Ionospheric monitoring with the Chilean GPS eyeball during the South American total solar eclipse on 2nd July 2019

Solar eclipses provide a natural experiment to study the ionosphere’s response to rapid, large reductions in solar irradiance. An eclipse interrupts photoionization and thermospheric heating, altering ion production and loss, temperature balance, and driving atmospheric gravity waves (AGWs). Prior studies have reported TEC depressions and eclipse-induced AGWs, with responses dependent on eclipse timing, location, season, and ionospheric layer altitude. The 2 July 2019 total solar eclipse traversed the southern Pacific to Chile and Argentina, with totality in Chile’s Coquimbo region during late afternoon local time. This study aims to quantify and interpret the ionospheric TEC response across and away from the totality path using a dense north–south array of 24 GPS stations in Chile, and to assess the role of eclipse-generated AGWs and background winds in shaping the observed TEC perturbations.

GPS-based TEC has become a primary tool for monitoring eclipse-related ionospheric changes. Prior works reported TEC depressions of ~1–3 TECu (e.g., Afraimovich et al., 1998) and up to ~40–50% in equatorial anomaly regions (Liu et al., 1998). Studies also documented AGWs and traveling ionospheric disturbances (TIDs) linked to eclipses (e.g., Jakowski et al., 2008; Nayak & Yiğit, 2018), with wave periods of tens of minutes. Mechanistically, eclipse-induced stratospheric ozone cooling produces a moving thermal disturbance that can generate AGWs (Chimonas & Hines, 1970). Background winds can modulate wave propagation, with waves more likely to penetrate when propagating against the wind direction. GNSS analyses of the 21 August 2017 eclipse over the U.S. showed widespread TEC depletion and localized enhancements attributed to interactions of eclipse-induced waves and orography (Coster et al., 2017). This study builds on these insights by leveraging a latitudinal array straddling the 2019 eclipse’s totality in Chile to examine contrasting responses north and south of the path.

- Network and period: 24 continuously operating Chilean GPS stations from the Centro Sismológico Nacional (CSN) spanning the totality path and up to ~80% obscuration on both sides. Data from 1–3 July 2019 were analyzed over 18:00–23:00 UTC, with 2 July as eclipse day and 1 and 3 July as reference days. - GPS processing: Dual-frequency L1 (1575.42 MHz) and L2 (1227.60 MHz) RINEX data were processed with GPS_GOPI to obtain slant TEC (STEC). Vertical TEC (VTEC) was derived via a mapping function under a single-layer ionosphere model (SLIM) at 350 km, assigning ionospheric pierce points (IPP). - Data selection: Satellite elevation >20° to reduce multipath and environmental noise. Analyses focused on a common satellite track (PRN 13), which provided broad temporal coverage and elevation >50° during the eclipse interval (~19:30–22:00 UTC). - Unperturbed baseline: The mean VTEC of 1 and 3 July provided the unperturbed reference for each station. Eclipse-day VTEC (2 July) was compared to this baseline to quantify changes near the maximum eclipse time. - Eclipse geometry: Ground-level obscuration and totality path were compiled; conditions were also estimated at 350 km using the Verhulst et al. method to compare with ionospheric responses. - Wave analysis: VTEC time series were high-pass filtered to remove >2 h trends and analyzed via Morlet wavelet transform to identify AGWs/TIDs. Representative stations spanning far south, near-totality (south/north of the central line), and far north were examined. - Space weather context: Solar flares (SpaceWeatherLive) and geomagnetic indices (Kp, Dst from WDC Kyoto) were checked for 1–3 July 2019 to rule out external geomagnetic drivers. - Background winds: ERA5 reanalysis (31 km resolution) vector winds and wind speed at ~80 km altitude were mapped over South America at 0, 6, 12, and 18 UTC on 2 July 2019 to assess wind directions relevant to AGW propagation.

