Energy transfer between the solar wind and Earth’s magnetosphere at the dayside magnetopause is the first step in the Dungey cycle and subsequent circulation of energy through the coupled solar wind-magnetosphere system. As part of Science Objective A, STORM answers two fundamental science questions: 1) How does global magnetopause reconnection control the flow of solar wind energy into the magnetosphere? 2) What are the spatial and temporal properties of this interaction as a function of solar wind conditions?

Understanding the dynamics of reconnection is essential to quantifying the flow of energy in the solar wind-magnetosphere interaction. Global simulations, in-situ, and remote measurements provide evidence for different magnetopause reconnection modes. Reconnection can be steady (Sonnerup et al., 1981; Trattner et al., 2021) or bursty (Russell & Elphic, 1978; Pfau-Kempf et al., 2020; Zou et al., 2022), and may (Lockwood et al., 1989) or may not (Le et al., 1993) be triggered by southward IMF turnings. Reconnection can be localized to less than 1 h of local time (Marchaudon et al., 2004; Oksavik et al., 2004; Oksavik et al., 2005; Fear et al., 2007) or extend across 5–8 h of local time on the dayside magnetopause (Lockwood et al., 1990; Milan et al., 2016), it may occur at off-equatorial pre- and post-noon (Crooker, 1979; Sandholt and Farrugia, 2003) or subsolar locations (Sonnerup, 1974; Mozer et al., 2002). Reconnection at the dayside magnetopause can take place gradually over 1–2 h causing an earthward erosion of the magnetopause of 1–2 R_E (Aubry et al., 1970) and an equatorward motion of the cusp aurora of more than 8° (∼800 km in the auroral oval) (Burch, 1972). This corresponds to a protracted interval of localized dayside proton precipitation resulting in auroral intensities of 1 kR, with a longitudinal width of ∼10° (∼290 km) that can move 3 h in local time over a few hours (Frey et al., 2003; Phan et al., 2003). Alternatively, reconnection may take place abruptly in ∼1–3 min bursts each ∼7–10 min that generate flux transfer events with dimensions ∼1 R_E normal to the magnetopause (Russell and Elphic, 1978). This leads to recurrent auroral brightenings followed by the equatorward motion of the dayside auroral oval in 100 km steps (Sandholt et al., 1998) and corresponding earthward and equatorward jumps of the magnetopause and cusp.

STORM distinguishes between steady and unsteady modes of magnetopause reconnection by tracking the motion of the magnetopause (XRI), dayside auroral oval (FUV), and intensity of proton auroral precipitation (FUV). To capture the location of the subsolar magnetopause over a broad range of solar wind conditions equivalent to a 5.2 R_E range of magnetopause locations from its canted location on STORM some 30 R_E from Earth, XRI requires a 10° × 10° field of view (FOV). To distinguish between steady and unsteady reconnection and erosion of the magnetopause, XRI must track spatiotemporal variations in the density gradients that indicate lines of sight tangent to the magnetopause with the 0.25 R_E and 3 min resolution needed to track 1–2 R_E erosion of the magnetopause (Aubry et al., 1970) over 60–120 min. From some vantage points, STORM would observe lines of sight that lie tangent to locations near the subsolar magnetopause; however, lines of sight tangent to the magnetopause away from the subsolar point can just as readily be used to identify magnetopause motion and determine the global spatiotemporal dependence of magnetic reconnection on the orientation of the interplanetary magnetic field (Collier and Connor, 2018). To quantify the spatial and temporal dynamics of reconnection, FUV must track the latitude of the dayside auroral oval with 100 km resolution and 3 min cadence to distinguish between magnetopause reconnection modes that drive 8° motion (Burch, 1972) over 60–120 min. FUV must also observe the intensity of the proton aurora with a spatial resolution and cadence of 100 km and 3 min to distinguish between proton precipitation driven by steady and unsteady dayside reconnection (Frey et al., 2003). The cadences and spatial resolutions for the XRI and FUV imagers readily distinguish between modes of dayside reconnection (Figure 2).

