betaFeedback

WorksPhysics and Astronomy

Detection and Statistics of large-scale Sector Boundary Crossings in the Solar Wind: The SBCross Catalog

Abstract

Heliospheric sector boundary crossings (SBCs), composed of the Heliospheric Current Sheet (HCS), Heliospheric Plasma Sheet (HPS), and surrounding plasma halo, represent fundamental structures in the solar wind that influence interplanetary dynamics and space weather. We present the SBCross catalog, an automated, publicly available database containing 1,076 SBC events derived from in-situ observations spanning more than 30 years (1995–2026), covering Solar Cycles 23 and 24 alongside the rising and maximum phases of Solar Cycle 25. By combining interplanetary magnetic field (IMF) and 255~eV suprathermal electron pitch angle distributions (PAD) data, our algorithm reliably captures large-scale polarity transitions, with 76%76\% of detected SBCs exhibiting clear HPS signatures (β>1β> 1, npeak>10 cm−3n_{\mathrm{peak}} > 10\mathrm{\,cm^{-3}}). Across the catalog, we find an average occurrence rate of ≈2.6\approx 2.6 events per Carrington rotation, a median event duration of ≈9.0\approx 9.0,hours, and a median physical thickness of ≈7.4×106\approx 7.4 \times 10^6,km. Furthermore, approximately ≈15%\approx 15\% of detected SBCs interact with interplanetary coronal mass ejections (CMEs) on average, with this proportion rising up to 33%33\% during solar maximum. These CME-associated events exhibit systematically longer durations (median of 10.010.0,hours), 54%54\% larger physical thicknesses, and higher average solar wind velocities. Notably, all high-velocity SBC events (>600 km s−1> 600\mathrm{\,km\,s^{-1}}) are exclusively driven by CME interactions, demonstrating local plasma pile-up and structural entanglement resulting from fast CMEs overtaking the ambient slow background. The SBCross catalog serves as an updated successor to the classic SBC list by L. Svalgaard, delivering a standardized, open-access resource for the heliophysics community to systematically investigate sector boundary transitions.

Introduction

The heliosphere is continuously being replenished by the solar wind, a magnetized plasma stream expanding outward from the solar corona into interplanetary space [39]. Embedded within this global outflow is the Heliospheric Current Sheet (HCS; e.g., [49, 54, 46]), which represents the fundamental boundary separating regions of opposite magnetic polarity. Enclosing the HCS is the Heliospheric Plasma Sheet (HPS; [6]), a region characterized by enhanced proton density, elevated plasma β\beta, a depressed helium-to-hydrogen ratio (He/H), and reduced magnetic field strength and temperature ([17, 5, 56]). Observational manifestations of transitioning across these large-scale structures in in-situ data are defined as sector boundary crossings (SBCs). Crucially, SBCs are identified by directional reversals in the interplanetary magnetic field (IMF) accompanied by abrupt shifts in the suprathermal electron heat flux ([12, 21, 10]). Consequently, pitch angle distributions (PADs) of suprathermal electrons are frequently used to trace these magnetic polarity transitions (e.g., [11, 18]).

The location where the suprathermal electron strahl flips alignment from parallel to antiparallel (or vice versa) relative to the magnetic field is traditionally termed the “true” sector boundary [20, 22]. While the true sector boundary is often synonymous with the HCS [3], N. U. Crooker et al. [9] demonstrated that the two features can become spatially or temporally misaligned due to interchange reconnection. Furthermore, a single sector boundary transition may contain multiple embedded HCS crossings [13, 28].

The structure of the region where the large-scale IMF configuration flips can be understood as a multilayer hierarchy. At the core lie one or multiple thin HCSs [13], high-β\beta current layers lasting from minutes to hours that maintain dynamic pressure balance with their surroundings. Within this localized view, the characteristic drop in magnetic field magnitude across the crossing is compensated primarily by a sharp rise in plasma density [56]. Individual HCSs exhibit typical thicknesses ranging from 600 km to 1.1×1061.1 \times10^{6} km, with an average scale on the order of 10410^{4} km [56, 29, 4].

Surrounding the thin HCS is the broader HPS, whose spatial extent spans from 3.2×1053.2 \times10^{5} km to 4.7×1064.7 \times10^{6} km [56, 29, 45] (C.-C. Wu et al. 2019). High-resolution analyses around the HPS reveal that this central plasma enhancement is itself nested within a broader, lower-density halo spanning several degrees in heliospheric latitude [2, 27]. Superposed epoch studies centered on HCS events demonstrate that these nested structures represent the interplanetary extension of coronal helmet streamers, where dense streamer stalks (∼1∘\sim1^\circ–2∘2^\circ wide) are embedded within wider coronal plasma outflows (∼15∘\sim15^\circ–20∘20^\circ) [5, 17, 2].

The spatial alignment of the HCS and the surrounding dense HPS plays a critical role in modulating the propagation dynamics and geoeffectiveness of Interplanetary Coronal Mass Ejections (CMEs). When fast, transient CMEs erupt from the solar corona, their path through interplanetary space is fundamentally shaped by these large-scale ambient structures. A fast CME overtaking a slow-moving HPS leads to a dynamic collision; the high-density plasma of the sheet resists the expansion of the CME, leading to a plasma pile-up and strong structural deformation of the CME’s shock front [34, 30]. This structural distortion and compression can amplify the ambient plasma density and plasma β\beta at the interface boundary. Crucially, when the magnetic flux rope of an CME interacts with the HCS/HPS structure, the localized magnetic field is compressed, amplifying any southward magnetic field component (BzB_z) within the interaction region [34, 52]. When these heavily compressed, high-density structures interact with Earth’s magnetosphere, they deliver a spike in dynamic pressure along with possibly long periods of strong southward BzB_z [14]. This combination regularly triggers unexpectedly severe, multi-phase geomagnetic storms that are more intense than those driven by isolated CMEs traveling through a quiet solar wind [e.g. 51, 23]. Several observational and modeling studies have also shown that CMEs are frequently deflected toward the HCS, which can lead to signification deviations in their expected propagation paths and complicate space weather forecasting [e.g. 19, 24, 55, 53]. The HCS / HPS may also substantially affect the propagation, transport and characteristics of solar energetic particles (SEPs; e.g., [1, 33].

