Key points
AMPERE field-aligned currents have been combined with the SAMI3 model to estimate the high-latitude potential
Independent validation of the new technique using satellite drifts and ground-based TEC indicates good agreement overall
The SAMI3 model performs better in this case when using AMPERE-derived potentials than when using the empirical Weimer potential

1. Introduction

Over the past several decades, great progress has been made in specifying the ionosphere through first-principles modeling and data assimilation. Ionospheric specification at high latitudes remains challenging because of the influence of the solar wind and magnetosphere on the ionosphere, especially in terms of plasma convection (e.g. Spencer and Mitchell, 2007; Bust et al., 2007; Chartier et al., 2019), as well as the relatively poor data coverage compared to other regions. The northern high latitudes are becoming increasingly relevant in terms of space weather due to the opening of the Arctic to shipping and mineral exploration. This is reflected in a focus on the region from several U.S. government agencies, notably the Department of the Air Force in its Arctic Strategy (Barrett et al., 2020) and the National Science Foundation’s Big Idea: Navigating the New Arctic. The lack of ground networks in this area results in increased reliance on wireless signals, which are susceptible to ionospheric disturbances.
Data from the Active Magnetosphere and Planetary Electrodynamics Response Experiment (AMPERE, Anderson et al., 2000; 2002, Waters et al., 2001; 2020) provide the most reliable and ubiquitous means of specifying high-latitude ionospheric electrodynamics, with field-aligned current measurements available from over 70 satellites in six polar low-Earth orbits, yielding global measurement coverage at all latitudes every 10 minutes. The AMPERE dataset allows for the resolution of the full Region 1 and Region 2 current system in near real time and across both northern and southern polar caps (e.g. Anderson et al., 2008). Understandably, this dataset has featured prominently in many recent efforts to specify the high-latitude electrodynamics, such as Assimilative Mapping of Ionospheric Electrodynamics (Cousins et al., 2015) and Assimilative Mapping of Geospace Observations (Matsuo et al., 2019).
To determine the high-latitude potential from the field-aligned current distribution, it is necessary to know the ionospheric perpendicular conductances (Pedersen and Hall). The most important determining factors of the conductances are the solar EUV flux, the average energy and energy flux of energetic particles and the neutral upper atmospheric density and composition. Several approaches exist to address this problem. The simplest is the assumption of constant conductance (used by Merkin and Lyon, 2010), while others use empirical relations and modeling to estimate conductance from other parameters (e.g. Robinson and Vondrak, 1984; Robinson et al., 1987; Zhang et al., 2015; McGranaghan et al., 2016). First-principles models of the ionosphere/upper atmosphere (e.g. Quegan et al., 1982; Roble et al., 1988; Huba et al., 2000; 2008) also contain all the parameters needed to calculate the conductance (altitudinal profiles of electron density and neutral atmospheric density).
A major obstacle to the development of reliable high-latitude electrodynamic solutions has been the challenge of validation. The high-latitude electric potential is not directly observable, and even its derivative, the high-latitude electric field, is observed only through its effect on ions via the E xB drift. As a result of this lack of direct comparison data, many recent developments in E xB drift solutions have been published without validation (e.g. Bristow et al., 2016; Gjerloev et al., 2018). In order to validate convection maps, some authors have used ionospheric F-region electron density structures as a tracer of the ExB drift. Wang et al. (2018) and Ramirez et al. (2019) have used the observed motion of high-latitude electron density structures to assess the accuracy of Weimer’s (2005) empirical model and Ruohoniemi and Baker’s (1998) SuperDARN-driven high-latitude potentials.

2. Method

In this investigation, we aim to specify the high-latitude electrodynamics in order to predict the generation and development of ionospheric electron density structures (e.g. the tongue of ionization, polar cap patches). We then test our predictions against independent GPS-based TEC data showing those structures, and against satellite ion drift data. The approach is as follows:
a) determine the Pedersen (\(\Sigma_{P}\)) and Hall (\(\Sigma_{H}\)) conductances from SAMI3,
b) calculate the high-latitude potential (Ψ) using \(\Sigma_{P}\),\(\Sigma_{H}\) and AMPERE-observed field-aligned currents (J ),
c) use Ψ to calculate E xB plasma drifts in SAMI3
d) compare simulation results to data
The potential is calculated at a 10-minute cadence, matching the latitude re-sampling cadence of the Iridium satellites used for AMPERE. The outputs are the full set of electrodynamic parameters (\(\Sigma_{P}\), \(\Sigma_{H}\), \(\Sigma_{||}\), Ψ, J ) as well as the ionospheric electron density distribution, which can be vertically integrated to produce Total Electron Content (TEC). The electric potentials are compared against Defense Meteorological Satellite Program (DMSP) horizontal ion drift data (e.g. Hairston and Heelis, 1993) during the period of interest. The TEC output is compared against images from the University of Bath’s Multi-Instrument Data Analysis Software (Mitchell and Spencer, 2003; Spencer and Mitchell, 2007), which uses ground-based GPS data. The MIDAS algorithm’s ability to image the tongue of ionization and patches has previously been validated by Wang et al., 2018, Spencer and Mitchell, 2007, and others.
For this demonstration, we have applied the technique to a magnetically active period in the most recent solar maximum: 21-25 May 2014. This period contains relatively strong field-aligned currents, both because of the elevated magnetic activity and because the northern polar cap is in sunlight. Figure 1 shows the IMF Bz component and the Kyoto Disturbance Storm Time (Dst) and Auroral Electrojet (AE) indices for this period. On 23 May 2014, Bz reached a minimum of -7.5 nT at 19 UT, Dst reached a minimum of -36 nT at 21 UT, while the maximum AE of 909 nT occurred earlier, at 17 UT. The case could be summarized as a weak geomagnetic storm with substantial substorm activity.