STORM employs FUV observations to infer the location and extent of magnetic reconnection on the magnetopause. FUV must track the location and azimuthal extent of latitudinal motion in the electron auroral oval with a 100 km resolution to determine the azimuthal extent and flux opened by bursty reconnection and whether it is confined to the vicinity of local noon or occurs preferentially at locations away from local noon to distinguish between component and antiparallel reconnection models (Milan et al., 2000). Similarly, FUV must identify the locations where proton aurora occurs within the dayside auroral oval with a spatial resolution and cadence of at least 100 km (3.5° longitude at 75° cusp latitudes) to determine the extent and location of the reconnection line (component vs antiparallel) as projected into the ionosphere from observations of proton precipitation (Frey et al., 2003; Lockwood et al., 2003).

STORM’s in-situ measurements of the solar wind determine event occurrence patterns. IES must observe the local solar wind plasma input with a cadence of 2 min, commensurate with that of the magnetospheric imagers, to accurately identify the arrival times of solar wind features at the subsolar magnetopause, eliminate solar wind dynamic pressure variations as the cause of magnetopause motion and changes in the brightness of the dayside aurora, and identify signatures in the solar wind plasma that may trigger reconnection (Frey et al., 2019). MAG must observe the IMF with a cadence of 1.5 s or faster to employ cross-product and minimum variance algorithms that determine the locations where solar wind features first impact the magnetosphere and trigger reconnection with less uncertainty than observations from L1 (Crooker et al., 1982; Cameron and Jackel, 2016). Finally, the EUV must observe the ∼1–1.8 R_E sunward motion of the day and nightside plasmasphere over >60 min (Goldstein et al., 2003; Goldstein and Sandel, 2005), readily observable with the 0.18 R_E spatial resolution and 2 min cadence of EUV. This motion quantifies the changes in the strength of convective electric fields in the inner magnetosphere associated with dayside reconnection providing an additional diagnosis of the significance of different reconnection modes in the solar wind-magnetosphere interaction.

STORM’s mission design provides multiple ways to quantify the significance of individual dayside reconnection events: 1) via direct observations of the intensity and extent of the dayside proton aurora (Frey et al., 2003), 2) via the amount of dayside flux opened by equatorward motion of the electron auroral oval, 3) via the amount of flux eroded from the dayside magnetosphere as inferred from observations of eroding magnetopause locations and 4) via inferences of the dayside reconnection electric field imposed on the inner magnetosphere from observations of low energy plasmaspheric motion. Statistical studies determine the geomagnetic effectiveness of each dayside reconnection mode as a function of solar wind drivers and geomagnetic conditions. Requirements for XRI and FUV are the same as those listed above. FUV must be capable of continuously observing the location, extent, and intensity of the dayside auroral oval and proton aurora over days, corresponding to the time scale of geomagnetic storms (Lockwood et al., 2003; Frey, 2007; Milan et al., 2007).

STORM’s comprehensive global space-based observations 1) identify distinct modes of magnetopause reconnection, 2) determine their occurrence patterns, and 3) quantify their significance in the circulation of energy throughout the magnetosphere as a function of solar wind and geomagnetic conditions providing science closure to objective A. Figure 3 presents an idealized example of how STORM discriminates between modes of dayside reconnection thereby quantifying the physical processes governing the entrance of energy into the solar wind magnetosphere system. It is noteworthy to mention that though Figure 3 is an idealized example, the STORM team has developed algorithms to identify the magnetopause boundary which take into account both XRIs instrument response as well as background noise (Sibeck et al., 2018). Similarly, the team has been working with existing algorithms developed during the IMAGE era (Boakes et al., 2008; Milan et al., 2009a) to improve the identification of auroral features and boundaries with the FUV instrument. Other researchers have also been developing algorithms to identify the magnetopause from simulated soft X-ray instruments (Collier & Connor, 2018; Jorgensen et al., 2019a; Jorgensen et al., 2019b; Sun et al., 2020; Connor et al., 2021). Finally, the STORM team combines observations from FUV and XRI to separate magnetopause motion driven by reconnection and that driven by changes in solar wind dynamic pressure.