In this study, we present a comprehensive statistical analysis of SBCs and provide a catalog spanning over 30 years of observations (1995–2026), covering entirely Solar Cycles 23 and 24, as well as the rising and maximum phases of Solar Cycle 25. The study details the distinctive physical properties of SBCs and the surrounding plasma environment derived from a systematic combination of IMF and PAD data. A major focus is the identification and analysis of SBC events that directly interact with interplanetary CMEs. As discussed above, the plasma environment during SBCs can distort the local plasma topology, causing significant variations in CME geoeffectiveness, altering shock propagation dynamics, and creating substantial statistical scatter in their downstream profiles compared to isolated, non-interacting events [40, 51]. By linking our SBC events to an existing in-situ CME catalog, this study offers a reliable statistical baseline for exploring how these large-scale background structures influence, shape, and accelerate severe space weather in the near-Earth environment.

Methodology

Data

In this study we utilize low-res OMNI solar wind data (1 hour resolution; [38]) and Electron Pitch Angle Distribution (PAD) data from the Three-Dimensional Plasma instrument (3DP; [31]) on-board the Wind spacecraft. Specifically, we are employing the EESA-Low (Electron Electrostatic Analyzer-Low) sensor data to retrieve PAD data in the 255 eV channel. We utilize the full available data from 1995 to 2026.

To identify potential associations between the detected SBC events and CMEs, we utilize the Richardson-Cane CME list (R&C list; [41]) to extract the disturbance onset and the end of the magnetic obstacle until September 2025.

SBC detection

To extract and characterize heliospheric SBCs and the associated plasma environments such as the HCS, HPS and surrounding halo, we utilize two independent solar wind data products. The events derived from each product are subsequently combined to construct the final catalog.

Primarily, we utilize normalized pitch angle distribution (PAD) data at 255 eV to derive an asymmetry index (Iasym255eVI_{\mathrm{asym}}^{255\mathrm{eV}}) from the 15° (parallel) and 165° (anti-parallel) bins, defined as:

Iasym255eV=PAD15∘−PAD165∘PAD15∘+PAD165∘.(1)I_{\mathrm{asym}}^{255\mathrm{eV}} = \frac{\mathrm{PAD}_{15^\circ} - \mathrm{PAD}_{165^\circ}}{\mathrm{PAD}_{15^\circ} + \mathrm{PAD}_{165^\circ}}. \tag*{(1)}

This asymmetry index, which reflects the tentative outward or inward direction of the magnetic field, is binned to a 1-hour resolution to match the solar wind data and is further smoothed using a Gaussian kernel with a sigma (σ\sigma) of 6 hours. Utilizing this index, we define SBC events as transitions from 0.2 to −0.2 (and vice versa) occurring between two nearly uninterrupted 40-hour periods dominated by a single field direction (∣Iasym255eV∣≥0.2|I_{\mathrm{asym}}^{255\mathrm{eV}}| \ge0.2). The duration of the event is defined as the time elapsed between these two transitions.

The secondary approach for detecting SBC events utilizes the azimuth angle (AzAz) derived from the interplanetary magnetic field (IMF) components in the Geocentric Solar Ecliptic (GSE) coordinate system, computed as:

Az [rad]=arctan⁡(ByBx).(2)Az\ [\mathrm{rad}] = \arctan\left(\frac{B_y}{B_x}\right). \tag*{(2)}

To remove small-scale fluctuations and isolate large-scale sector transitions, we smooth the computed azimuth angle using a Gaussian kernel (σ=6,h\sigma= 6,\mathrm{h}). Smoothing is performed using circular averaging to prevent phase-wrapping artifacts at the periodic 0°/360° boundary. Data gaps shorter than 6 h were linearly interpolated, whereas larger gaps were assigned NaN values and explicitly excluded from the Gaussian smoothing process. From the distribution of these angles, we identify four distinct regions associated with inward (π\pi to π/2\pi/2) and outward (0 to −π/2-\pi/2) fields, as well as their respective transitions (π/2\pi/2 to 0 and −π/2-\pi/2 to −π-\pi). Similar to the PAD asymmetry index approach, sector boundary transitions are identified as the transition from an outward to an inward field (or vice versa). The periods of outward and inward fields must last nearly uninterrupted (>95%> 95\% of datapoints) for at least 40 hours. The duration of these transitions is defined as the time interval between the stable outward and inward field periods.

The parameters of the detection algorithm were determined through empirical optimization. The smoothing parameter (σ=6\sigma= 6 h) removes high-frequency fluctuations while preserving distinct sector boundary signatures. The 40 hour threshold was selected as a representative duration for a stable interplanetary magnetic sector (∼1.5−2\sim1.5 - 2 days), preventing short-lived polarity reversals or transient magnetic structures (e.g., small flux ropes or turbulence) from being misidentified as major sector boundary crossings. These choices were optimized by cross-referencing the automated detections against visual inspection and evaluating performance against the historical SBC catalog established by Leif Svalgaard ([48]) and hosted by the Wilcox Solar Observatory (WSO). Although an SBC catalog exists, a new one is necessary because the Svalgaard list is no longer maintained and provides only approximate boundary crossing times without detailed physical parameters. Sensitivity tests show that slight variations in these thresholds (e.g., 38 − 42 hours for polarity duration or σ=4−8\sigma= 4 - 8 hours for smoothing) yield nearly identical event selections, demonstrating that the final catalog is robust against minor parameter choices.

