OR/14/004 Worst case scenario research at NASA and Catholic University of America

Thomson, A W P (Editor), Beggan, C, Kelly, G, Baillie, O, Viljanen, A, and Ngwira, C. 2014. Project EURISGIC: worst case scenarios (Technical note D5.1). British Geological Survey Open Report, OR/14/004.

The worst-case scenarios work at NASA and The Catholic University of America (CUA) has been approaching the problem from three different viewpoints: i) event-based, ii) statistics and iii) theoretical modeling. The event-based and statistics approaches are described in detail in Pulkkinen et al. (2012); Ngwira et al. (2013a) and for brevity the results are not repeated here. Pulkkinen et al. (2012) developed full 1-in-100 year storm and geoelectric field scenarios as a function of local ground conductivity structures and geomagnetic latitudes. Time series representation of the scenarios were provided and applied to power systems in Virginia, US and in UK. The statistics-based scenarios have also been extended to cover all US physiographic regions derived by US Geological Survey (Pulkkinen and Ngwira, 2014).

Later work at NASA and CUA has been focusing on fully first-principles-based investigations of extreme geoelectric field and GIC phenomena. The driving motivation for such studies is that the maturity of modern first-principles-based space physics models allows us to start asking questions about hypothetical extreme storm situations not present in the observational geomagnetic data sets. More specifically, we want to use state-of-the-art space physics models to acquire information about theoretical extremes: how bad can space weather conditions get from the first-principles standpoint? To push our understanding on this topic we have utilized modern space physics model(s) hosted at Community Coordinated Modelling Center (CCMC) for solving the first-principles equations of the solar wind-magnetosphere-ionosphere system under extreme solar wind driving conditions. These solutions allow us to explore also extreme variations of the ground magnetic field and corresponding geoelectric fields. In Discussion and conclusions we describe the NASA and CUA work on the topic.

Using global 3D MHD simulations

The coupling of fast moving CME's to planetary magnetospheres has been a subject of great scientific interest. The magnetosphere is a highly complex nonlinear system whose large-scale state is controlled primarily by the orientation of the interplanetary magnetic field (IMF) and solar wind plasma properties. The transfer of mass, momentum and energy from the solar wind into the magnetosphere-ionosphere system produces various transition layers, the extended geomagnetic tail, and different dynamic current systems and auroral processes.

Three-dimensional (3D) global magnetohydrodynamics (MHD) models play a critical role in simulating the large-scale dynamics of magnetospheric plasmas. These first principles physics-based models represent a very important component of attempts to understand the response of the magnetosphere-ionosphere system to varying solar wind conditions (see e.g., Gombosi et al., 2000; Palmroth et al., 2004). Upstream solar wind parameters are used as driving conditions for many simulation models of the magnetosphere-ionosphere system, and the results (or performance of these models) are validated by comparing with ground-based or satellite observations. Understanding of the magnetosphere and ionosphere dynamics during extreme solar wind driving is still a major challenge, mainly because of lack of modern scientific data from such periods, as explained by Ridley et al., (2006).

Modeling ‘Carrington-type’ storm events

Ngwira et al., (2013b; 2014a) have introduced a 3D MHD modeling approach with specially refined components for modeling extreme space weather events. The core MHD model is based on University of Michigan Space weather Modeling Framework (SWMF) that uses the BATSRUS code (Powell et al., 1999) to predicts in a self-consistent manner the dynamic response of the large-scale magnetosphere to changing solar wind conditions. In this study, the low-latitude Colaba estimated minimum negative geomagnetic intensity was used as a benchmark for simulating space weather events that are constructed using extreme upstream solar wind input conditions. Historically, MHD models have typically been utilized for studying non-extreme events. So, the primary purpose of this work is to examine the simulated ground geomagnetic and geoelectric field response during extreme solar wind driving conditions.

We use the SWMF generated ground magnetic perturbations as our primary data for modeling the global ground induced geoelectric field distribution using the plane wave method. On the ground, the simulated CME shows strong geomagnetic and geoelectric field perturbation. Figure 4.1 displays example time series of ground induced geoelectric field components and magnetic perturbations at two active INTERMAGNET ground sites.

Figure 4.1    Example simulated time series of ground induced geoelectric field components Ex (top) and Ey (middle). Units are V/km. The bottom panels show the simulated time series of the horizontal ground magnetic field perturbations. The two selected high-latitude magnetometer locations are Ottawa (left) and Eskdalemuir (right).

Figure 4.2 shows the distribution of global maximum ground induced geoelectric fields determined for all INTERMAGNET sites. Simulation results for the Carrington-type event are displayed in Figure 5.2(a). It is clearly evident from the figure that the model is able to reproduce the global geoelectric field distribution.