Following dayside magnetopause reconnection, open magnetic flux is transported over the polar cap and stored within the magnetotail. This flux does not build up indefinitely. Eventually reconnection is triggered within the nightside plasma sheet. Reconnection closes the open magnetic flux and accelerates closed magnetic flux and plasma into the inner magnetosphere. Science objective B addresses this energy circulation and transfer through the magnetotail by answering two key questions: 1) How does magnetotail reconnection regulate the circulation of energy from the dayside, through the magnetotail, and into the inner magnetosphere? 2) What controls the occurrence and significance of differing reconnection modes?

In-situ and remote measurements, together with supporting results from global simulations, provide evidence for a number of distinct magnetotail reconnection modes (DeJong et al., 2007; Kissinger et al., 2012a; Walach et al., 2017) that cannot be deconvolved without observing the complete system. This includes the solar wind local to the magnetosphere, the microscale auroral dynamics associated with nightside reconnection, and the return of flux to the dayside magnetosphere and subsequent outward motion of the magnetopause boundary. STORM’s unique mission design provides all these essential observations.

Isolated substorms (Akasofu, 1964; Akasofu, 2017) require a period of energy storage following southward IMF turnings. These isolated tail modes exhibit a ∼1 h growth phase corresponding to plasma sheet thinning and expansion of the auroral oval. This leads to a transient, localized, onset marked by a discrete brightening of the nightside auroral oval in the vicinity of local midnight followed by a ∼30 min expansion phase. This expansion phase is characterized by dynamic aurora moving poleward and spanning multiple hours of local time across the night sky. The expansion phase is followed by a ∼1 h recovery phase during which aurora dim and the oval begins to expand again. Sawtooth events are sequences of three–eight substorms that occur every ∼2–4 h during a nearly continual growth phase (Borovsky et al., 1993; DeJong et al., 2007). Some reports indicate substorm sequences (duration) and properties (auroral brightness and oval motion) during several-daylong geomagnetic storms differ from those of non-storm time substorms (Baumjohann et al., 1996; Korth et al., 2002; Hoffman et al., 2010) while other reports indicate they do not (Hsu and McPherron, 2000).

Not all nightside reconnection modes follow the isolated substorm sequence. Poleward boundary intensifications (PBI), pseudobreakups, or steady magnetospheric convection (SMC) can occur with no growth, expansion, or recovery phases. PBIs occur as brightenings with ∼1° or more longitudinal extent along the most poleward arc (Zesta et al., 2002). They may be the optical signatures of nightside reconnection. The north/south oriented arcs, known as streamers, that ensue are thought to be the optical signatures of reconnection-induced flow bursts. Pseudobreakups occurring on 10 min time scales are auroral brightenings resembling those at substorm onset that do not engage the open field line region within the polar cap (Akasofu, 1964). SMCs are intervals of strongly southward IMF orientation greater than ∼6 h long without classical substorms. They exhibit multiple bursts of streamer activity corresponding to bursts of nightside reconnection that, in total, balance steady dayside reconnection (DeJong et al., 2009; Kissinger et al., 2012b; Walach and Milan, 2015).

Table 1 highlights the vast array of observational work, spanning over half a century, dedicated to identifying the drivers of nightside reconnection modes. It also illustrates the often-contradictory findings of these studies regarding the importance of various solar wind and magnetospheric conditions that favor the development of each mode. Although there have been extensive improvements in global simulations, these models frequently produce contradictory results (Gordeev et al., 2017). These contradictions are due to model limitations including, for example, simulations that do not fully couple all relevant plasma regimes, model results that depend on spatiotemporal resolution, and global simulations that do not fully include kinetic effects or realistic ionospheres and thermospheres. These limitations and contradictory results preclude disentangling the drivers shown in Table 1 using global simulations of the solar wind-magnetosphere interaction. STORM provides the observations needed to distinguish between diverse nightside reconnection modes, determine their occurrence patterns as a function of solar wind and geomagnetic conditions, and quantify the significance of each in the global circulation of energy throughout the magnetosphere.