By performing this automated analysis on the available data from the start of 1995 to the end of 2025, we derive two distinct sets of events that are subsequently combined, with confidence based on their detection source. If a transition is simultaneously detected by both the PAD asymmetry index and the IMF azimuth angle, the event is assigned the highest reliability flag (R=highR = \textit{high}), and the duration derived from the asymmetry index is utilized to define the length of the SBC event. When an event is uniquely identified in the PAD asymmetry data product but absent in the IMF azimuth angle data, it is included in the combined catalog with a moderate reliability flag (R=moderateR = \mathit{moderate}). Conversely, if an event is detected solely via the IMF magnetic field data, it is recorded and flagged with the lowest reliability (R=lowR = \mathit{low}). These reliability flags simply indicate the data source used for detection, allowing community users to filter the SBCross catalog according to their specific research needs. Further, a manual inspection is performed to identify and remove the automated detection artifacts. For all categorized events, their respective physical properties are systematically recorded in the Sector Boundary Crossing (SBCross) catalog. As a representative example of this detection process, Carrington rotation 2069 is illustrated in Figure 1.

Figure 1. Overview of solar wind plasma and interplanetary magnetic field parameters observed during Carrington rotation 2069 (April 16 to May 13, 2008). From top to bottom, the panels display: (a) proton bulk velocity (blue, left axis) and proton number density (orange, right axis); (b) IMF components (BxB_x, ByB_y, BzB_z) in Geocentric Solar Ecliptic (GSE) coordinates; (c) IMF azimuthal angle (dark red, left axis) superimposed with the suprathermal electron asymmetry index evaluated at 255 eV (purple, right axis); and (d) the corresponding normalized electron pitch angle distribution (PAD) spectrogram at 255 eV, color-coded on a logarithmic scale. The solid bars with the dotted lines show the detected SBC intervals and the dashed lines the sector boundary transitions from the L. Svalgaard catalog.

Computing SBC thickness

To determine the physical thickness of SBCs, we calculated its cross-sectional width by tracking the dynamic orientation of the solar wind. Because traditional Minimum Variance Analysis can degrade when applied to low-cadence, hourly averaged magnetic data across multi-hour intervals [47], we adopted a robust geometric approach based on the Parker spiral configuration. Rather than assuming a fixed 45∘45^\circ Parker spiral angle, we determined the local Parker spiral angle (ψ\psi) which represents the expected orientation of the magnetic sector front relative to the radial direction, using the mean solar wind speed (VV) measured during the crossing:

ψ=arctan⁡(Ω⊙rV)(3)\psi= \arctan\left(\frac{\Omega_{\odot} r}{V}\right) \tag*{(3)}

where Ω⊙≈2.865×10−6 rad s−1\Omega_{\odot} \approx2.865 \times10^{-6}\ \mathrm{rad}\ \mathrm{s}^{-1} is the sidereal solar rotation rate and r≈1 AUr \approx1\ \mathrm{AU} is the heliocentric distance. We then projected the velocity along the boundary normal to find the normal impact velocity, Vnormal=Vcos⁡ψV_{\mathrm{normal}} = V \cos\psi. The total physical thickness (LL) was calculated by multiplying this normal velocity by the SBC duration (Δt\Delta t):

L=Vcos⁡ψ×Δt(4)L = V \cos\psi\times\Delta t \tag*{(4)}

Accounting for changing solar wind speeds avoids projection errors and keeps our thickness calculations accurate throughout the solar cycle.

Results

Statistics

For our dataset spanning the period from 1995 to 2025, we identify 1,076 SBC events. Among these, 788 are detected simultaneously in both the PAD and IMF data, 80 are found exclusively in the PAD data, and 208 are identified solely in the IMF data. By filtering these

events for properties indicative of clear HPSs (β>1\beta> 1 and npeak>10 cm−3n_{\mathrm{peak}} > 10\ \mathrm{cm}^{-3}; according to D. Winterhalter et al. [56]; N. U. Crooker et al. [9]), we find that 817 (76%) of the detected SBC events seem to include at least one HPS. Of the 817 HPS-associated events, 612 (75%) belong to the high-reliability group, 60 (7%) to the moderate-reliability group, and 145 (18%) to the low-reliability group.

Figure 2 illustrates the distribution of key SBC parameters, specifically the event duration, average solar wind velocity, average plasma beta, average and peak plasma density. The distribution features a median duration of approximately 9 hours, characterized by a Gaussian-like profile for events lasting less than ≈24\approx24 hours, alongside a prominent tail of longer durations. A total of 123 events have a duration exceeding 24 hours and are hereafter referred to as long-duration SBCs, with the longest recorded event lasting 128 hours (5.3 days) in 2004.

Figure 2. Distribution of SBC crossing characteristics for the 1,076 SBC events, categorized by magnetic polarity orientation (+, − in red; −, + in blue): (a) duration of SBCs, with values exceeding 72 hours grouped into the final bin; (b) average solar wind bulk velocity (VV); (c) average plasma beta (β\beta); (d) average proton density (nn); and (e) peak proton density (npeakn_{\mathrm{peak}}). Vertical dashed lines indicate overall median values across all events.

By averaging the solar wind bulk velocity over each event’s duration interval, we find a mean velocity of ≈370 km s−1\approx370\ \mathrm{km}\ \mathrm{s}^{-1} and a median velocity of ≈360 km s−1\approx360\ \mathrm{km}\ \mathrm{s}^{-1}. 95 events show an average plasma beta (β\beta) below one, suggesting no HPS is contained within. We also find that the detected SBCs are high-density structures with a median average density of ≈11 cm−3\approx11\ \mathrm{cm}^{-3} and a median peak density of ≈17 cm−3\approx17\ \mathrm{cm}^{-3}. Over the entire baseline, we observe an average occurrence rate of 30 to 40 events per year, which corresponds to roughly ≈2.6\approx2.6 events per Carrington rotation.

