NSWP: Data-based Forecasting of the Geomagnetic Field with High Resolution in Space

磁场是控制磁层结构及其风暴时间动力学的基本参数。实现其及时、准确和可靠的预报是国家空间天气计划的主要目标之一。这对内磁层尤其重要，因为那里是磁暴和辐射带扰动发生的地方，也是目前第一原理模型能力最有限的地方。特别是磁场动力学是控制辐射带径向输运和加速的关键因素。最近开发的一项技术基于赤道流场的可扩展模型，该模型使用大量航天器数据，已被证明可以显著提高磁层磁场经验图像的空间分辨率。由于高空间分辨率所需的数据积累时间过长，可能会抹去重要的动力学效应，因此设计了一种新的非线性数据集技术，通过将模型拟合到局部数据子集来描述磁层各状态的空间结构。它既包括给定状态下的实际数据，也包括在全局参数、太阳风电场、地磁活动指数Sym-H及其时间导数空间中与当前状态相邻的其他时间间隔（如其他磁暴的相似相位）的数据。利用该模型获得的磁暴结构和动力学的初步结果与现场地球同步数据、IMAGE航天器观测数据以及从铱星星座数据推断出的场向电流图像一致，表明该技术为从过去事件中提取有关风暴时间电流和磁场的重要新信息提供了一种强大的新方法。该项目的目标是利用行星际介质数据作为唯一的模式输入，将目前的高分辨率模式转变为成熟的预报工具。当前模式中通过symm - h指数提供的关于磁层状态的缺失信息，将通过预测的symm - h以及对太阳风和IMF参数及其时间历史的更详细描述，在其预测版本中提供。这个项目将分三步完成。首先，预测的Sym-H或Dst指数，已经从现有的全球预测模型中获得，将被用作实际Sym-H指数的代理。其次，将阐述一种新的数据拟合程序，其中仅使用太阳风和IMF数据及其时间历史。第三，新工具将使用现场数据和基于实际Sym-H指数的全范围风暴的现有高分辨率模型进行验证和优化。该研究基于11年的GOES、IMP 8、Polar、Geotail、Cluster、ACE和Wind航天器观测数据，使用了迄今为止最大的原位空间磁场数据和并发行星际介质数据汇编数据库进行实证建模研究。如果有新的THEMIS任务的数据，也将添加到数据集中。这项研究的最终成果将是一套空间天气预报代码，该代码指定磁层磁场，其空间分辨率为几个地球半径，时间分辨率为亚暴时间尺度。为了提供快速预测，新的代码将被并行化，并在本地集群和超级计算机上进行测试。

The magnetic field is a fundamental parameter that governs the structure of the magnetosphere and its storm-time dynamics. Achieving its timely, accurate, and reliable forecasting is one of principal goals of the National Space Weather Program. It is especially important for the inner magnetosphere, where magnetic storms and radiation belt disturbances occur, and where the capabilities of the present-day first-principle models are most limited. In particular, the dynamics of the magnetic field is a key factor controlling the radial transport and acceleration of the radiation belts. A recently developed technique based on an extensible model for the field of equatorial currents that uses large sets of spacecraft data has been shown to dramatically improve the spatial resolution of the empirical picture of the magnetospheric magnetic field. Since the data accumulation, necessary for high resolution in space, may be too long and smear out important dynamical effects, a new nonlinear data-binning technique has been devised, where the spatial structure of each state of the magnetosphere is described by fitting the model to a local subset of data. It includes both the actual data obtained for the given state and data from other time intervals (e.g., similar phases of other magnetic storms), neighboring the present state in the space of global parameters, solar wind electric field, geomagnetic activity index Sym-H, and its time derivative. Initial results for magnetic storm structure and dynamics made with the model are consistent with in situ geosynchronous data, IMAGE spacecraft observations and the picture of field-aligned currents inferred from the Iridium constellation data, indicating that the technique offers a powerful new way to extract important new information on the storm-time currents and magnetic field from the past events. The goal of the project is to transform the present high-resolution model into a fully-fledged forecasting tool by using the interplanetary medium data as the only model input. Missing information on the state of the magnetosphere, available in the current model through the Sym-H index, will be provided in its forecasting version through a predicted Sym-H and through a more detailed description of the solar wind and IMF parameters and their time histories. The project will be done in three steps. First, the predicted Sym-H or Dst indices, already available from existing global forecasting models, will be used as a proxy of the actual Sym-H index. Second, a new data-fitting procedure will be elaborated, in which only solar wind and IMF data are used together with their time histories. Third, the new tool will be validated and optimized using in situ data and the already available high-resolution model based on the actual Sym-H index for the full range of storms. The proposed study uses the largest assembled database of in-situ space magnetic field data and concurrent interplanetary medium data ever compiled for empirical modeling studies, based on 11 years of GOES, IMP 8, Polar, Geotail, Cluster, ACE, and Wind spacecraft observations. When available, data from the new THEMIS mission will also be added to the data set. The final product of the study will be a set of space weather forecasting codes specifying the magnetospheric magnetic field with the resolution in space of a few Earth radii and the temporal resolution up to substorm time scales. To provide fast predictions the new codes will be parallelized and tested on local clusters and supercomputers.