Collaborative Research: CMG: Geomagnetic Data Assimilation and Dynamo Modeling

了解地球磁场的起源仍然是地球科学中最具挑战性的问题之一，这个问题对于了解地球深部的动力学及其与地球系统其他部分的相互作用至关重要。近年来，在我们对该领域起源的理解方面，在两个不同的方面都取得了很大进展。首先，数值地球发电机模型已经取得了进展，数值模型产生了与地球磁场形态和时间行为相似的磁场。其次，新一代卫星磁场测量和地磁场建模的进展相结合，为地球发电机过程提供了新的观测洞察力。但是，这两种方法在很大程度上是相互孤立的。有必要进行协作研究，将地磁观测同化到岩心发电机模型中，以(1)更好地了解和预测地磁场的变化，并(2)提供对岩心磁场产生动力学的更多洞察。在这个项目中，将定义和开发一个数据同化系统，其中包括MoSST岩心动力学模型(模块化和可缩放模型)和地磁场观测。该系统基于集合卡尔曼滤波(EnsKF)，这是一种从预报集合估计预报误差协方差的最优同化方法。这种方法不需要建立发电机模型的伴随项，并降低了计算误差协方差的计算成本。这项研究的学术价值包括：(1)首次在地磁/地发电机领域建立同化系统；(2)通过地面观测和数值核心动力学模型预测地球磁环境的变化。拟议研究的更广泛影响包括：(1)将地磁观测应用于发电机数值模拟研究；(2)将核心动力学应用于地磁场模拟研究；(3)将核心动力学研究应用于其他地球物理研究，例如地球自转变化和时变重力场；(4)新的地磁同化教育/培训项目

Understanding the origin of the Earth's magnetic field remains one of the most challenging problems in Earth Science, a problem that is central to understanding the dynamics of the Earth's deep interior and its interactions with other parts of the Earth system. Recent years have seen great advances on two separate fronts in our understanding of the origin of the field. First, numerical geodynamo modeling has progressed to the point where numerical models generate magnetic fields of similar morphology and temporal behavior to the Earth's field. Second, the combination of a new generation of satellite magnetic field measurements and advances in geomagnetic field modeling are providing new observational insight into the geodynamo process. But, these two approaches exist in large part in isolation from one another. A collaborative research is necessary to integrate these approaches: assimilating geomagnetic observations into core dynamo modeling to (1) understand better and forecast variations of the geomagnetic field, and to (2) provide additional insight into the dynamics of magnetic field generation in the core.In this project, a data assimilation system will be defined and developed with the MoSST core dynamics model (a modular and scalable model) and observations of the geomagnetic field. The proposed system is based on the ensemble Kalman filter (EnsKF) which is an optimal assimilation method to estimate the forecast error covariance from an ensemble of forecasts. This method eliminates the need to develop an adjoint of the dynamo model, and reduces the computational cost of calculating the error covariances. The intellectual merits of the proposed research include (1) the first-ever assimilation system in geomagnetic/geodynamo field; (2) prediction of changes in Earth's magnetic environment with surface observation and numerical core dynamics model. The broader impacts of the proposed research include (1) application of geomagnetic observations to numerical dynamo modeling research; (2) application of core dynamics to geomagnetic field modeling research; (3) application of core dynamics research to other geophysical studies, e.g. Earth's rotation variation and time variable gravity field; (4) new education/training project in geomagnetic assimilation