Prominence of the training data preparation in geomagnetic storm prediction using deep neural networks

Abstract

The direct interaction between large-scale interplanetary disturbances emitted from the Sun and the Earth’s magnetosphere can lead to geomagnetic storms representing the most severe space weather events. In general, the geomagnetic activity is measured by the Dst index. Consequently, its accurate prediction represents one of the main subjects in space weather studies. In this scenario, we try to predict the Dst index during quiet and disturbed geomagnetic conditions using the interplanetary magnetic field and the solar wind parameters. To accomplish this task, we analyzed the response of a newly developed neural network using interplanetary parameters as inputs. We strongly demonstrated that the training procedure strictly changes the capability of giving correct forecasting of stormy and disturbed geomagnetic periods. Indeed, the strategy proposed for creating datasets for training and validation plays a fundamental role in guaranteeing good performances of the proposed neural network architecture.

Renewability and Robustness Analysis and Review for Sustainable-Technology Propulsion Systems in Modern Transportation Infrastructure Administration

“…Such geomagnetic storms can affect modern electronics, power transmission, and communication equipment [120,121], damaging or temporarily disabling electronic circuits. Although prediction models using deep neural networks were successfully implemented [122], artificial intelligence prototypes are often prone to misleading identifications. Such situations would include falsely predicting an event or not detecting a real occurrence.…”

Ring Current Local Time Dependence during geomagnetic storms using equatorial Dst - proxies

“…Most models use solar wind data and prior Dst index as prediction parameters. Cristoforetti et al (2022) The authors were looking for triggering mechanisms for these large storms. They discovered there is a correlation between the electric field measured at L1 and the Dst-index.…”

Classifying Interplanetary Discontinuities Using Supervised Machine Learning

“…The complex nature of the solar‐terrestrial system imposes more advanced tools to be used in computational space physics. In recent years, there has been a clear growth of published articles on applied machine learning (ML) techniques in space plasmas, such as solar wind characterization and prediction (Li et al., 2020; Upendran et al., 2020), space whether research (Camporeale, 2019; Camporeale et al., 2018), forecasting radiation belt dynamics (Bernoux et al., 2021), and geomagnetic storm prediction (Cristoforetti et al., 2022). ML algorithms can be used to build models based on a training data set, and then try to make predictions without being explicitly programmed how to do so.…”