Figure 3 displays a superposed epoch analysis (SEA) of the solar wind bulk velocity, number density, integrated electron flux (255 eV channel), magnetic field magnitude and plasma beta for all 1,076 events, centered on the midpoint of the detected SBC intervals. The characteristic minimum/maximum values along with their 10th–90th and 20th–80th percentile ranges are summarized in Table 1.

ParameterFeatureΔt\Delta tMedian20th–80th Percentile10th–90th Percentile
Bulk Velocity VV (km s−1^{-1})Min−6.92-6.92 h351.0[318.0, 402.0][304.0, 442.0]
Number Density nn (cm−3^{-3})Max+0.58+0.58 h11.60[7.10, 18.20][5.60, 23.48]
Plasma Beta β\betaMax+0.58+0.58 h3.23[1.37, 8.14][0.80, 13.92]
Electron Flux Φ255eV\Phi_{255\mathrm{eV}} (cm−2^{-2} s−1^{-1} sr−1^{-1})Min−3.42-3.42 h1.54[0.84, 3.15][0.60, 4.72]
Magnetic Field Magnitude ∣B∣\lvert B\rvert (nT)Min+0.08+0.08 h4.22[2.39, 7.39][1.71, 9.73]

Table 1. Summary of superposed epoch analysis (SEA) extremum values and percentile ranges for all detected SBC events (N=1,076N = 1{,}076), corresponding to Figure 3. For each parameter, the median minimum or maximum value is listed along with its timing relative to the epoch center (Δt\Delta t), as well as the 20th–80th and 10th–90th percentile ranges. Note that the uncertainty in Δt\Delta t due to data cadence is approximately ±1\pm1 h.

Figure 3. Superposed epoch analysis of solar wind plasma and IMF magnetic field parameters relative to the center of the 1,076 detected SBCs (denoted by the vertical dashed line at Time = 0 days). The median duration of SBC events is indicated by the vertical dotted lines (at ±5.5\pm5.5 hours). From top to bottom, the panels display the temporal profiles of: (a) solar wind bulk velocity (VV), (b) solar wind number density (nn), (c) integrated electron flux in the 255 eV energy bin, (d) the IMF magnitude ∣B∣|B|, and (e) the plasma beta (β\beta). For all parameters, the solid curves trace the median profiles, while the darker and lighter shaded envelopes represent the 20th^{\mathrm{th}}–80th^{\mathrm{th}} and 10th^{\mathrm{th}}–90th^{\mathrm{th}} percentile ranges, respectively.

In the velocity profile (Figure 3a), a minimum occurs prior to the center at Δt=−6.92 h\Delta t = -6.92\ \mathrm{h}. Conversely, a pronounced density peak (Figure 3b) follows shortly after the midpoint (Δt=+0.58 h\Delta t = +0.58\ \mathrm{h}), coinciding with the maximum in plasma beta β\beta (Figure 3e). The integrated 255 eV electron flux (Figure 3c) reaches a minimum at Δt=−3.42 h\Delta t = -3.42\ \mathrm{h}. Finally, the magnetic field magnitude ∣B∣|B| (Figure 3d) shows a clear minimum centered on the midpoint (Δt=+0.08 h\Delta t = +0.08\ \mathrm{h}), corresponding to the expected magnetic field drop at the heliospheric current sheet crossing.

These distinct multi-instrument signatures show that our methodology successfully detects and extracts SBCs that statistically feature a central HCS often associated with a HPS structure. Furthermore, the velocity profile exhibits an increase after the average SBC interval, which is clearly associated with the onset of a Stream Interaction Region (SIR) and a trailing high-speed solar wind streams (HSSs). The corresponding post-interval enhancements in ∣B∣\lvert B\rvert and electron flux appear to map to the compressed plasma of the stream interaction regions and their respective stream interfaces preceding the HSSs (as shown by N. U. Crooker et al. [7]). However, distinguishing the density increase caused by the SIR from that of the SBC plasma environment remains difficult.

SBCs and CMEs

Interaction of CMEs with SBCs can significantly alter their geoeffectiveness. In space weather forecasting, particularly the heliospheric current sheet as discussed by M. Temmer et al. [52, 51], warrants further attention as crucial parameter. As such, we investigate SBC events that can be associated with in-situ measured CMEs as listed in the R&C list. We define an SBC event as CME-associated if any portion of a CME interval overlaps with the SBC interval, or if the start or end of a CME occurs within 12 hours of the SBC boundary boundaries. We further sort these interaction into three groups, where a CME precedes a SBC (occurs before but has no overlap), a CME succeeds a SBC (occurs after without an overlap) and events where the R&C listed interval overlaps with the SBC interval.

We find that 166 of the SBC events are interacting with at least one CME (note that we did not check for association with multiple CMEs), which corresponds to about 15% of all recorded events. As expected, the occurrence of CME-associated events is strongly modulated by the solar cycle (Figure 4d). While the annual count of total SBC events shows no solar cycle dependence (Spearman correlation coefficient ρ=0.07\rho= 0.07, p=0.699p = 0.699), CME-associated SBC counts exhibit a strong positive correlation with the annual average sunspot number (ρ=0.70\rho= 0.70, p<0.001p < 0.001). Consequently, the relative fraction of CME-associated SBC events is also strongly correlated with the sunspot number (ρ=0.75\rho= 0.75, p<0.001p < 0.001), reaching up to 33% during solar maximum. Figure 4 further shows the difference in duration (a) and solar wind velocity (b) between SBC events with and without associated CMEs. The distribution clearly reveals that CME related events last longer (median duration is ≈1.5−2\approx1.5 - 2 hours longer) and have a faster average solar wind velocity (about ≈60\approx60 km s−1^{-1} faster). We further find that the proportion of SBC events associated with CMEs is larger for long duration events (29%) than of events with a duration of below one day (14%). Figure 4c shows the distribution of SBC thickness across all 1,076 events, yielding an overall median thickness of ≈7.4×106\approx7.4 \times10^{6} km (10th–90th percentile range: [3.9,23.0]×106[3.9, 23.0] \times10^{6} km). When categorized by CME association, CME-associated events exhibit a median thickness of 11.1×10611.1 \times10^{6} km, which is roughly 54% larger than that of non-CME events (7.2×1067.2 \times10^{6} km). Similar to our findings for event duration, we find no direct correlation between the solar cycle (as approximated by the sunspot number) and SBC thickness (not shown). However, solar maximum features a significantly larger proportion of CME-associated events, which are statistically both longer in duration and physically thicker. All results discussed in this section are statistically significant based on Mann-Whitney UU tests (H. B. Mann & D. R. Whitney [37]), with non-normality confirmed via Shapiro–Wilk tests (S. S. Shapiro & M. B. Wilk [44]). A detailed summary of these statistical results is provided in Appendix Table 2.

