2 Database and methodology
One of the requirement to perform the statistical analysis is long-term availability of all-sky imaging observations for the arc detection. An OMTI (Optical Mesosphere Thermosphere Imager) (Shiokawa et al., 1999, 2009) all-sky imager (ASI) (camera no. 7) has been operating at Athabasca (54.6°N, 246.3°E, magnetic latitude: 61.5°N, L = 4.4), Canada, since September 3, 2005. Magnetic midnight for this ASI occurs at ~8.1 UT. In the present study, we have used 15-years of continuous ASI observations from 2006-2020. The camera has a 180° field-of-view fish-eye lens, seven band-pass optical filters, and a thermoelectrically cooled CCD with 512 × 512 pixels. In order to increase the signal to noise ratio, CCD images are processed with 2 × 2 binning, so that the images had a resolution of 256 × 256 pixels. The seven band-pass filters of this camera allow measurement of airglow/auroral emissions at specific wavelengths: OI at 557.7 nm, OI at 630.0 nm, Hβ at 486.1 nm, Na at 589.3 nm, OH bands at 720–910 nm, OI at 844.6 nm, and nominal background at 572.5 nm. In this study, we have used images at all the wavelengths except for OH bands. The details about the exposure times of the images at different wavelengths are given in Yadav et al. (2021a).
The raw images are projected into the geographical latitude/longitude coordinates by assuming that the emission intensity has a peak at 250 km altitude for 630.0 nm and 120 km for the other wavelengths. We have constructed North-South (NS) keograms by stripping the slices of airglow images at the longitude of Athabasca (246.3°E). The keograms facilitate the study of temporal variations in the auroral emission intensities and their latitudinal motion.
As mentioned earlier, proton arcs are not considered in this study. Subauroral proton arcs, which exhibit emission in 557.7 nm, 630.0 nm, and 486.1 nm (Hβ), have been shown to occur simultaneously with Pc 1 geomagnetic pulsations in the frequency range of the EMIC wave (e.g., Sakaguchi et al., 2008). An induction magnetometer has been operating at Athabasca since September 2005. The induction magnetometer measures variation of a three-component geomagnetic field with a sensitivity of 0.45 (V/nT) at 6 Hz with a turnover frequency of ∼6 Hz. The dynamic magnetic field spectra used to decipher the presence of Pc 1 geomagnetic pulsations is available at the ISEE magnetometer website (https://stdb2.isee.nagoya-u.ac.jp/magne/induction/index.html). Detached red+green arcs with simultaneous occurrence of Pc 1 geomagnetic pulsation were eliminated from the present study.
To study the solar flux and geomagnetic activity dependence of detached arcs, we obtained the yearly solar flux and geomagnetic indices from the NASA’s Space Physics Data Facility (SPDF) OMNI database. The 1-min resolution geomagnetic indices were also obtained from the OMNI database, used to examine the geomagnetic conditions during the detachment of arcs. The magnetometer data at Athabasca (ATHA), Fort Smith (FSMI; Lat 60.0°N, Long 248.2°N; magnetic latitude: 67.28°N), and Boulder (BOU; Lat 40.1°N, Long 254.7; magnetic latitude: 48.7°N) were used to access the substorm activity in the longitude zone of Athabasca ASI. Fort Smith is located in the auroral oval, whereas Boulder is a middle latitude station.
The magnetic field (MAG) and particle data from MAGnetospheric Electron Detector (MAGED) and MAGnetospheric Proton Detector (MAGPD) onboard Geostationary Operational Environmental Satellite (GOES) were used to identify the substorm injection signatures at the geosynchronous orbit. We used data from GOES-15 [135°W], GOES-14 [105°W], GOES-12 [75°W], and GOES-11 [135°W]. The GOES satellites were located between ~19 to 24 MLT for all STEVE events. The longitudinal difference of GOES-15 and GOES-11 with Athabasca is ~63 degree, whereas GOES-14 and GOES-12 is ~38 degree and 8 degree, respectively. In this paper, we have shown only particle data from GOES-15 to demonstrate the dispersionless injection prior to three STEVE events occurring during 2018-2019.
We have used the Space Physics Environment Data Analysis Software (SPEDAS) tool (Angelopoulos et al., 2019) to download and analyze the 1-min geomagnetic activity indices, ground-based magnetic field, and GOES particle flux data. We used median subtracted X-component magnetograms in this study.