STORM’s extended array of high-resolution ground-based imagers together with the global auroral imager distinguish between proposed interaction modes. FUV must observes the nightside electron aurora with a spatial resolution of at least 100 km (∼1° latitude) and cadence of 3 min to track the ∼5° equatorward motion and rapid expansion of the nightside aurora to distinguish each reconnection mode (Frey et al., 2004a). STORM determines whether substorms are isolated, sawtooth, or occurring during geomagnetic storms by their repetition times and the state of geomagnetic activity as measured by the Disturbance-Storm Time (Dst) index. Due to limited ground-based coverage, FUV’s global auroral imaging is the only robust method available for identifying long duration events, including SMCs, sawtooth substorms, and geomagnetic storms.

STORM identifies SMC events on the basis of 1) MAG observations of strongly southward IMF orientations, 2) FUV observations of a steady proton aurora in the dayside oval indicative of continuous dayside reconnection (a prerequisite for balanced reconnection), 3) FUV observations of an auroral oval with nearly constant dimensions (thereby requiring the FUV to have a FOV that covers the entire auroral oval), and 4) ASI observations of repetitive auroral streamers. This method of identifying SMCs is superior to those relying on ground-based magnetometers (Kissinger et al., 2011), which can only infer intervals of SMC, as these methods provide no measure of the amount of open flux stored in the magnetosphere. The ASI array complements FUV and must provide observations with spatial resolutions of 20 km and cadences of 30 s to identify auroral streamers (Gallardo-Lacourt et al., 2014) with azimuthal scale sizes of ∼0.5°–1.0° that correspond to 25–50 km at geomagnetic latitudes of ∼65°. STORM employs ASI and FUV observations to identify PBIs and pseudobreakups in the nightside auroral oval. Finally, STORM’s circular orbit and 30 R_E radius enable its imagers to observe phenomena continuously throughout the various time scales (minutes to days) that encompass all nightside interaction modes, including those of multi-day geomagnetic storms.

STORM determines the occurrence patterns of magnetotail response modes as a function of solar wind and geomagnetic conditions, including solar wind and magnetospheric triggers (McPherron et al., 2008; DeJong et al., 2009; Walach et al., 2017). The extensive list of proposed drivers for magnetotail reconnection modes (Table 1) demonstrates the need for STORM’s proposed observations to test these hypotheses systematically both event and statistical bases. This includes comprehensive end-to-end substorm models such as that proposed by Lyons et al. (2016). As illustrated in Figure 4, this model emphasizes the role of meso-scale features within the global circulation of energy throughout the magnetosphere in triggering substorms. A burst of dayside reconnection causes the dayside auroral oval to jump equatorward and the magnetopause to jump earthward. This burst of dayside reconnection generates a plasma structure that propagates anti-sunward and triggers a PBI along the nightside auroral oval. In turn, the PBI launches an equatorward-moving auroral streamer that initiates substorm onset. STORM is ideally instrumented to test and verify global model predictions for dayside, nightside, and solar wind triggers, or the lack thereof (Zesta et al., 2002; Zesta et al., 2006; Murphy et al., 2014), for various magnetotail modes. This is because STORM uniquely provides observations of the dayside magnetopause and auroral oval locations identifying whether reconnection is steady or bursty, the microscale streamers that may trigger substorm onset, and the global auroral response to geomagnetic substorms. STORM provides MAG and IES observations with at least 1.5 s and 2 min cadences from vantage points near Earth. Note that the performance of existing instruments far exceeding these cadences (Figure 2). Studies have shown that timing accuracy decreases with distance from the magnetosphere (Collier et al., 1998; Richardson and Paularena, 1998; Weimer et al., 2003). Because solar wind features have a variety of scale sizes (Crooker et al., 1982; Richardson and Paularena, 1998; Matsui et al., 2002; Weimer et al., 2003), the probability of observing a feature which will impact the subsolar magnetopause decreases with increasing distance away from the Sun-Earth line. Vantage points near Earth minimize these concerns allowing STORM to determine whether solar wind triggers exist, reach the magnetopause, or initiate substorms with greater amplitudes than those that are not triggered (Lyons, 1996; Hsu and McPherron, 2009).

STORM quantifies the significance of individual magnetotail reconnection modes by determining the amount of open polar cap magnetic flux closed, the strength of inner magnetosphere electric fields, rate of convection, the amount of magnetic flux returned to the dayside magnetosphere, and (on a statistical basis) the occurrence rates of the individual mechanisms for differing solar wind, geomagnetic conditions, and histories. FUV tracks the global dimensions of the electron auroral oval with a resolution of 1° to quantify its size and calculate cross polar potential drops, reconnection rates, and flux closure by nightside reconnection (Milan et al., 2007). Simultaneous XRI and FUV must identify the sunward motion of the magnetopause with 0.25 R_E resolution and poleward motion of the dayside auroral oval with 1° resolution to determine if individual modes of nightside reconnection return flux to the dayside (Siscoe et al., 2011) and to quantify the significance of this flux return in the global circulation of energy throughout the magnetosphere. EUV must track the boundary of the plasmasphere with a cadence of 2 min and 0.18 R_E resolution to quantify the strength of convection and the driving convective electric fields (Goldstein et al., 2005).

STORM combines unique and comprehensive global space- and ground-based observations to 1) identify distinct modes of magnetotail reconnection, 2) determine their occurrence patterns, and 3) quantify their significance in the circulation of energy throughout the magnetosphere as a function of solar wind and geomagnetic conditions. Figure 5 demonstrates how STORM differentiates between substorms initiated by auroral streamers and those initiated without auroral streamers; case studies and statistics allow STORM to determine the importance of each mode in energy transport through the magnetotail.

The return of closed magnetic flux to and acceleration and transport of plasma into the inner magnetosphere is the next sequence of energy flow in the Dungey cycle. The acceleration and transport of plasma energizes the ring current and dynamic processes within the inner magnetosphere and on the dayside control the subsequent decay of terrestrial ring current. Science Objective C investigates energy sources and sinks for the ring current by answering three underlying questions: How efficiently do the magnetotail response modes described in Science Objective B energize ring current ions? How do transport and loss mechanisms affect subsequent evolution of the energized ring current? What are the occurrence patterns of the source, transport, and loss mechanisms?

Ring current ions with energies of 30–300 keV at distances of 4–5 R_E from Earth originate in the Earth’s plasma sheet. Each magnetotail reconnection mode: isolated substorms, sawtooth substorms, storm-time substorms, and steady magnetospheric convection (SMC)—supplies ions to the ring current in the form of sequential bursty bulk flows (auroral streamers), substorm injections, and enhanced convection and inner magnetospheric electric fields. Large scale injections associated with isolated substorms provide enhancements by a factor of two to four in the intensity of ∼100 keV ions within a pre-midnight wedge with a local time extent that ranges from ∼2 to >7 h that can penetrate to within ∼4 R_E from Earth (Reeves et al., 1990; Reeves et al., 1991; Reeves et al., 1992; Friedel et al., 1996). Particle injections may penetrate deeper for substorms that consume more magnetic flux (Boakes et al., 2009). Sawtooth and storm-time substorms enhance ring current intensities more effectively than isolated substorms, in part because of the continued and prolonged injection of particles (Reeves and Henderson, 2001) with greater intensities (Cai et al., 2006; Clauer et al., 2006; Walach et al., 2017). However, storm-time and sawtooth modes may inject ions onto open drift paths, leading to only a limited enhancement of the partial ring current in the dusk sector. Prolonged SMC intervals lead to the energization of 5–10 keV ions by factors of up to 10 during ∼7 min long traversals from the nightside near X_GSE = −12 R_E to locations as close to Earth as X_GSE = −3 R_E (Artemyev et al., 2015; Ukhorskiy et al., 2018). These injections result in substantial contributions to the ring current (Gkioulidou et al., 2014).

Ring current intensities do not increase indefinitely. During storms ring current intensities initially decay rapidly over ∼8 h during the early recovery phase and then more slowly over the next several days (Hamilton et al., 1988). Proposed mechanisms for ring current loss include charge exchange, wave-particle induced precipitation, and magnetopause shadowing/outflow. Daglis et al. (1999) concluded that charge exchange with the exosphere is the main mechanism for ring current decay. The fast initial and subsequent slow decay in ring current intensities may result from the higher charge exchange rate of oxygen and slower charge exchange rate of protons (Hamilton et al., 1988; Jordanova et al., 1996; Jordanova et al., 1998; Daglis et al., 1999). By contrast, Liemohn et al. (1999) maintain that the storm time ring current is a partial ring current on open drift paths from which loss via drifts to the magnetopause predominates during periods of enhanced convection. In-situ observations suggesting greater ion intensities post-than pre-noon support this hypothesis (Kronberg et al., 2015). As the convection electric field decays during the late recovery phase, the slower charge exchange loss dominates ring current decay on completely closed drift paths (Takahashi et al., 1990; Liemohn et al., 2001). Electromagnetic ion cyclotron (EMIC) wave-particle interactions and precipitation may become important during the main phase of geomagnetic storms (Gonzalez et al., 1989), particularly at low (10 keV) energies where charge exchange cross sections are 70 times greater than those at higher (100 keV) energies (Smith and Bewtra, 1978; Runov et al., 2016). It is estimated that EMIC wave-driven precipitation accounts for ∼1–7% of the ion loss (Kozyra et al., 1997; Jordanova et al., 1998). However, recent work suggests EMIC waves are more widespread than originally thought, especially during the main phase of geomagnetic storms (Usanova et al., 2010; Tetrick et al., 2017), which may also explain the two-stage decay of the ring current during storms. In contrast, Thorne and Horne (Thorne and Horne, 2004) suggest the presence of O⁺ in the ring current during the storm main phase limits EMIC wave growth and subsequently any loss via EMIC wave-particle interactions.

STORM provides the observations needed to comprehensively evaluate the energization and decay of the ring current (Kozyra and Liemohn, 2003). Section 3.2 addressed how STORM distinguishes between each nightside reconnection mode and quantifies the occurrence and significance of these modes to magnetospheric convection and magnetic flux balance. STORM uses three methods to evaluate the contribution of each nightside mode to the energization of the ring current: 1) ENA tracks the intensity of energetic neutral atoms from the ring current as a function of time to quantify how individual injection events and prolonged intervals of nightside reconnection contribute to the energization of the ring current (Jorgensen et al., 1997). 2) ENA determines the depth to which ring current ion injections penetrate, their azimuthal extent, the degree to which ions are energized, their spectral slopes, and H, O composition on a case-by-case and statistical basis for each event category. 3) EUV tracks the motion of the plasmapause to distinguish between and quantify steady and impulsive electric fields applied to the inner magnetosphere and the resulting magnetospheric convection.

