率网格估计，可靠准确地预测与天气相关的作物产量损失。重要的是，我们的工作将产生新的工具和方法，以满足指数保险行业的两个需求：（i）通过设计天气指数触发器来降低时间基础风险，这些天气指数触发器准确反映了产量对极端事件的敏感性在生长季节如何变化，以及（ii）通过利用捕捉天气条件、作物发育、实地条件和管理做法。与HDFC ERGO General Insurance合作，HDFC ERGO General Insurance是为印度各地的小农户提供基于天气指数的保险的主要供应商，我们将应用这些方法来设计和测试基于天气指数的新保险产品，以保护旁遮普和哈里亚纳邦（印度的粮仓）的主要农业州的农民免受极端温度和强降雨事件的综合生产风险。利用最近IFPRI-HDFC基于图片的作物保险（PBI）项目收集的独特实地数据，我们将进行农业经济影响评估，以量化基础风险的降低，农民福利的增加以及我们新合同相对于当前指数保险产品和政府面积产量保险计划的保险需求。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型小农农业。

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.