Figure 4. Distribution of SBC characteristics for CME-associated (orange) and non-CME (blue) events: (a) SBC duration, with values exceeding 72 hours grouped into the final bin; (b) average solar wind bulk velocity (VV); (c) physical thickness of the transition regions, with values exceeding 50×10650 \times10^{6} km grouped into the final bin; and (d) full time series from 1996 to 2026 comparing the annual frequency of total detected SBC events (blue bars, left axis), the annual CME association ratio (orange dashed line, right axis), and the monthly sunspot number from SILSO1 (grey curve, far-right axis). Dotted vertical lines in panels (a)–(c) indicate the overall, CME-associated, and non-CME median values.

Figure 5 shows the mean bulk velocity of the SBC events plotted against their duration, with color indicating CME association. We find that bulk velocity and event duration are essentially uncorrelated across all categories (ρ≈−0.19\rho\approx-0.19 to 0.090.09, with p>0.10p > 0.10 for all CME-

Figure 5. Scatter plots of average solar wind velocity versus SBC duration, categorized by CME association: (a) non-CME events (N=903N = 903); (b) CME before the SBC (N=26N = 26); (c) CME after the SBC (N=16N = 16); and (d) CME overlapping with the SBC (N=124N = 124). Error bars represent standard deviations.

associated subsets). However, 17 out of the 22 SBC events with an average velocity exceeding 550 km s−1^{-1}, as well as all events exceeding 600 km s−1^{-1}, overlap with CMEs. Furthermore, CMEs occurring even shortly after an SBC event do not systematically alter the observed average velocity or duration of the SBC structure compared to isolated events. In contrast, SBC events that directly overlap with CMEs exhibit a significantly larger scatter in both duration and velocity, despite lacking any correlation.

The SBCross Catalog

Using the automated method described in this study, we developed a new, comprehensive catalog of SBCs, including their surrounding plasma properties. The catalog is published online at https://sbcross.helsinki.fi, hosted by the University of Helsinki, Finland, in collaboration with the University of Graz, Austria. The detection algorithm is permanently archived on Zenodo2 and maintained on GitHub3. Currently, the catalog spans data from 1995 to 2026 and contains the date, duration, and multiple plasma and magnetic field properties, as well as a confidence estimation and CME association. We aim to maintain the catalog through ongoing updates by S.G. Heinemann, D. Price, and M. Temmer. The static dataset is archived on Zenodo4.

Discussion

Using PAD and IMF data, we identified 1,076 magnetic field sector changes of which 76% are associated with high-density HPS structures. The identified events are compiled in the SBCross catalog. We evaluated the performance of our identification method by benchmarking our catalog against the comprehensive sector boundary list compiled by L. Svalgaard ([48]). Out of the 1,076 total events in our catalog, 1,030 fall within the temporal overlap period with Svalgaard’s list (1995 to June 2024, marking its final update), compared to the 844 events recorded by Svalgaard. Overall, 748 entries in the SBCross catalog correspond directly to events in Svalgaard’s list, representing an 89% agreement within a ±1.5-day window of the cataloged timestamps. Additionally, our automated method identified 182 events not present in Svalgaard’s list, whereas 96 entries from Svalgaard’s catalog were not detected by us. A detailed comparison and visual inspection of these discrepancies revealed two main factors driving the variations between the catalogs. First, events listed by Svalgaard but missed by SBCross often occur during complex intervals with overlapping CMEs or CME-like structures. These disturbances obscure clear sector boundary transitions in PAD or IMF orientation, precluding automated detection, while a small subset showed no distinct SBC signatures at all. Second, SBCross identifies additional boundary crossings because it requires a shorter minimum sector duration before and after a reversal (40 hours compared to Svalgaard’s 4 days). This shorter stability requirement allows the algorithm to detect distinct, well-defined SBCs bounded by shorter-lived IMF sectors that Svalgaard’s catalog filtered out. Across the full catalog, we find an average of ≈ 2.6 events per Carrington rotation. This value is within the range reported by WSO (2–4 days), though lower than the 3.4 events per rotation calculated by K. Liou & C.-C. Wu [32], who manually identified 1,197 HCS-focused events from Wind data between 1995 and 2020.

From the duration of the SBCs, we estimated their median thickness to be ≈ 7.4×1067.4 \times10^{6} km, with 10th and 90th percentiles of 3.9×1063.9 \times10^{6} km and 2.3×1072.3 \times10^{7} km, respectively. Comparing this to measurements of the HPS, which range from 3.2×1053.2 \times10^{5} to 4.7×1064.7 \times10^{6} km ([56, 29, 45] (C.-C. Wu et al. 2019)), clearly highlights that the intervals defined in this study contain not only a possible HPS but also the surrounding plasma halo. This is further supported by the average plasma beta of 3.5 within the SBCs. L. Klein & L. F. Burlaga [26] identified two types of SBCs: those with durations < 10 min and those with durations ≥ 3 hours. Due to our detection methodology (e.g., smoothed hourly data), all of our detected SBCs fall into the second category.

