CSISA Data Repository

In keeping with the open data policies of the U.S. Agency for International Development (USAID) and Bill & Melinda Gates Foundation, the Cereal Systems Initiative for South Asia (CSISA) has launched the CSISA Data Repository to ensure public accessibility to key data sets, including crop cut data- directly observed, crop yield estimates, on-station and on-farm research trial data and socioeconomic surveys. CSISA is a science-driven and impact-oriented regional initiative for increasing the productivity of cereal-based cropping systems in Bangladesh, India and Nepal, thus improving food security and farmers’ livelihoods. CSISA generates data that is of value and interest to a diverse audience of researchers, policymakers and the public. CSISA’s data repository is hosted on Dataverse, an open source web application developed at Harvard University to share, preserve, cite, explore and analyze research data. CSISA’s repository contains rich datasets, including on-station trial data from 2009–17 about crop and resource management practices for sustainable future cereal-based cropping systems. Collection of this data occurred during the long-term, on-station research trials conducted at the Indian Council of Agricultural Research – Research Complex for the Eastern Region in Bihar, India. The data include information on agronomic management for the sustainable intensification of cropping systems, mechanization, diversification, futuristic approaches to sustainable intensification, long-term effects of conservation agriculture practices on soil health and the pest spectrum. Additional trial data in the repository includes nutrient omission plot technique trials from Bihar, eastern Uttar Pradesh and Odisha, India, covering 2012–15, which help determine the indigenous nutrient supplying ability of the soil. This data helps develop precision nutrient management approaches that would be most effective in different types of soils. CSISA’s most popular dataset thus far includes crop cut data on maize in Odisha, India and rice in Nepal. Crop cut datasets provide ground-truthed yield estimates, as well as valuable information on relevant agronomic and socioeconomic practices affecting production practices and yield. A variety of research data on wheat systems are also available from Bangladesh and India. Additional crop cut data will also be coming online soon. Cropping system-related data and socioeconomic data are in the repository, some of which are cross-listed with a Dataverse run by the International Food Policy Research Institute. The socioeconomic datasets contain baseline information that is crucial for technology targeting, as well as to assess the adoption and performance of CSISA-supported technologies under smallholder farmers’ constrained conditions, representing the ultimate litmus test of their potential for change at scale. Other highly interesting datasets include farm composition and productive trajectory information, based on a 20-year panel dataset, and numerous wheat crop cut and maize nutrient omission trial data from across Bangladesh.

遵循美国国际开发署（USAID）和比尔及梅琳达·盖茨基金会（Bill & Melinda Gates Foundation）的开放数据政策，南亚粮食系统倡议（Cereal Systems Initiative for South Asia，CSISA）已推出CSISA数据仓库，以确保公众能够访问关键数据集，包括作物收获数据（直接观察）、作物产量估算、站内和农场研究试验数据以及社会经济调查。CSISA是一项以科学为驱动、以影响为导向的区域性倡议，旨在提升孟加拉国、印度和尼泊尔以谷物为基础的种植系统的生产力，从而改善食品安全和农民生计。CSISA产生的数据对于研究人员、政策制定者和公众等多个领域都具有价值和意义。CSISA的数据仓库托管于哈佛大学开发的开源网络应用程序Dataverse，用于共享、保存、引用、探索和分析研究数据。该仓库包含丰富的数据集，包括关于2009-17年期间作物和资源管理实践的站内试验数据，旨在为可持续发展的未来谷物种植系统提供支持。这些数据收集于在印度比哈尔邦的印度农业研究委员会（ICAR）东部地区研究综合体进行的长期站内研究试验。数据涵盖了关于作物系统可持续增密、机械化、多元化、可持续增密的未来方法、保护农业实践对土壤健康和病虫害谱的长期影响等方面的农艺管理信息。仓库中还包含2012-15年从比哈尔、东乌达耶普拉德什和奥里萨印度进行的养分缺失图技术试验数据，有助于确定土壤的本土养分供应能力。这些数据有助于开发针对不同类型土壤的最有效的精准养分管理方法。迄今为止，CSISA最受欢迎的数据集包括印度奥里萨州的玉米作物收获数据和尼泊尔的稻米作物收获数据。作物收获数据集提供了经过地面验证的产量估算，以及关于影响生产实践和产量的相关农艺和社会经济实践的有价值信息。从孟加拉国和印度还可以获取关于小麦系统的各种研究数据。不久还将上线更多的作物收获数据。仓库中包含与作物系统相关的数据和社会经济数据，其中一些数据与国际食物政策研究所（International Food Policy Research Institute）运行的Dataverse进行了交叉列出。社会经济数据集包含对技术目标定位至关重要的基线信息，以及评估在小型农户受限条件下CSISA支持技术的采纳和表现，这代表了它们在规模上产生变革的终极试金石。其他极具兴趣的数据集还包括基于20年面板数据的农场构成和生产轨迹信息，以及来自孟加拉国的多种小麦作物收获数据和玉米养分缺失试验数据。