- Spatial TEC response: Minimal VTEC change at stations within totality (~0.39 TECu typical magnitude). South of totality, VTEC decreased relative to the unperturbed baseline, with maximum depression magnitude ~2.24 TECu (e.g., station cmba). North of totality, VTEC increased relative to baseline, with maximum magnitude ~3.89 TECu (e.g., station cifu). - Timing: TEC response commonly lagged eclipse onset by ~30 min (changes starting near ~20:00 UTC) and recovered ~20 min before eclipse end (~21:30 UTC), consistent with delayed ionospheric response to obscuration. - Latitudinal trend: The magnitude of VTEC deviation from baseline was smallest within the totality zone and increased with distance north or south from the central totality line (at 350 km height). - AGWs/TIDs: Wavelet analysis showed AGWs with periods ~30–60 min. These were strongest at stations north of the totality line (e.g., hsco and stations further north), weaker near/just south of totality, and largely absent at the far-southern station (chda) during the interval. - Background conditions: Space weather was quiet (max Kp = 3 at 03 UTC on 1 July; minimum Dst ≈ −18 nT at 22 UTC on 1 July; no flares). ERA5 winds were predominantly westerly, with southwesterlies over southern Chile during peak eclipse hours; winds converged north and south of Chile and strengthened to ~60 m/s to the east. This pattern favors northward AGW propagation and inhibits southward propagation. - Ancillary observations: VTEC time series on 1–3 July exhibited an overall increasing trend during the analysis window, particularly at northern stations, possibly related to PRN13’s northward IPP trajectory; the precise cause is left for future study.

The contrasting VTEC responses—decreases south of totality and increases north—together with minimal changes within totality, point to an interplay between direct eclipse-driven reductions in ionization and AGW-induced plasma density perturbations. Eclipse-induced stratospheric cooling generates AGWs that propagate upward and couple into the thermosphere-ionosphere, manifesting as TIDs. ERA5 background winds during the eclipse favored northward propagation (opposing winds aiding penetration) while inhibiting southward propagation, consistent with strong AGW signatures and positive VTEC perturbations north of the path and weaker signatures southward. The minimal change at totality can be understood as a near-cancellation between reduced photoionization and concurrent AGW-driven density modulation. The coastal Chilean setting bounded by the Andes suggests potential contributions from eclipse-induced mountain waves interacting with ozone-cooling-driven disturbances, analogous to localized TEC enhancements seen during the 2017 U.S. eclipse. Quiet geomagnetic conditions strengthen the conclusion that the observed TEC structures are eclipse/AGW-driven rather than magnetospheric.

Using a dense array of 24 Chilean GPS stations during the 2 July 2019 total solar eclipse, the study reveals a distinctive latitudinal TEC response: minimal changes within totality, significant decreases south of totality (up to ~2.24 TECu), and significant increases north of totality (up to ~3.89 TECu). Wavelet analysis identified strong AGWs with 30–60 min periods predominantly north of totality, consistent with ERA5 winds that supported northward wave propagation and impeded southward propagation. The findings highlight an interplay between eclipse-induced reductions in ionization and AGW/TID-driven density perturbations that can produce spatially asymmetric TEC responses. Future work should include long-term VTEC analyses to clarify diurnal and seasonal behaviors and dedicated observations (e.g., radar/lidar) to detect and characterize eclipse-induced mountain waves and better constrain coupling mechanisms.

- Lack of direct atmospheric wave observations (e.g., radar or lidar) prevents verification of eclipse-induced mountain waves and detailed characterization of AGW properties. - Analysis focused on a single GNSS satellite track (PRN 13) for uniformity; multi-satellite synthesis could improve spatial sampling. - Short temporal baseline (1 day before/after) limits assessment of broader diurnal/seasonal variability; authors note the need for long-term datasets. - Some inconsistencies between sign conventions in tabulated VTEC changes and narrative descriptions may complicate quantitative interpretation, though qualitative spatial patterns are robust.