As shown in Figure 3, this approach successfully captures the density enhancement and magnetic field minimum typically associated with the HCS, HPS, and surrounding halo during a sector boundary transition. The superposed epoch analysis yields velocity and density profiles similar to those presented by J. T. Gosling et al. [17] for 23 well-defined sector boundaries.

C.-C. Wu et al. (2019) noted in their review a clear solar cycle dependence in HPS thickness, characterized by a linear relationship between logarithmic thickness and sunspot number. For the SBCs identified in this study (76% showing HPS signatures), we find virtually no correlation between sunspot number and either event duration (ρ≈−0.12\rho\approx-0.12, p<0.001p < 0.001) or physical thickness (ρ≈−0.03\rho\approx-0.03, p=0.234p = 0.234).

We further note that SBCs are not exclusively tied to stream interaction regions (SIR; J. T. Gosling et al. [17]; ); G. Borrini et al. [5]). We observe a distinct rise in solar wind velocity approximately one day after the density peak, which is accompanied by a peak in the suprathermal electron flux indicative of a stream interaction region (e.g., [16, 40]). Subsequent manual inspection reveals that the density peak in SBCs often precedes an SIR, regularly coincides with one, or exhibits a very narrow separation, manifesting as a distinct double-peak structure in the density profile (see also [43], Figure 3.16).

The solar cycle modulation observed in CME-associated SBC events peaks at approximately 33% during solar maximum and drops nearly to zero during solar minimum, agreeing well with the established solar cycle dependence of CME occurrence rate ([25]). High solar wind velocities (> 600 km s−1^{-1}) are observed exclusively during direct overlaps of CMEs and SBCs, providing strong in-situ evidence of faster CMEs actively interacting with a slower background SBC. This dynamic collision can cause local plasma pile-up, which expands the event duration and generates highly complex ejecta ([9, 36]).

Interestingly, we find no significant correlation when a CME merely precedes or follows an SBC event without overlapping. This implies that structural modifications require direct physical interaction, supporting the hypothesis that an encroaching CME does not simply displace the sector boundary structure, but embeds itself within or may even replace the local HCS/HPS structure ([15]). This mechanism explains the large scatter in duration and velocity observed during SBC events that interact with CMEs; individual CME–SBC collisions are inherently disruptive ([42, 35]) and unpredictable, leaving behind a severely distorted or entangled solar wind structure (e.g., [50]).

Furthermore, N. U. Crooker et al. [8] argued that CMEs associated with SBCs may utilize coronal streamer structures as funnels into the heliosphere, a phenomenon that may be directly linked to their enhanced geoeffectiveness.

Summary and Conclusions

In this study, we presented a comprehensive catalog of 1,076 sector boundary crossings (SBCs) spanning from 1995 to 2026. Beyond identifying these distinct events, we characterized their physical properties—focusing on event durations and the surrounding plasma environment, and investigated potential interactions between SBC structures and CMEs. Our primary conclusions are summarized as follows:

  • We introduced the SBCross catalog, an actively maintained dataset of 1,076 sector boundary crossings published online at https://sbcross.helsinki.fi. This catalog is designed to serve as a modern, updated successor to the classic sector boundary crossing list compiled by L. Svalgaard.

  • We find that approximately three out of four SBCs (76%) are associated with at least one HPS structure.

  • At 1 AU, SBCs exhibit a median duration of approximately 9 hours, with a notable population of long-duration events (>1> 1 day) accounting for 11% of all cases. Beyond duration, our broader statistical comparison across plasma and magnetic parameters (including velocity, density, plasma beta, and physical thickness) demonstrates significant systematic differences between event subpopulations.

  • On average, approximately 15% of the identified SBC events were directly associated with CMEs, with this fraction reaching up to 33% during solar maximum. These interaction events are characterized by systematically longer durations and higher average velocities compared to isolated SBC crossings. Notably, all high-velocity SBC events (>600 km s−1> 600\ \mathrm{km}\ \mathrm{s}^{-1}) were found to be exclusively driven by these CME interactions.

Acknowledgments

We acknowledge use of NASA/GSFC’s Space Physics Data Facility’s OMNIWeb/CDAWeb service, and OMNI data. SGH acknowledges funding from the Research Council of Finland (Academy Fellowship) [370747; RIB-Wind]. SGH thanks his chair for always having his back. This research was funded in whole or in part by the Austrian Science Fund (FWF) [10.55776/PAT2280725]. MO is part-funded by Science and Technology Facilities Council (STFC) grant number UKRI1207 and Natural Environment Research Council (NERC) grant number NE/Y001052/1. SGH, EK and DP acknowledge the Research Council of Finland Centre of Excellence project SpaceResilience [1374096]. ERC Synergy project GER-ACLE (4100617), funded by the EU under grant agreement 101166910. Open access funded by Helsinki University Library.

Author Contributions

SGH conducted the analysis and led the manuscript preparation. DP prepared the online catalog. MT contributed the core concept for this study. All authors participated in revising and editing the manuscript.

Appendix

Additional Figures

Table 2. Statistical summary comparing All, CME-associated, and Non-CME events split by non-parametric (Median) and parametric (Mean) metrics. Non-normality (Shapiro-Wilk test p<0.001p < 0.001) is denoted by ✓ for All/CME/Non-CME groups. IQR represents the Interquartile Range (75th−25th75^{\text{th}} - 25^{\text{th}} percentile), and ±SD denotes the standard deviation. The pp-values are derived from a two-sided Mann-Whitney UU test comparing CME vs. Non-CME events.

