Domain of Influence analysis: implications for Data Assimilation in space weather forecasting

TL;DR

This paper applies Domain of Influence analysis to space weather models to identify optimal observation locations and improve data assimilation, enhancing forecasting of solar-terrestrial interactions and CME impacts.

Contribution

It introduces the use of DOI and RA analyses on space weather models to determine key observation points and model characteristics for better data assimilation.

Findings

DOI analysis highlights critical observation regions like the magnetospheric plasma sheet.

Time-dependent models are more useful for data assimilation than static models.

Initial steps toward identifying key heliospheric parameters for CME prediction.

Abstract

Solar activity, ranging from the background solar wind to energetic coronal mass ejections (CMEs), is the main driver of the conditions in the interplanetary space and in the terrestrial space environment, known as space weather. A better understanding of the Sun-Earth connection carries enormous potential to mitigate negative space weather effects with economic and social benefits. Effective space weather forecasting relies on data and models. In this paper, we discuss some of the most used space weather models, and propose suitable locations for data gathering with space weather purposes. We report on the application of \textit{Representer analysis (RA)} and \textit{Domain of Influence (DOI) analysis} to three models simulating different stages of the Sun-Earth connection: the OpenGGCM and Tsyganenko models, focusing on solar wind - magnetosphere interaction, and the PLUTO model, used to simulate CME propagation in interplanetary space. Our analysis is promising for space weather purposes for several reasons. First, we obtain quantitative information about the most useful locations of observation points, such as solar wind monitors. For example, we find that the absolute values of the DOI are extremely low in the magnetospheric plasma sheet. Since knowledge of that particular sub-system is crucial for space weather, enhanced monitoring of the region would be most beneficial. Second, we are able to better characterize the models. Although the current analysis focuses on spatial rather than temporal correlations, we find that time-independent models are less useful for Data Assimilation activities than time-dependent models. Third, we take the first steps towards the ambitious goal of identifying the most relevant heliospheric parameters for modelling CME propagation in the heliosphere, their arrival time, and their geoeffectiveness at Earth.