CAREER: Development of Computational Methods for the New Generation of Air Quality Models

空气质量计算机模型是环境研究中广泛使用的调查工具，模拟了许多物理和化学过程，并研究了它们对大气污染物浓度的综合影响。这些模型也是监管和政策界的重要工具，用于根据国家环境空气质量标准的要求制定最优的大气污染物排放控制策略。模拟结果的可信度严重取决于所采用的数值方法的准确性，以及基础实施的稳健性。这项研究将开发最先进的计算方法和软件技术，用于空气质量建模工作中的时间步长方法。将开发特殊用途的算法，用于同时处理箱式模型过程，包括气溶胶动力学、气相和液相化学以及相间传质。更高的精度和更少的计算时间是目标。颗粒物(气溶胶)过程最近已成为环境科学的优先重点领域。在模型中加入气溶胶过程会导致总体计算时间增加数量级。这项研究将扩展到求解控制粒子演化的积分-微分方程组的领域；具体地说，该工作将完善理论框架，并将为气溶胶动力学-化学模拟建立快速可靠的数值技术。与物理/化学过程理解和数值算法设计相结合，软件技术是这些模型的重要组成部分。该项目的另一个目标是探索软件设计技术和工具，使建模软件更易于使用、维护和开发。由此产生的通用欧拉模型的面向对象版本将说明软件构建的新方法，并将构成开发和测试专门数值方法的理想环境。教育目标是加强面向计算的学科的跨学科教育。由于未来的许多科学技术和许多新兴行业将跨越传统学科的边界，跨学科教育变得越来越重要。一些最新的高需求领域涉及计算和信息技术以及其他领域：教学和研究将支持密歇根科技大学与计算相关的尖端理科学士项目的发展，并继续为计算科学和工程博士项目做出贡献。

Air quality computer models are widely used investigation tools in environmental research;many physical and chemical processes are modeled and their integrated impacts on atmospheric pollutantconcentrations studied. These models are also important tools for regulatory and policy communities, andare used to develop optimal emission control strategies for atmospheric pollutants, as required bythe National Ambient Air Quality Standards.The level of confidence in modeling results depends critically on the accuracy of numerical methodsemployed, as well as on the robustness of the underlying implementation. This research willdevelop state-of-the-art computational methods and software techniques for use in air quality modelingwork on time stepping methods. Special purpose algorithms will be developed for simultaneous treatment of box model processes, including aerosol dynamics, gas and liquid phase chemistry and interphase mass transfer. Higher accuracies and lower computational times are targeted.Particulate matter (aerosol) processes have recently became a priority focus area in environmental science. Incorporating aerosol processes in the models leads to an order of magnitude increase in the overall computational time. The research will extend into the area of solving the integro-differential equations thatgovern particle evolution; specifically, the proposed work will improve the theoretical framework and willbuild fast and reliable numerical techniques for aerosol dynamics-chemistry simulations.Together with the physical/chemical process understanding and numerical algorithm design, software tech-nology is an important component of these models. Another objective of the project is to explore softwaredesign techniques and tools that will make the modeling software easier to use, to maintain and to develop.The resulting object-oriented version of a generic Eulerian model will illustrate the new approach to soft-ware construction and will constitute an ideal environment for developing and testing specialized numericalmethods.The educational goal is to enhance interdisciplinary education in computationally-oriented disciplines.Since much of the science and technology of the future and many of the new industries will cross theboundaries of traditional disciplines, interdisciplinary education becomes increasingly important. Some ofthe newest high demand areas involve computing and information technology together with other fields:teaching and research will support the development of cutting-edge computationally-involved B.S. programs at Michigan Technological University, and continue to contribute to the Ph.D. program in Computational Science and Engineering.