ParameterAll EventsCME EventsNon-CME EventsShapiro-WilkMann-Whitney UU
Medians and Interquartile Ranges [IQR]
SBC Duration [h]9.00[7.00]10.00[14.00]8.50[6.00]✓ / ✓ / ✓< 0.001
Avg Velocity [km s−1^{-1}]357.67[71.38]420.33[108.91]352.44[61.93]✓ / ✓ / ✓< 0.001
Avg Density [cm−3^{-3}]11.26[7.39]9.15[7.00]11.58[7.11]✓ / ✓ / ✓< 0.001
Peak Density [cm−3^{-3}]16.60[12.30]16.80[13.80]16.60[12.00]✓ / ✓ / ✓0.319
Avg Beta3.51[4.48]1.37[2.10]4.11[4.80]✓ / ✓ / ✓< 0.001
SBC Thickness [10610^{6} km]7.38[5.55]11.13[17.21]7.18[4.86]✓ / ✓ / ✓< 0.001
Means and Standard Deviations [±SD]
SBC Duration [h]12.37 ± 11.6316.60 ± 14.8811.60 ± 10.77✓ / ✓ / ✓< 0.001
Avg Velocity [km s−1^{-1}]372.53 ± 63.61433.91 ± 92.83361.25 ± 48.92✓ / ✓ / ✓< 0.001
Avg Density [cm−3^{-3}]12.53 ± 6.4810.71 ± 6.8612.87 ± 6.36✓ / ✓ / ✓< 0.001
Peak Density [cm−3^{-3}]19.76 ± 11.7119.90 ± 13.9219.73 ± 11.26✓ / ✓ / ✓0.319
Avg Beta5.22 ± 5.352.34 ± 3.005.76 ± 5.52✓ / ✓ / ✓< 0.001
SBC Thickness [10610^{6} km]11.31 ± 12.5919.51 ± 19.889.80 ± 10.04✓ / ✓ / ✓< 0.001

Table 2.

