The Youngest (74 ka) Toba Tuff (YTT) super-eruption is traditionally interpreted as a single, short-lived event. However, new paleomagnetic data from both southern and northern non-welded exposures of YTT’s Units 2, 3, and 4 (in ascending order) challenge this model. Unit 2 samples were obtained from a thin ash layer sandwiched between two ignimbrites at two localities on the northern and southern sides of the caldera, whereas samples from Units 3 and 4 were collected from ignimbrites at one locality each on the southern side of the caldera. Mineral chemistry of matrix-glass and biotite from thin ash layers in both northern and southern deposits shows compositions consistent with YTT. Mean characteristic remanent magnetization (ChRM) directions from Units 3 and 4 are indistinguishable and consistent with those of the welded YTT. Mean ChRM directions of Unit 2 from two different localities are indistinguishable and exhibit a statistically distinct paleomagnetic direction from those of Units 3 and 4. Using a secular variation rate of 0.02 degrees/year, the estimated minimum time gap between Units 2 and 3–4 is ~ 807 ± 193 years. Temporal gaps recorded in non-welded proximal deposits are consistent with recent interpretations from distal marine tephra, suggesting that the YTT eruption was episodic and may have spanned several hundred to thousands of years. Graphical Abstract

Mt. Guntur (West Java, Indonesia) has remained in surface repose since its last significant eruption in 1847 CE. Nevertheless, persistent volcano-tectonic seismicity and seismic tomographic anomalies indicate the possible presence of an active, melt-bearing magmatic system beneath the volcano. This study examines whole-rock geochemistry, high-resolution mineral chemistry, quantitative textural analysis (Crystal Size Distribution; CSD), and geothermobarometry of the youngest eruptive products to reconstruct the subvolcanic architecture. The results delineate a vertically extensive transcrustal plumbing system, consisting of a shallow, crystal-rich reservoir (1.0–3.0 kbar; approximately 3.5–11 km) and a deeper staging zone (approximately 3.0–3.5 kbar; 11–13 km, assuming an average crustal density of 2.7 g/cm³). This configuration spatially corresponds with seismic Low Velocity Zones (LVZs) and high-attenuation anomalies identified at 5–8 km depth in previous geophysical studies. Textural and micro-analytical data demonstrate a marked shift in magmatic dynamics over time. The mature GL-18 unit (>50 vol% crystallinity) records prolonged, closed-system fractional crystallization within a stable reservoir. It represents differentiated magma confined within a rigid crystal framework. In contrast, the younger basaltic units (GL-17, GL-11, GL-10) exhibit evidence of active disruption driven by successive mafic recharge events. These recharges largely replaced the resident interstitial melt, as indicated by primitive mineral compositions and distinct textural features that reflect pulsed thermal perturbations. The historical 1847 CE GL-17 eruption reached its final pre-eruptive equilibration at shallow depths (less than 1.5 kbar; approximately 1110–1125 °C), where the primitive magma was thermally buffered by cooler, partially crystallized portions of the resident reservoir. These findings classify Mt. Guntur as a system with Potentially Active Magmatic Storage (PAMS). Although persistent magmatic heat sustains regional geothermal fields, ongoing recharge dynamics suggest a continued potential for future volcanic unrest. As petrology alone cannot predict eruptions, a multidisciplinary monitoring framework that integrates continuous geophysical, deformation, and gas observations is necessary to mitigate risks in the Garut Basin.