The solar corona is structured by the Sun’s magnetic field, which evolves over the 11-year solar cycle. During solar minimum, the global magnetic field is dominated by a low-order, largely axisymmetric dipolar component, giving rise to long-lived polar coronal holes and an enhanced fraction of open magnetic flux. In contrast, solar maximum is characterised by a far more complex and multipolar topology driven by the frequent emergence of active regions at low and mid-latitudes, which disrupts the large-scale dipole and redistributes open flux across the solar surface (e.g., Mackay and Yeates, 2012; Babcock, 1961). These shifts in magnetic topology govern the global behaviour of coronal plasma, dictating the distribution of open versus closed field lines and, consequently, the global balance of plasma confinement and outflows. Extreme ultraviolet (EUV) spectroscopy provides direct diagnostics of these dynamics. Doppler shifts trace bulk motions along the line of sight, while excess line broadening (often expressed as a non-thermal velocity) can reflect unresolved wave activity, turbulence, or multi-component flows (e.g., Doschek et al., 2008). In spatially resolved solar observations, coronal holes are commonly associated with blue-shifted emission indicative of outflows (e.g., Hassler et al., 1999), whereas active regions show a mix of red-shifts in cooling loops and blueshifts at footpoints and peripheries (e.g., Tian et al., 2021). Non-thermal broadening is also frequently enhanced in regions exhibiting upflows, although the physical drivers are not understood, in part due to instrumental limitations (e.g., Doschek, 2012).

Stellar spectra are spatially unresolved and combine emission from multiple magnetic structures (e.g., Kowalski, 2024), so stellar atmospheric dynamics cannot be studied with solar-like spatial detail. Sun-as-a-star observations, where emission from the solar disk is spatially integrated (as in stellar spectra) but where the spatially-resolved contributors to the emission are known, can be useful in understanding how disk-integrated spectra represent the atmospheres of stars (Toriumi, 2026). Toriumi et al. (2020), for example, showed that integrated multi-wavelength emission retains information on the presence and evolution of active regions, plages, and coronal loops. One complication is that stellar activity studies usually use chromospheric diagnostics such as Ca II H&K (e.g., Costes et al., 2021), whereas resolved solar coronal studies often use optically thin EUV lines (e.g., McKevitt et al., 2026a). Nevertheless, on long timescales chromospheric and coronal diagnostics broadly trace the same underlying cycle in stars, and solar magnetic flux correlates with atmospheric emission from the chromosphere to the corona (Toriumi and Airapetian, 2022).

In these solar-stellar contexts solar atmospheric irradiance is typically measured using full-disk spatially-resolved broadband imaging, given the availability of such data from the Atmospheric Imaging Assembly onboard the Solar Dynamics Observatory (SDO/AIA; Pesnell et al., 2012; Lemen et al., 2012). The Extreme ultraviolet Variability Experiment on SDO (EVE; Woods et al., 2012) provides full-disk Sun-as-a-star spectra, but cannot isolate contributions from specific solar regions. No observatory currently provides sustained, high-cadence, spatially-resolved full-disk spectral diagnostics of the solar atmosphere (Ugarte-Urra et al., 2023). Small-field-of-view spectrographs such as the EUV Imaging Spectrometer onboard Hinode (Hinode/EIS; Kosugi et al., 2007; Culhane et al., 2007) provide unmatched plasma diagnostics, but typically cover less than 10% of the full solar disk so can miss large-scale coupling and distant destabilisation (Schrijver et al., 2013; Schrijver and Higgins, 2015). Therefore full-disk spectroscopic studies require mosaics assembled from sequences of individual observations.

Coronal activity cycles are known on stars (e.g., Hempelmann et al., 2006; Robrade et al., 2012). The ‘Sun as an X-ray star’ series (e.g., Orlando et al., 2000) used solar differential emission measure (DEM) analysis from the Soft X-ray Telescope on Yohkoh (Tsuneta et al., 1991; Ogawara et al., 1991) to infer the fraction of stellar disks occupied by active regions (filling factor). Orlando et al. (2001) found that active region filling factor is the main cause of solar-cycle emission-measure-variability. Also, Morgan and Taroyan (2017) used broadband SDO/AIA imaging to show that solar-cycle-related variability in full-disk emission is driven by the presence of active regions, but that active regions themselves display no clear solar-cycle-related variability in average temperatures, emission measures, or magnetic field strength. Such work supports the hypothesis that stellar coronal activity cycles are caused by active region filling factor. However, approaching this with spectroscopic data, rather than broadband imaging, would prove useful in challenging the hypothesis further.

To this end, we compile and analyse a set of 18 spatially resolved full-disk spectroscopic mosaics taken between 2013 and 2024 by Hinode/EIS, processed using the EISMaps pipeline (McKevitt et al., 2026a), and perform an analysis of the variability of log T K ∼ 6.2 plasma. We also release full-disk maps of intensity, Doppler velocity, and non-thermal velocity for log T K ∼ 4.7–7.1 plasma, which were processed but whose analysis fell outside the scope of this study.

