Collaborative Research: CDI-Type II--The Open Wildland Fire Modeling E-community: A Virtual Organization Accelerating Research, Education, and Fire Management Technology

与野火相关的损失以及理解、预测和控制野火的愿望促使研究人员创建更详细和更强大的计算模型。然而，必要的复杂性和现实主义需要复杂的科学计算和分析过程，而这种过程本身已成为进一步进步和教育的障碍。此外，这项研究是由地理上分散的跨学科小组进行的，使他们的模型互操作是一个挑战。该项目致力于通过最先进的软件工程、可视化和协作工具来克服这种复杂性和协作面临的挑战。结果将在野火研究、教育、培训和管理中得到应用。该项目将提供基于网络的服务，在该服务中，访问者可以透明地与火灾行为模型、天气模型和数据源的组合进行交互，并提供一个软件环境，研究人员可以在其中贡献他们的软件和数据。这个门户将植入调查人员及其合作者的现有软件，吸引新的合作者，并作为一个充满活力的在线社区的焦点。调查人员已经获得了一些著名的火灾研究人员、教育工作者和管理人员的承诺。身临其境的、类似视频游戏的、动态的高性能可视化将作为可免费下载的PC客户端软件提供。门户背后的计算引擎将在本地计算机集群和远程超级计算机上运行。该项目将建立在数值天气预报、基于组件的软件工程和3D可视化的现有社区软件的基础上。它将通过提供通用的开源环境、新的可访问的数据存储库和协作工具来增加研究的影响。要实现这一愿景，需要在野地火灾的数学建模、软件工程和合作的社会模式方面取得变革性的进展。调查员将在这些领域最新成就的基础上，在每个领域作出更多重大贡献。数据同化的新方法将使将实时数据纳入野火模型成为可能。基于组件的软件的改进将有可能将计算和分析的复杂性分离到单独的模型中，并使连接这些模型的过程自动化。高度集成的可视化环境将有助于探索协作数据环境中的假设场景。耦合的模型将使人们能够更好地理解超越野火动态的复杂性。该项目将培养新一代接受网络培训的学生，为他们提供最先进的野火天气模型和高效火灾管理的实践经验。其基于网络的服务将在各级教育中受益。它将建立工具，以最大限度地减少世界各地跨学科小组之间、研究机构和教育机构之间以及研究和业务火灾预测之间的分离。该项目包括通过NCAR活动和活动面向普通公众和K-12教育，以及将电子社区纳入本科生和研究生课程。

Abstract The losses associated with wildfires and the wish to understand, predict, and control wildfires have driven researchers to create ever more detailed and powerful computational models. However, the necessary sophistication and realism requires scientific computing and analysis processes of a complexity that has itself become a hindrance to further progress and education. In addition, the research is done by geographically disperse interdisciplinary groups, and making their models interoperate is a challenge. This project strives to overcome this complexity and challenges to collaboration by state-of-the-art software engineering, visualization, and collaboration tools. The results will find applications in wildfire research, education, training, and management. The project will provide a web-based service in which a visitor can transparently interact with combinations of fire behavior models, weather models, and data sources, and a software environment where researchers can contribute their software and data. This portal will be seeded with existing software of the investigators and their collaborators, attract new collaborators, and serve as a focus for a vibrant online community. The investigators have already obtained the commitment of a number of prominent fire researchers, educators, and managers to participate. Immersive, video game-like, dynamic high-performance visualization will be provided as freely downloadable PC client software. The computational engine behind the portal will run on a local computer cluster and remote supercomputers. The project will build on existing community software in numerical weather prediction, component based software engineering, and 3D visualization. It will increase the impact of the research by providing a common open-source environment, new accessible data repositories, and collaboration tools. Fulfilling this vision will require transformational advances in mathematical modeling of wildland fires, software engineering, and social modes of collaboration. The investigators will build on their recent achievements in these areas to make significant additional contributions in each. New methods in data assimilation will make it possible to incorporate real-time data into wildfire models. Improvements in component-based software will make it possible to isolate the complexity of computing and analysis into separate models and automate the process of connecting these models. A highly integrated visualization environment will facilitate exploring what-if scenarios in a collaborative data environment. The coupled models will enable a new level of understanding of complexity that goes beyond the wildfire dynamics. The project will create a new generation of cyber-trained students provided with hands-on experience with state of the art wildfire-weather models and efficient fire management. Its web-based service will be beneficial in education at all levels. It will build tools to minimize the separation between interdisciplinary groups around the world, between research and educational institutions, and between research and operational fire forecasting. The project includes outreach to general public and K-12 education through NCAR events and activities, and integration of the e-community into undergraduate and graduate courses.