By tracking the time-history of ring current intensities observed by ENA, STORM quantifies the loss rates of ring current ions driven by magnetopause shadowing (West et al., 1972), charge exchange, and precipitation. STORM evaluates the significance of these loss mechanisms in isolation during individual events in three ways: 1) IES observing intervals of enhanced solar wind pressure that compress the magnetopause, making magnetopause shadowing (aka outflow) more likely. As illustrated in Figure 6, the loss rate due to magnetopause shadowing is then determined by comparing pre- and post-noon ENA emissions (Brandt, 2002) on open outer magnetospheric drift paths that originate from and lead to the magnetopause where the location is determined to within 0.25 R_E by XRI observations. 2) ENA directly measuring loss via charge exchange as a function of species (H, O), energy, location, and time (Keika et al., 2006). 3) FUV observing precipitation in the proton aurora (Frey et al., 2004b; Fuselier, 2004; Frey, 2007; Spasojevic and Fuselier, 2009; Spasojević et al., 2013) with at least a100 km spatial resolution to identify the extent and magnitude of scattered ring current ions. ENA tracks the decay in ring current intensities throughout such precipitation events.

Beyond these topics, STORM can test how oxygen may regulate the storage and flow of energy from the magnetotail. Enhanced ionospheric outflow and low energy oxygen in the plasma sheet may result in stronger geomagnetic storms (Glocer et al., 2009). However, enhanced oxygen densities may move the reconnection line closer to Earth, resulting in less heating, less plasma access to the inner magnetosphere, and a weaker ring current (Ilie et al., 2015). ENA provides observations of total ring current and energized oxygen intensities to discriminate between these hypotheses. These energized plasma populations are used to deduce the presence of lower energy outflow and plasma sheet oxygen and thereby determine the overall role of magnetospheric oxygen in controlling the flow of energy in the coupled solar wind-magnetosphere system.

STORM provides continuous observations of the ring current to quantify the sources and sinks of ring current energy. STORM builds on science objective B to distinguish between and quantify how energy is transferred from the nightside magnetotail to the inner magnetosphere in the form of ring current enhancements. STORM quantifies individual ring current ion loss processes, determines their occurrence patterns, and calculates their relative contribution to ring current dynamics. For example, Figure 6 demonstrates how STORM differentiates between magnetopause shadowing and charge exchange losses in the ring current. Finally, although all the instruments on STORM observe many geomagnetic storms from start to finish, these instruments observe even more individual initial, main, and recovery phases, enabling the mission to conduct statistical studies of each phase.

The inner magnetosphere is not the final stop in the circulation of energy in the coupled solar wind-magnetosphere system. Rather the inner magnetosphere hosts dynamic processes that have important feedback effects on the solar wind-magnetosphere interaction. To quantify the circulation of energy in the solar wind-magnetosphere system it is imperative to separate the effects of the solar wind and storage of energy in the tail from the effects of inner magnetospheric plasmas. Science Objective D investigates energy feedback from the inner magnetosphere by answering the following question: How do the ring current and plasmaspheric plumes affect the outer magnetosphere including day and nightside reconnection and the location of the dayside magnetopause?

Plasmaspheric plumes frequently become entrained in magnetopause reconnection (McFadden et al., 2008), particularly during geomagnetic storms when plumes persist for days (Borovsky and Denton, 2008). Since reconnection rates are inversely proportional to plasma densities (Cassak & Shay, 2007), the arrival of a high density plasmaspheric plume at the magnetopause quenches reconnection. Simulations show that plumes locally reduce the reconnection rate by a factor of 2 (Borovsky et al., 2008), however this may be compensated for by enhancing reconnection rates at magnetopause locations just outside the regions where the plumes encounter the magnetopause such that the net dayside reconnection rate remains unchanged (Ouellette et al., 2016).

