Science Through Machine Learning: Quantification of Post‐Storm Thermospheric Cooling
Science Through Machine Learning: Quantification of Post‐Storm Thermospheric Cooling
复制标题
DOI：
10.1029/2022sw003189
复制
发表时间：
2022-06
期刊：
影响因子：
--
通讯作者：
R. Licata;P. Mehta;D. Weimer;D. Drob;W. Tobiska;J. Yoshii
中科院分区：
文献类型：
--
作者：
R. Licata;P. Mehta;D. Weimer;D. Drob;W. Tobiska;J. Yoshii
机器学习（ML）模型是通用的函数逼近器，如果使用正确，可以将观测数据集的信息内容以函数形式汇总，以供科学和工程应用。ML优于参数模型的一个好处是，没有关于特定基函数的先验假设，这可能会限制可以建模的现象。在这项工作中，我们在三个数据集上开发了ML模型：空间环境技术高精度卫星阻力模型（HASDM）密度数据库，来自Jacchia-Bowman 2008经验热层密度模型（JB 2008）的时空匹配数据集，以及来自CHAllenging Minisatellite Payload（CHAMP）的加速度计衍生密度数据集。这些ML模型与海军研究实验室质谱仪和非相干散射雷达（NRLMSIS 2.0）模型进行了比较，以研究中热层风暴后冷却的存在。我们发现NRLMSIS 2.0和JB 2008-ML都没有考虑风暴后的冷却，因此在强地磁暴之后的时期表现不佳（例如，2003年万圣节风暴（Halloween Storms）相反，HASDM-ML和CHAMP-ML确实显示了风暴后冷却的证据，表明这种现象存在于原始数据集中。结果表明，根据风暴的位置和强度，风暴后1 - 3天可能会出现高达40%的密度下降。
Machine learning (ML) models are universal function approximators and—if used correctly—can summarize the information content of observational data sets in a functional form for scientific and engineering applications. A benefit to ML over parametric models is that there are no a priori assumptions about particular basis functions which can potentially limit the phenomena that can be modeled. In this work, we develop ML models on three data sets: the Space Environment Technologies High Accuracy Satellite Drag Model (HASDM) density database, a spatiotemporally matched data set of outputs from the Jacchia‐Bowman 2008 Empirical Thermospheric Density Model (JB2008), and an accelerometer‐derived density data set from CHAllenging Minisatellite Payload (CHAMP). These ML models are compared to the Naval Research Laboratory Mass Spectrometer and Incoherent Scatter radar (NRLMSIS 2.0) model to study the presence of post‐storm cooling in the middle‐thermosphere. We find that both NRLMSIS 2.0 and JB2008‐ML do not account for post‐storm cooling and consequently perform poorly in periods following strong geomagnetic storms (e.g., the 2003 Halloween storms). Conversely, HASDM‐ML and CHAMP‐ML do show evidence of post‐storm cooling indicating that this phenomenon is present in the original data sets. Results show that density reductions up to 40% can occur 1–3 days post‐storm depending on the location and strength of the storm.