Geospatial Resource for Agricultural Species and Pests with integrated workflow modelling to support Global Food Security (GRASP-GFS): a prototype

获取广泛的信息，从严谨的科学结果到“听说”农民的知识，对于能够有针对性地在社区或国家层面实现粮食安全规划的努力变得至关重要。此外，在不断变化的气候和市场中设计科学的干预策略是一项根本性的挑战。移动通信、传感器平台、空间搜索和普适计算等数据收集技术的发展正在从根本上改变农业研究。然而，需要将原始数据转化为有用的情报和知识以改善地球环境、经济和社会福祉的跨学科研究仍然受到学科和组织筒仓和遗留概念的限制，并且不存在或不严格的方法来量化任何收集的数据集中固有的不确定性。grip - gfs将使用地理空间锚定的“基因型”数据库整合原则来查询这些多维数据信息，包括论文、报告、土著、社会经济和农民的知识。该框架将通过在选定的粮食安全地理工作流模型中使用不同组成部分的质量加权描述符，实现不确定性评估。由地理空间科学方法驱动的跨学科专业知识将用于开发与粮食安全相关的所有学科的综合框架。驱动焦点将是农业物种种质的基因型特征，数据按地理空间起源排序，更高层次的描述符为“农业性状”。一个特别新颖的方面是将气候记录或情景和土地状况数据与已知（和新的）作物、动物和农业重要微生物物种特征来源相结合。这将使人们能够从新的角度看待遗传多样性，确定新的种质资源和性状变异来源，通过结合农业生态建模和匹配原则来定位合适的种质资源，通过考虑当地市场和农民知识的附加信息来规划可能产生最大影响的育种目标。这些建模能力将来自于将上述每个模型框架在一个通用的方法中，该方法允许基于数据和过程的语义描述组合工作流，并对不确定性/错误传播进行工作流质量评估。以英国小麦作物和马来西亚班巴拉花生为模型的两个用例将展示利用作物特定数据和处理模型来预测这两种作物的地理空间性状变异的方法。为支持马来西亚的未来作物研究中心（CFFRC），包括主要病虫害在内的农业物种综合地理空间平台将使未来的研究人员能够在为原型数据库的人口做出贡献之前，根据其主题要求制定数据处理和整合。从传感器网络到遥感的数据捕获，包括来自农民的群体外包，将进一步整合到GRASP平台中，允许对工作流程建模和多尺度情景风险评估进行其他改进。使用开放地理空间联盟（OGC）开发的开放标准和互操作性原则，平台可交付的软件将在开源许可下发布，以实现广泛使用和进一步开发，并确保项目的可持续性。将发布带有汇编的当前数据库（带有更新设施）的桌面界面和网络界面，使广大受众即使在互联网连接薄弱的偏远地区也能使用。grip - gfs旨在与其他全球倡议，如GEOSS（全球地球观测系统）和GeoNode联系起来，以发展实验、经济和社会科学家之间富有成效的互动。

Access to a wide range of information, from rigorous scientific results to 'hear-say' farmer's knowledge is becoming critical to be able to target efforts in achieving food security planning at community or country levels. Also, designing scientific and intervention strategies within changing climates and markets is a fundamental challenge. Developments of technologies for data collection in mobile communications, sensor platforms, spatial search and pervasive computing are fundamentally changing research in agriculture. However, inter-disciplinary research needed to transform raw data into useful intelligence and knowledge to improve the planet's environmental, economic and societal well being is still constrained by disciplinary and organisational silos and legacy concepts and an non-existent or non-rigorous approach to quantifying the uncertainty intrinsic in any collected dataset. GRASP-GFS will use a geospatially-anchored 'genotype' database integration principle to query such multidimensional data information, including papers, reports, indigenous, socio-economic and farmer's knowledge. This framework will enable uncertainty assessments through the use of quality weighting descriptors of the different components within a chosen geo-workflow model for food security.Cross-disciplinary expertise driven from geospatial sciences methodologies will be used to develop this integrating framework across all subjects relevant to Food Security. The driving focus will be the agricultural species germplasm for genotype characteristics with the data ordered by geospatial origin with the higher level descriptor being the 'agricultural trait'. A particular novel aspect is the combined use of climate records or scenarios and land ground condition data with known (and new) sources of traits in crop, animal and microbial species of agricultural importance. This will permit new perspectives on genetic diversity, identifying new sources of germplasm and sources of trait variation, geolocating suitable germplasm by a combination of agro-ecological modelling and matching principles, planning breeding objectives with the greatest likely impact by taking into acccount the added information of local market and farmer knowledge. These modelling capabilities will come from framing each above model within a generic approach allowing workflow composing based on semantic description of data and processes and workflow quality assessment for uncertainty/error propagation. Two use cases modelling with wheat crop in the UK and bambara groundnut in Malaysia will demonstrate the approach with crop specific data and processing models to forecast geospatial trait variation for these two crops.Supporting the Crops for the Future Research Centre (CFFRC) in Malaysia, the GRASP integrated geospatial platform for agricultural species, including major pests and diseases, will allow future investigators to shape the data handling and integration according to their subject requirements, before contributing to the population of the prototype database. Data capture from sensor network to remote sensing, including crowd-sourcing from farmers will be further integrated within the GRASP platform allowing other refinements of the workflow modelling and multiple scale scenario risk assessments.Using open standards and interoperability principles developed by the Open Geospatial Consortium (OGC), the platform deliverable software will be released under open source license to enable wide use and further developments also ensuring sustainability of the project. Both desktop interfaces and web interfaces with compiled current databases (with updating facilities) will be released, enabling wide audience usage even from remote places with weak internet connections. The GRASP-GFS aims to link with other global initiatives such as GEOSS (Global Earth Observation System of Systems) and GeoNode to develop productive interaction between bench, economic and social scientists.