A strong storm-time ring current that endures from hours to days can affect the dynamics of the dayside and nightside magnetosphere. The ring current enhances northward magnetic field strengths in the dayside and nightside outer magnetosphere, including locations near the magnetopause. As shown in Figure 7, corresponding enhancements in the magnetospheric pressure cause the magnetopause to move outward (Schield, 1969) and compete with the inward erosion of the magnetopause resulting from Region one field-aligned currents associated with reconnection on the dayside magnetopause (Sibeck et al., 1991). Tsyganenko and Sibeck (1994) estimated the effect of the ring current, if any, on the magnetopause location to be small; however, global MHD models with no ring current systematically predict the magnetopause closer to Earth than observed, suggesting the effect of the ring current on magnetopause location is not negligible (García & Hughes, 2007). MHD simulations and simple calculations indicate that even the modest enhancements in ring current strength corresponding to small geomagnetic storms can move the subsolar magnetopause outward by ∼0.3–0.8 R_E (Samsonov et al., 2016). Observationally, the results are mixed; Petrinec and Russell (1993) reported that the ring current has only a small effect on the magnetopause location. However, a number of researchers report that enhanced partial ring currents in the dusk magnetosphere cause a clear outward motion of the duskside magnetopause (Wrenn et al., 1981; McComas et al., 1993; Dmitriev et al., 2004; Dmitriev et al., 2005; Dmitriev and Suvorova, 2012), although some researchers dispute this effect (McComas et al., 1994).

In the magnetotail, enhanced northward magnetic fields may prevent the magnetospheric magnetic field from assuming the stretched magnetotail configuration needed to initiate substorms. Consequently, periods of enhanced ring current may require greater amounts of open magnetospheric flux to initiate substorms and also lead to more intense bursts of nightside reconnection (Milan et al., 2009b).

To investigate the effect of plumes on dayside reconnection STORM utilizes EUV to identify plumes and determine their arrival time and location at the magnetopause from their sunward velocity. At geosynchronous orbit, plumes typically have widths of 1–6 R_E and move outward/sunward at velocities of ∼10 km/s, with the velocity diminishing with age and increasing with the strength of magnetospheric convection (Borovsky and Denton, 2008), placing a requirement on EUV cadence and spatial resolution (10 km/s is 0.5 R_E in 10 min). FUV then determines whether plumes halt equatorward erosion of the dayside auroral oval from electron observations and diminish the proton flux precipitating into the high latitude dayside ionosphere from proton aurora observations. XRI further supplements these observations by tracking the magnetopause location, where inward motion due to reconnection should be quenched by the arrival of a plume.

When STORM is located at high latitudes, ENA observes energetic neutral atoms coming from the ring current at all local times and radial distances, enabling direct determination of ring current and partial ring current strengths (Fok et al., 2003) not possible by other observations such as ground-based magnetometers which observe the integrated effects of several current systems (e.g., the magnetopause, Region 1 and 2, ring current, cross-tail, and electrojet currents) (O’Brien and McPherron, 2000). From similar vantage points, XRI observes the location of the global dayside magnetopause, including outward displacements due to the ring current and dawn/dusk asymmetries resulting from the partial ring current. Working together, ENA and XRI determine whether enhanced ring current strengths push the entire magnetopause outward, and/or whether enhanced partial ring current strengths push the post-noon magnetopause outward. To quantify the effect of the ring current on nightside reconnection, ENA observes variations in ring current strength from equatorial and polar vantage points. When at polar locations, STORM employs FUV to determine the size and shape of the polar cap and therefore the amount of open magnetic flux needed to trigger substorm onset. When STORM is at equatorial latitudes, the ASI array identifies the time and location of substorm onset in a limited local time sector. Working together, these instruments determine whether greater amounts of open flux are needed to trigger substorms during periods when ring current strengths are enhanced (Milan, 2009).

STORM uses simultaneous and global observations of the ring current, plasmasphere, magnetopause, and auroral oval to determine the effects of the ring current and cold plasmaspheric plasma on the dayside and nightside magnetosphere. Images of the ring current, plasmasphere and plumes, magnetopause location, and auroral oval allow STORM to quantify the effect of the ring current and cold plasma on the location of the dayside magnetopause and rate of dayside reconnection, as well as the latitude of substorm onset in the nightside auroral oval. Science Objective D completes STORM’s comprehensive investigation and quantitative determination of the circulation of energy throughout the magnetosphere in the coupled solar-wind magnetosphere system. Figure 8 demonstrates how STORM quantitatively tests the relationship between ring current strength and the amount of open flux required to trigger a substorm.