Hinode/EIS performs different full-disk mosaic observations using both it is narrow slits (e.g., Brooks et al., 2015), and it is slots ¹ (e.g., Warren et al., 2014). In this study we use 18 full-disk spectroscopic narrow-slit mosaic scans, taken between 2013 and 2024, for which Level-1 data are available ² . To enable the full-disk scan to be completed within a typical 48-h Hinode observing window, the widest 2 arcsec slit is used and the scan step size is set to 4 arcsec (sparse rastering). The plate scale of 1 arcsec results in spectrally-resolved measurements of plasma in pixels of 4 arcsec (X) × 1 arcsec (Y).

This study finds that disk-integrated coronal EUV intensity tracks the solar cycle closely (r = 0.98; Figure 3), and that the distributions of Doppler and non-thermal velocity within the magnetic structures responsible for that emission show no clear modulation across the same interval. This strong cycle dependence at the integrated level, and near-invariance at the per-structure level, supports the hypothesis that active region filling factor is the primary driver of coronal intensity variability, rather than any change in the spectroscopic characteristics of the regions themselves. We note, however, that spectroscopic characteristics such as composition and thermal structure were not covered in this study. Orlando et al. (2001) reached a similar conclusion using analysis of soft X-ray DEM variability between the maximum and the minimum of solar cycle 22. Morgan and Taroyan (2017) subsequently supported this result using SDO/AIA broadband imaging across most of cycle 24 and some of cycle 25, demonstrating that the average temperature, emission measure, and photospheric magnetic field strength of individual active regions showed no systematic cycle-phase dependence. The spectroscopic analysis we present is sensitive to plasma motions and wave activity in a way that broadband imaging and DEM analysis is not, and so further supports the filling factor hypothesis.

The persistence of the well-known anti-correlation between upflows and non-thermal broadening in active regions throughout the cycle implies that the mechanism(s) responsible (e.g., impulsive footpoint heating, unresolved Alfvénic wave activity, or reconnection-driven outflow) operate whenever the magnetic topology of an active region is present, regardless of the global solar magnetic field. We caution, however, that our sample size of active region plasma at solar minimum is limited. The moderate correlation between active region intensity normalised by solid angle with the solar cycle (r = 0.52; Figure 3) is therefore curious. A pure filling-factor model would predict no residual correlation once solid angle is accounted for. The fact that one remains at moderate significance suggests that either the typical active region near solar maximum carries higher intrinsic intensity, perhaps reflecting higher photospheric flux and more complex magnetic geometry, or that diffuse emission from the outskirts of active regions, not fully captured by the SHARP-based masking, contaminates the solid angle normalisation. The former would be consistent with the known dependence of coronal luminosity on total unsigned magnetic flux, which depends not only on the area occupied by magnetic structures but also on their field strength (Pevtsov et al., 2003). In that sense, normalising only by active region solid angle would not necessarily remove all solar-cycle-dependent variability. Distinguishing between enhanced unsigned magnetic flux per unit active region area near solar maximum and contamination from diffuse emission at active region outskirts would require more directly comparing the spectroscopic dataset against HMI magnetic flux, a step that falls outside the scope of the present analysis.

There is some small variation seen with solar cycle in the peak non-thermal velocity values in the quiet Sun plasma distributions, where solar maximum is associated with slightly lower non-thermal velocities. This could be due to stronger emission reducing artificial enhancers of non-thermal velocity such as signal-to-noise ratio and stray light effects. However, intensity versus non-thermal velocity histograms would be required to properly consider this, something outside the scope of this study. The next-generation high-throughput SOLAR-C/EUVST spectrometer will enable such measurements to a much higher precision (McKevitt et al., 2026b). The quiet-Sun distributions appear to show variability resembling that of the full disk. We find this logical given the majority of the disk in all cases is quiet Sun plasma. However, we note that the active region masking technique we apply may also influence the quiet Sun distributions, where diffuse or extended emission from nearby active regions is considered as quiet Sun plasma.

The original motivation for assembling this dataset was to connect spatially resolved coronal spectroscopy to Sun-as-a-star observables, and in particular to ask whether any solar-cycle variability in disk-integrated line profiles can be explained by changes seen within specific magnetic structures. Modelling of X-ray activity cycles in solar analogues (e.g., Hempelmann et al., 2006; Robrade et al., 2012; Orlando et al., 2017; Coffaro et al., 2020) has consistently considered varying active region filling factors as the mechanism linking chromospheric and coronal cycles, with per-region plasma treated as more fixed. That has rested largely on the assumption that coronal loop physics is determined locally. Our results suggest, albeit with limited data points, that the rise phase of the solar cycle and the related magnetic complexity may complicate this relationship.