References

  1. [1]Battarbee, M., Dalla, S., & Marsh, M. S. 2017, ApJ, 836, 138, doi: 10.3847/1538-4357/1/138
  2. [2]Bavassano, B., Woo, R., & Bruno, R. 1997, Geophys. Res. Lett., 24, 1655, doi: 10.1029/97GL01630
  3. [3]Behannon, K. W., Neubauer, F. M., & Barnstorf, H. 1981, J. Geophys. Res., 86, 3273, doi: 10.1029/JA086iA05p03273DOI
  4. [4]Blanco, J. J., Rodriguez-Pacheco, J., Hidalgo, M. A., & Sequeiros, J. 2008, Journal of Atmospheric and Solar-Terrestrial Physics, 70, 226, doi: 10.1016/j.jastp.2007.08.030DOI
  5. [5]Borrini, G., Gosling, J. T., Bame, S. J., Feldman, W. C., & Wilcox, J. M. 1981, J. Geophys. Res., 86, 4565, doi: 10.1029/JA086iA06p04565DOI
  6. [6]Burlaga, L. F., Mish, W. H., & Whang, Y. C. 1990, J. Geophys. Res., 95, 4247, doi: 10.1029/JA095iA04p04247DOI
  7. [7]Crooker, N. U., Appleton, E. M., Schwadron, N. A., & Owens, M. J. 2010, Journal of Geophysical Research (Space Physics), 115, A11101, doi: 10.1029/2010JA015496DOI
  8. [8]Crooker, N. U., Gosling, J. T., & Kahler, S. W. 1998, J. Geophys. Res., 103, 301, doi: 10.1029/97JA02774DOI
  9. [9]Crooker, N. U., Huang, C.-L., Lamassa, S. M., et al. 2004a, Journal of Geophysical Research (Space Physics), 109, A03107, doi: 10.1029/2003JA010170DOI
  10. [10]Crooker, N. U., Kahler, S. W., Larson, D. E., & Lin, R. P. 2004b, Journal of Geophysical Research (Space Physics), 109, A03108, doi: 10.1029/2003JA010278DOI
  11. [11]Crooker, N. U., Larson, D. E., Kahler, S. W., Lamassa, S. M., & Spence, H. E. 2003, Geophys. Res. Lett., 30, 1619, doi: 10.1029/2003GL017036DOI
  12. [12]Crooker, N. U., Lazarus, A. J., Lepping, R. P., et al. 1996, Geophys. Res. Lett., 23, 1275, doi: 10.1029/96GL00031DOI
  13. [13]Crooker, N. U., Siscoe, G. L., Shodhan, S., et al. 1993, J. Geophys. Res., 98, 9371, doi: 10.1029/93JA00636DOI
  14. [14]Echer, E., Gonzalez, W. D., Tsurutani, B. T., & Gonzalez, A. L. C. 2008, Journal of Geophysical Research (Space Physics), 113, A05221, doi: 10.1029/2007JA012744DOI
  15. [15]Forsyth, R. J., Balogh, A., Smith, E. J., & Gosling, J. T. 1997, Geophys. Res. Lett., 24, 3101, doi: 10.1029/97GL03099DOI
  16. [16]Gosling, J. T. 1996, ARA&A, 34, 35, doi: 10.1146/annurev.astro.34.1.35DOI
  17. [17]Gosling, J. T., Borrini, G., Asbridge, J. R., et al. 1981, J. Geophys. Res., 86, 5438, doi: 10.1029/JA086iA07p05438DOI
  18. [18]Gosling, J. T., Eriksson, S., & Schwenn, R. 2006, Journal of Geophysical Research (Space Physics), 111, A10102, doi: 10.1029/2006JA011863DOI
  19. [19]Isavnin, A., Yourildas, A., & Kilpua, E. K. J. 2014, SoPh, 289, 2141, doi: 10.1007/s11207-013-0468-4DOI
  20. [20]Kahler, S., & Lin, R. P. 1994, Geophys. Res. Lett., 21, 1575, doi: 10.1029/94GL01362DOI
  21. [21]Kahler, S. W., Crooker, N. U., & Gosling, J. T. 1996, J. Geophys. Res., 101, 24373, doi: 10.1029/96JA02232DOI
  22. [22]Kahler, S. W., & Lin, R. P. 1995, SoPh, 161, 183, doi: 10.1007/BF00732092DOI
  23. [23]Kataoka, R., Shiota, D., Kilpua, E., & Keika, K. 2015, Geophys. Res. Lett., 42, 5155, doi: 10.1002/2015GL064816DOI
  24. [24]Kay, C., Opher, M., & Evans, R. M. 2015, ApJ, 805, 168, doi: 10.1088/0004-637X/805/2/168DOI
  25. [25]Kilpua, E., Koskinen, H. E. J., & Pulkkinen, T. I. 2017, Living Rev. Sol. Phys., 14, 5, doi: 10.1007/s41116-017-0009-6DOI
  26. [26]Klein, L., & Burlaga, L. F. 1980, J. Geophys. Res., 85, 2269, doi: 10.1029/JA085iA05p02269
  27. [27]Lacombe, C., Salem, C., Mangeney, A., et al. 2000, Annales Geophysicae, 18, 852, doi: 10.1007/s00585-000-0852-x
  28. [28]Lavraud, B., Fargette, N., Réville, V., et al. 2020, ApJL, 894, L19, doi: 10.3847/2041-8213/ab8d2dDOI
  29. [29]Lepping, R. P., Szabo, A., Peredo, M., & Hoeksema, J. T. 1996, Geophys. Res. Lett., 23, 1199, doi: 10.1029/96GL00658DOI
  30. [30]Liewer, P. C., Hall, J. R., Howard, R. A., et al. 2011, Journal of Atmospheric and Solar-Terrestrial Physics, 73, 1173, doi: 10.1016/j.jastp.2010.09.004DOI
  31. [31]Lin, R. P., Anderson, K. A., Ashford, S., et al. 1995, SSRv, 71, 125, doi: 10.1007/BF00751328DOI
  32. [32]Liou, K., & Wu, C.-C. 2021, ApJ, 920, 39, doi: 10.3847/1538-4357/ac1586DOI
  33. [33]Liou, K., & Wu, C.-C. 2024, ApJ, 966, 16, doi: 10.3847/1538-4357/ad33c2DOI
  34. [34]Liu, Y. D., Luhmann, J. G., Mösti, C., et al. 2012, The Astrophysical Journal, 746, L15, doi: 10.1088/2041-8205/746/2/L15DOI
  35. [35]Liu, Y. D., Zhu, B., Ran, H., et al. 2024, ApJ, 963, 85, doi: 10.3847/1538-4357/ad1e56DOI
  36. [36]Lugaz, N., Temmer, M., Wang, Y., & Farrugia, C. J. 2017, SoPh, 292, 64, doi: 10.1007/s11207-017-1091-6DOI
  37. [37]Mann, H. B., & Whitney, D. R. 1947, The Annals of Mathematical Statistics, 18, 50, doi: 10.1214/aoms/1177730491DOI
  38. [38]Papitashvili, N. E., & King, J. H. 2020, OMNI Hourly Data [1996-2026], https://doi.org/10.48322/1shr-ht18 NASA Space Physics Data Facility
  39. [39]Parker, E. N. 1958, ApJ, 128, 664, doi: 10.1086/146579DOI
  40. [40]Richardson, I. G. 2018, Living Rev. Sol. Phys., 15, 1, doi: 10.1007/s41116-017-0011-zDOI
  41. [41]Richardson, I. G., & Cane, H. V. 2010, SoPh, 264, 189, doi: 10.1007/s11207-010-9568-6DOI
  42. [42]Romeo, O. M., Braga, C. R., Badman, S. T., et al. 2023, ApJ, 954, 168, doi: 10.3847/1538-4357/ace62eDOI
  43. [43]Schwenn, R. 1990, in Physics of the Inner Heliosphere I, ed. R. Schwenn & E. Marsch, 99DOI
  44. [44]Shapiro, S. S., & Wilk, M. B. 1965, Biometrika, 52, 591, doi: 10.2307/2333709
  45. [45]Simunac, K. D. C., Galvin, A. B., Farrugia, C. J., et al. 2012, SoPh, 281, 423, doi: 10.1007/s11207-012-0156-9DOI
  46. [46]Smith, E. J. 2001, J. Geophys. Res., 106, 15819, doi: 10.1029/2000JA000120DOI
  47. [47]Sonnerup, B. U. Ö., & Scheible, M. 1998, ISSI Scientific Reports Series, 1, 185
  48. [48]Svalgaard, L. 1972, J. Geophys. Res., 77, 4027, doi: 10.1029/JA077i022p04027DOI
  49. [49]Svalgaard, L., & Wilcox, J. M. 1976, Nature, 262, 766, doi: 10.1038/262766a0DOI
  50. [50]Temmer, M. 2021, Living Reviews in Solar Physics, 18, 4, doi: 10.1007/s41116-021-00030-3
  51. [51]Temmer, M., Heinemann, S. G., Dresing, N., Dumbovic, M., & Asvestari, E. 2026, ApJ, 1004, 203, doi: 10.3847/1538-4357/ae6dabDOI
  52. [52]Temmer, M., Dumbović, M., Martinić, K., et al. 2025, A&A, 695, A58, doi: 10.1051/0004-6361/202451479DOI
  53. [53]Wang, J., Hoeksema, J. T., & Liu, S. 2020, Journal of Geophysical Research (Space Physics), 125, e27530, doi: 10.1029/2019JA027530DOI
  54. [54]Wilcox, J. M., Hoeksema, J. T., & Scherrer, P. H. 1980, Science, 209, 603, doi: 10.1126/science.209.4456.603DOI
  55. [55]Winslow, R. M., Lugaz, N., Schwadron, N. A., et al. 2016, Journal of Geophysical Research: Space Physics, 121, 6092, doi: 10.1002/2015JA022307DOI
  56. [56]Winterhalter, D., Smith, E. J., Burton, M. E., Murphy, N., & McComas, D. J. 1994, J. Geophys. Res., 99, 6667, doi: 10.1029/93JA03481DOI

Paper details

vlnc:W_3jf5qh8a
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 2

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents