Magma Chamber Response to Ice Unloading: Applications to Volcanism in the West Antarctic Rift System Authors: A. N. Coonin, C. Huber, J. Troch, M. Townsend, K. Scholz, B. S. Singer Published: 2024 ABSTRACT AbstractVolcanic activity has been shown to affect Earth's climate in a myriad of ways. One such example is that eruptions proximate to surface ice will promote ice melting. In turn, the crustal unloading associated with melting an ice sheet affects the internal dynamics of the underlying magma plumbing system. Geochronologic data from the Andes over the last two glacial cycles suggest that glaciation and volcanism may interact via a positive feedback loop. At present, accurate sea‐level predictions hinge on our ability to forecast the stability of the West Antarctic Ice Sheet, and thus require consideration of two‐way subglacial volcano‐deglaciation processes. The West Antarctic Ice Sheet is particularly vulnerable to collapse, yet its position atop an active volcanic rift is seldom considered. Ice unloading deepens the zone of melting and alters the crustal stress field, impacting conditions for dike initiation, propagation, and arrest. However, the consequences for internal magma chamber dynamics and long‐term eruption behavior remain elusive. Given that unloading‐triggered volcanism in West Antarctica may contribute to the uncertainty of ice loss projections, we adapt a previously published thermomechanical magma chamber model and simulate a shrinking ice load through a prescribed lithostatic pressure decrease. We investigate the impacts of varying unloading scenarios on magma volatile partitioning and eruptive trajectory. Considering the removal of km‐thick ice sheets, we demonstrate that the rate of unloading influences the cumulative mass erupted and consequently the heat released into the ice. These findings provide fundamental insights into the complex volcano‐ice interactions in West Antarctica and other subglacial volcanic settings. Introduction The ongoing stability of the West Antarctic Ice Sheet plays a crucial role in the prediction of modern sea‐level rise. One of the Earth's largest reservoirs of land ice, the West Antarctic Ice Sheet is particularly vulnerable to collapse as it is largely grounded below sea‐level (marine‐based). As sea‐level rises, such a marine‐based ice sheet becomes increasingly submerged, accelerating the retreat of the grounding line (Gomez et al., 2020). Further, the bedrock underneath the ice sheet slopes downward from the current margins toward the interior of the ice sheet, promoting runaway retreat (Fretwell et al., 2013; Van Wyk de Vries et al., 2018). Factors affecting the stability of the West Antarctic Ice Sheet include atmospheric CO2 emissions (e.g., Sutter et al., 2023), glacial isostatic adjustment and grounding line dynamics (e.g., Barletta et al., 2018; Gomez et al., 2015), ice shelf melting and calving due to interaction with a warming ocean (e.g., Stevens et al., 2020), and buttressing along the bedrock (e.g., Huybers et al., 2017; Pegler, 2018). While various studies investigate these parameters, one often overlooked factor is the potential for rapid ice loss to affect volcanic activity in the West Antarctic Rift System (WARS), which may promote further ice melting. The West Antarctic Ice Sheet sits atop the WARS, a partially hidden complex network of active rifts comprising one of Earth's largest volcanic provinces. More than 100 eruptive centers are thought to be currently active, some exposed and some subglacial (LeMasurier et al., 1990; Smellie & Edwards, 2016; Van Wyk de Vries et al., 2018). While eruptions from kilometers‐deep subglacial magmatic systems may not be directly observed, digital elevation models utilizing ice‐penetrating radar indicate intact subglacial volcanic cones (Behrendt et al., 2002; Corr & Vaughan, 2008; Schroeder et al., 2014; Van Wyk de Vries et al., 2018). Given that rifting in West Antarctica initiated around 66 Ma and glaciation began around 34 Ma, the presence of intact subglacial cones today suggests ongoing volcanic and/or magmatic activity (Spiegel et al., 2016). Otherwise, basal friction with the ice sheet would have thoroughly eroded these structures (Van Wyk de Vries et al., 2018). Additionally, high regional heat fluxes (Schroeder et al., 2014; Shapiro & Ritzwoller, 2004), seismic activity (Lough et al., 2013), and geomagnetic anomalies (Blankenship et al., 1993) in West Antarctica suggest that the rift remains active to the present (Geyer et al., 2023). The impacts of volcanic processes on the cryosphere and Earth's climate system, including decreased ice albedo (e.g., Bray, 1979; Möller et al., 2019) and outgassing of sulfur aerosols and CO2 (e.g., Aubry et al., 2022; Handler, 1989; Huybers & Langmuir, 2009) have been investigated over a range of timescales. While it is widely understood that nearby magmatic and/or volcanic activity can accelerate ice melting, expansive geochronological records from the Andes provide strong evidence that changes in ice volume have influenced volcanic activity over the last deglaciation (Singer et al., 2024). However, little is known regarding the mechanisms by which a shrinking ice load at the surface affects magmatic systems at depth. Several solid Earth phenomena have been proposed to be triggered by deglaciation and to play into complex two‐way processes: (a) changes in the crustal stress field, (b) deepening of the intersection of the adiabat and the solidus, and (c) a decrease in lithostatic pressure. As the ice load above magma chamber retreats, the crustal stress field changes with time. These variations in stress conditions affect the likelihood that the magma chamber ruptures and transports magma through the crust. Flexure of the crust during isostatic rebound creates tensile stresses in the upper crust, promoting dike formation (Mora & Tassara, 2019; Wilson & Russell, 2020). This principle applies regardless of the nature of the load being removed; for example, Satow et al. (2021) suggested that a decrease in sea‐level above the crust is capable of triggering dike injection from the Santorini magma chamber. In addition, a shrinking ice load can increase the depth at which the geotherm crosses the solidus of mantle rock; as pressure from the overburden is reduced, the rock is enabled to melt deeper in the mantle, thereby extending the zone of decompression melting. The consequence of this augmented zone of melting is an increased melt flux to the magma chamber (Sigmundsson et al., 2010; Van Wyk de Vries et al., 2018). Jull and McKenzie (1996) modeled increased melt generation in response to the removal of an axisymmetric ice sheet above a mid‐ocean ridge, with application to Iceland. Building on this work, Maclennan et al. (2002) examined this effect regarding volcanism in Iceland following the Last Glacial Maximum; they found that immediately after deglaciation, Icelandic eruption rates were 30–50 times higher than present‐day levels and persisted for over 1,000 years post‐deglaciation, owing to increased melt generation via decompression. One intuitive consequence of unloading above a crustal magma chamber is the decrease in lithostatic pressure (pressure due to the weight of the overlying material), which affects the thermodynamic state and stability of magmatic phases at depth. Jellinek et al. (2004) developed a theoretical model illustrating that periodic glacial loading changes (mirroring glacial cycles) can affect the frequency of dike propagation from a magma chamber. As the magmatic overpressure exerts tensile stresses on the magma chamber walls, fractures can dilate, facilitating dike propagation (Jellinek & DePaolo, 2003). Further, Jellinek et al. (2004) argue that glacial loading changes may be especially influential with respect to the lithostatic stress conditions on shallow (5–8 km depth within the crust) magma chambers. However, the effect of such decreasing lithostatic pressure on internal magma chamber dynamics has yet to be investigated. While past studies have proposed a correlation between deglaciation and volcanic activity (Huybers & Langmuir, 2009; Rawson et al., 2016), the physical mechanisms linking dynamic ice loads and the evolution of underlying magmatic systems remain unclear. Consequently, assessments of the stability of the West Antarctic Ice Sheet and projections of its contribution to future sea‐level rise lack potentially crucial information. Additionally, in West Antarctica, unloading‐triggered volcanic activity could pose additional hazards to future field campaigns. This highlights the need for physics‐based modeling of subglacial magma chambers subjected to retreating ice loads to quantify the magmatic and/or volcanic effects of a time dependent removal of overburden pressure. To address this gap, we simulate the evolution of magma chambers with depth, size, composition, and volatile content in the range of those to be expected beneath the West Antarctic Ice Sheet, subject to cooling (leading to crystallization and exsolution), viscoelastic stress relaxation, magmatic recharge, and ice unloading (see Figure 1). Specifically, we test different rates of decreasing lithostatic pressure due to a shrinking surface load. Chen et al. (2006) estimate the rate of deglaciation‐related surface unloading to be on the order of 10−5Pa·s−1 in West Antarctica based on GRACE satellite data. We consider this a conservative estimate for the modern‐day rate of ice unloading in West Antarctica, given the continuous increase in atmospheric CO2 levels through to the present, and test rates of ice removal two orders of magnitude higher to represent accelerated ice loss in West Antarctica. Some ice streams of West Antarctica have experienced even higher rates of ice thinning, on the order of 10−0.5Pa·s−1, in the last decade (Hogg et al., 2021). For simplicity, we assume a linear lithostatic pressure drop over time for the removal of a finite surface load. We apply this forcing to a thermomechanical magma chamber model (Degruyter & Huber, 2014; Scholz et al., 2023) parametrized for West Antarctic basalt magmas with mixed CO2 and H2O contents to simulate and characterize the response of magma chamber evolution and eruption behavior to ice unloading. While we examine West Antarctic magmatic and/or volcanic activity here, our physical model of unloading can be extended to investigate the response of magma chambers (in other settings) to the removal of any uniform surface load, such as sudden reductions in sea‐level over submarine volcanoes and erosion of overlying material during landslides or flank collapse. Schematic of the thermomechanical magma chamber model with