ITR/AP(Geo): Collaborative Proposal For First Generation Model And Data Assimilation System To Reduce Volcanic Hazards

该基金由美国国家科学基金会/信息技术研究小额资助计划和EAR/仪器和设施计划资助，将开发用于火山灾害评估的第一代数据和模型同化系统。从本质上讲，该系统将使用贝叶斯预测技术来构建危害估计，同时结合先前的地质信息和近实时数据类型。主要目标是通过信息技术方法的应用证明，即使在火山危机期间，这些不同的地质数据类型也可以顺利地结合起来，以提高对危害和风险的估计。重点将放在与火山灰分散和积聚有关的危害上，这既是因为这种火山危害的重要性，也是因为我们以前在预测火山灰危害方面的经验。模型将用于将在几座火山（科利马火山、塞罗内格罗火山、蒙特塞拉特火山、波波卡皮塔尔火山）观测到的火山地层学与一系列输入参数联系起来，包括喷发持续时间、初始气体含量和气象条件。我们计划探索几种技术（反演、简化模型、数据库开发），将观测到的火山变形模式直接与tephra沉降模型中使用的参数估计联系起来。该系统将使用并行算法构建，以利用廉价且广泛可用的PC集群。通过ITR拨款的支持，我们计划首次系统地将地质记录、过程模型和近实时监测结果联系起来，从而清楚地展示巨大的优势。具体地说，这种方法将进一步使人们清楚地认识到变形、岩浆过程以及控制毯体分散和聚集的基本参数之间的联系。从这个意义上说，这里提倡的系统方法的结果将为最有可能减少火山危害的火山学研究提供指导。这项研究将作为迈阿密大学、南佛罗里达大学和西南研究所（非营利）之间的合作提案进行。此外，与布里斯托尔大学的国际合作将为该项目带来关于破伤风扩散模型的专门知识和关于蒙特塞拉特重要火山爆发的数据。学生将参与项目的所有阶段，包括地质研究、模型开发和软件开发。经过适当的测试后，我们计划通过开源代码和向火山学界提供PC集群实施方面的帮助，使这个项目的结果广泛可用。＊＊＊

0130602ConnorThis grant, supported through funds from the NSF/Information Technology Research - Small Grants Program and the EAR/Instrumentation and Facilities Program, will enable the development of a first-generation data and model assimilation system for volcanic hazard assessment. Essentially, this system will use Bayesian forecasting techniques to construct hazard estimates that incorporate both prior geologic information and near-real-time data types. The primary goal is to demonstrate, through the application of information technology methodologies, that such disparate geologic data types can be combined smoothly, even during volcanic crises, to yield improved estimates of hazard and risk. The focus will be on hazards associated with the dispersion and accumulation of volcanic tephra (volcanic ash), both because of the importance of this volcanic hazard and our previous experience in forecasting tephra hazards. Models will be used to relate observed tephra-stratigraphies at several volcanoes (Colima, Cerro Negro, Montserrat, Popocateptal) to sets of input parameters, including eruption duration, initial gas content, and meteorological conditions. We plan to explore several techniques (inversion, simplified models, database development) to relate observed patterns of volcano deformation directly to the parameter estimates used in tephra fallout models. The system will be constructed using parallel algorithms to exploit inexpensive and widely available PC clusters. Through the auspices of an ITR grant, we plan to clearly demonstrate the huge advantages realized by, for the first time, systematically linking the geologic record, process models, and near-real time monitoring results. For tephra accumulation hazards specifically, this approach will further lead to a clear understanding of the links between deformation, magma processes, and the fundamental parameters governing tephra dispersion and accumulation. In this sense, the results of the systematic approach advocated here will provide guidance about the volcanological research most likely to results in volcanic hazard reduction. This research will be conducted as a collaborative proposal between the University of Miami, the University of South Florida, and Southwest Research Institute (nonprofit). In addition, international collaboration with the University of Bristol will bring expertise on tephra dispersion modeling and data on the important Montserrat eruptions to the project. Students will contribute to all phases of the project, including geological research, model development, and software development. After appropriate testing, we plan to make the results of this project widely available by open-sourcing the codes and providing assistance in implementation of PC clusters to the volcanological community. ***