Interplanetary origin of geomagnetic activity in the declining phase of the solar cycle

Interplanetary magnetic field (IMF) and plasma data are compared with ground‐based geomagnetic Dst and AE indices to determine the causes of magnetic storms, substorms, and quiet during the descending phase of the solar cycle. In this paper we focus primarily on 1974 when the AE index is anomalously high . This year is characterized by the presence of two long‐lasting corotating streams associated with coronal holes. The corotating streams interact with the upstream low‐velocity (300–350 km s −1 ), high‐density heliospheric current sheet (HCS) plasma sheet, which leads to field compression and ∼ 15‐ to 25‐nT hourly average values. Although the B z component in this corotating interaction region (CIR) is often < −10 nT, typically the field directionality is highly variable, and large southward components have durations less than 3 hours. Thus the corotating stream/HCS plasma sheet interaction region can cause recurring moderate (−100 nT ≤ Dst ≤ −50 nT) to weak (−50 nT ≤ Dst ≤ −25 nT) storms, and sometimes no significant ring current activity at all ( Dst > −25 nT). Storms of major ( Dst ≤ −100 nT) intensities were not associated with CIRs. Solar wind energy is transferred to the magnetosphere via magnetic reconnection during the weak and moderate storms. Because the B z component in the interaction region is typically highly fluctuating, the corresponding storm main phase profile is highly irregular. Reverse shocks are sometimes present at the sunward edge of the CIR. Because these events cause sharp decreases in field magnitude, they can be responsible for storm recovery phase onsets. The initial phases of these corotating stream‐related storms are caused by the increased ram pressure associated with the HCS plasma sheet and the further density enhancement from the stream‐stream compression. Although the solar wind speed is generally low in this region of space, the densities can be well over an order of magnitude higher than the average value, leading to significant positive Dst values. Since there are typically no forward shocks at 1 AU associated with the stream‐stream interactions, the initial phases have gradual onsets. The most dramatic geomagnetic response to the corotating streams are chains of consecutive substorms caused by the southward components of large‐amplitude Alfvén waves within the body of the corotating streams. This auroral activity has been previously named high‐intensity long‐duration continuous AE activity (HILDCAAs). The substorm activity is generally most intense near the peak speed of the stream where the Alfvén wave amplitudes are greatest, and it decreases with decreasing wave amplitudes and stream speed. Each of the 27‐day recurring HILDCAA events can last 10 days or more, and the presence of two events per solar rotation is the cause of the exceptionally high AE average for 1974 (higher than 1979). HILDCAAs often occur during the recovery phase of magnetic storms, and the fresh (and sporadic) injection of substorm energy leads to unusually long storm recovery phases as noted in Dst . In the far trailing edge of the corotating stream, the IMF amplitudes become low, <3 nT, and there is an absence of large‐amplitude fluctuations (Alfvén waves). This is related to and causes geomagnetic quiet. There were three major ( Dst ≤ −100 nT) storms that occurred in 1974. Each was caused by a nonrecurring moderate speed stream led by a fast forward shock. The mechanisms for generating the intense interplanetary B s which were responsible for the subsequent intense magnetic storms was shock compression of preexisting southwardly directed B z ( B s ) for the two largest events and a magnetic cloud for the third (weakest) event. Each of the three streams occurred near a HCS crossing with no obvious solar optical or X ray signatures. It is speculated that these events may be associated with flux openings associated with coronal hole expansions. In conclusion, we present a model of geomagnetic activity during the descending phase of the solar cycle. It incorporates storm initial phases, main phases, HILDCAAs, and geomagnetic quiet. It uses some of the recent Ulysses results. We feel that this model is sufficiently developed that it may be used for predictions, and we encourage testing during the current phase of the solar cycle.