To further test the MHD model performance, a simulation of a portion of the Halloween superstorm on October 29, 2003 was carried out using the same model settings. Then, we compared the modeled maximum geoelectric field to the observed for the same portion of the storm event. Results are provided in Figure 4.2(b) showing a comparison for the Halloween storm of the geoelectric field determined from model derived magnetic perturbations (blue) and the geoelectric field determined from observations (red) with a very good agreement. This gives us confidence in the MHD code performance and the results since the model is generally able to capture important features, such as the geoelectric field transition region between the middle and high latitudes.

Figure 4.2    Global distribution of the peak geoelectric fields determined for: (a) the Carrington-type event simulation, and (b) for the Halloween storm event, i.e., simulation in blue and observations in red. Each ‘*’ represents a specific ground magnetometer site, and the time of the peak electric field varies from site-to-site. The vertical red dashed lines show the locations of the transition regions between middle and high latitudes.

Additionally, Figure 4.2(a) also clearly shows that the location of the latitude threshold boundary, which is the transition region between the middle and high latitude likely caused by the auroral electrojet current, shifted to 40° geomagnetic latitude. This is much lower than previously determined (50–55° geomagnetic) for observed severe geomagnetic storm events (see reports by Pulkkinen et al. 2012; Ngwira et al., 2013a). Furthermore, the strongly shifted latitude threshold boundary implies that the region of large ground induced electric fields is displaced further equatorward due to a shift of the auroral current system, thereby may affect power grids in regions normally far away from the auroral zone, such as Southern states of continental USA or Central and Southern Europe.

The maximum high-latitude geoelectric field of 26 ± 4 V/km presented in Figure 4.2(a) for the Carrington-type event is in close agreement with predicted theoretical maximum for the 100-year scenario (20 V/km) reported by Pulkkinen et al., (2012).

Modeling 23 July 2012 extreme space weather event

On July 23, 2012, a CME was hurled away from the Sun's active region AR1520 with a Rare speed (‘R’-type) of approximately 2500 ± 500 km/s (Baker et al., 2013; Ngwira et al., 2013b). This particular CME was not Earth-directed, but was the fastest ever observed in-situ by NASA's STEREO-A spacecraft and had particularly large IMF components (Russell et al., 2013). Events such as the July 23rd, 2012 CME event offer an unprecedented opportunity to explore the effects of extreme space weather. In our study Ngwira et al., (2013b), we considered NASA's STEREO-A spacecraft in-situ observations to represent the upstream L1 solar wind boundary conditions that are used as driving conditions for the global SWMF MHD model. Figure 4.3 shows the interplanetary conditions associated with this CME. Our primary goal was to examine the geomagnetically induced electric field response that this R-type space weather event could have generated had it hit the Earth.

Figure 4.3    Solar wind in-situ observations from the STEREO-A spacecraft. From top to bottom are: the IMF By, IMF Bz, plasma bulk speed (Vsw), the velocity Vy (solid) and Vz (dashed) components, the solar wind density (Np) and the temperature (Temp). Note that the density (red trace) is derived using the WSA-ENLIL 3D MHD heliosphere model due to challenges in extracting the PLASTIC density data.

Figure 4.4 depicts the maximum induced geoelectric fields at all the individual ground sites used in this study. The figure shows two simulation results, i.e., geoelectric field simulated using STEREO in-situ real-time density (blue) and the other using WSA-ENLIL model density (red). Since our interest is in a worst-case scenario, therefore we only discuss results simulated using the WSA-ENLIL density. Here, the latitude threshold boundary (red dashed line) is located around 50 degrees MLAT and is consistent with observations for severe geomagnetic storms (Thomson et al., 2011; Pulkkinen et al., 2012; Ngwira et al., 2013a). The location of these transition regions between middle and high latitudes is related to the dynamics (strengthening and widening) of the auroral current system Ngwira et al., (2013a).

Figure 4.4    The maximum induced ground electric field (blue) simulated using STEREO in-situ density and (red) simulated using WSA-ENLIL model density. Note that the time of the maximum field varies from site-to-site. The red and blue dashed lines show the locations of the mid- to high latitude geoelectric field transition regions, as discussed earlier.

Ngwira et al., (2013b) showed that the largest simulated induced geoelectric fields were observed on the nightside in the European high-latitude sector. One of the interesting features of the result in Figure 4.4 is the value of the maximum ground induced geoelectric field with a peak value of 14.38 V/km. This value is 3 V/km higher than the value determined for previously observed events during the period 1989 to 2005 (11.4 V/km). It was determined by Ngwira et al. (2013b) that the largest geoelectric field peaks were driven by substorm-type dynamics in the simulation.