simulated ice unloading from this study. Transparent arrows represent ice unloading as a decrease in the ice layer thickness over time. Methods To simulate WARS volcanoes, we employ a thermomechanical magma chamber model that includes water and carbon dioxide as magmatic volatile species (Degruyter & Huber, 2014; Scholz et al., 2023) tailored to West Antarctic Rift basalt compositions. Most exposed WARS deposits are part of the Marie Byrd Land (MBL) Province of West Antarctica (LeMasurier, 2013). To develop melting curves and H2O‐CO2 solubility parametrizations specific to WARS basalts, we utilize whole rock compositions of basalts from MBL from the GEOROC database (https://georoc.eu/) to compute an average anhydrous magmatic composition (DIGIS Team, 2023). The major oxide contents of the average anhydrous WARS basalt composition used in this study and several parameters for the magma chamber model are provided in Table 1. Sample compositions and locations used to determine the average composition as well as their associated publications are provided in Table S1 of the Supporting Information S1. Table 1Anhydrous Composition and Model Parameters for WARS BasaltMajor oxides for anhydrous WARS basalt compositionSiO2 TiO2 Al2O3 FeOT MnOMgOCaONa2OK2OP2O5 Totalwt. %46.883.2915.1812.360.196.7710.383.121.140.70100.0Parameters for WARS magma chamber modelSymbolDefinitionValueUnits cm Specific heat capacity of melt 1,142 Jkg−1K−1 cx Specific heat capacity of crystals 1,160 Jkg−1K−1 cCO2 Specific heat capacity of CO2 fluid 1,200 Jkg−1K−1 cH2O Specific heat capacity of H2O fluid 3,880 Jkg−1K−1 MH2O Molar mass of water 18.02×10−3 kgmol−1 MCO2 Molar mass of CO2 44.01×10−3 kgmol−1 Le Latent heat of exsolution 610×103 Jkg−1 Lx Latent heat of crystallization 470×103 Jkg−1 ∆Pc Critical overpressure 20×106 Pa αr Crustal thermal expansion coefficient 10−5 K−1 αm Melt thermal expansion coefficient 10−5 K−1 αx Crystal thermal expansion coefficient 10−5 K−1 βm Bulk modulus of melt 1.2×1010 Pa βx Bulk modulus of crystal 1010 Pa βr Bulk modulus of crust 1010 Pa κ Thermal diffusivity of crust 10−6 m2s−1 ρm Melt density 2,420 kgm−3 ρx Crystal density 2,900 kgm−3 ρr Density of crust 2,500 kgm−3 Parameter space for rhyolite‐MELTS simulationsPressure (MPa)100–400 (intervals of 50 MPa)wt. % H2O0.25–6 (intervals of 0.25 wt. %)wt. % CO2 0–1.1 (intervals of 0.1 wt. %) Note. Anhydrous oxide abundances in wt. %, the parameter space covered by the rhyolite‐MELTS v.1.1.0 simulations, and the derived parameters for the WARS magma chamber model. Anhydrous Composition and Model Parameters for WARS Basalt Note. Anhydrous oxide abundances in wt. %, the parameter space covered by the rhyolite‐MELTS v.1.1.0 simulations, and the derived parameters for the WARS magma chamber model. We derive estimates for the range of initial H2O and CO2 contents from volatile compositions presented in Oppenheimer et al. (2011), Moussallam et al. (2014), and Lowenstern (2001). Combining these with our anhydrous WARS basalt composition, we simulate thermodynamic closed‐system equilibrium crystallization using rhyolite‐MELTS (Gualda et al., 2012). While rhyolite‐MELTS v.1.1.0 is optimized for silicic systems, it is the only version of MELTS that accounts for both water and carbon dioxide at crustal pressures (Ghiorso & Gualda, 2015). After conducting multiple isobaric down‐temperature crystallization runs at various pressures, we fit the outputs of the MELTS simulations via quadratic regression to develop mathematical expressions for (a) crystallinity and (b) CO2 and H2O solubilities as functions of temperature, pressure, and bulk volatile contents. We use a constant latent heat of melting computed as the median latent heat of melting over the entire temperature range from each MELTS simulation, averaged over all MELTS runs. Constant values for density, specific heat, and compressibility of the melt are computed as the median value of each respective parameter across the super‐liquidus portion of each individual MELTS simulation, averaged over all crystallization calculations. Similarly, for the crystal density and specific heat, we apply the same approach, instead using the portion of the MELTS simulation over which the magma is greater than 50% crystals by volume. Melting Curves We use 2184 isobaric crystallization simulations on rhyolite‐MELTS v.1.1.0 with the average WARS basalt composition calculated from normalized anhydrous whole rock compositions over a range of fixed pressures and different initial CO2 and H2O conditions, as described in Table 1 (All MELTS calculations conducted at the oxygen fugacity of the NNO buffer.) From the raw MELTS outputs, we calculate volume fractions for melt, crystals/solids, and magmatic volatile phase (MVP). Employing the Matlab curve fitting application (cftool), we determine the most suitable function to describe the crystal volume fraction in the magma with respect to temperature. Huber et al. (2009) postulated that the melt fraction (f) of magmas can be parametrized as a power law function with an exponent (b) describing the effect of magma composition: (1) f(T)=T−TsolTliq−Tsolb,0 1). Within this range of magma chamber conditions, ice unloading raises the total erupted mass even at the most conservative unloading rate (0.03m/yr). Deeper in the crust, the viscous relaxation of accumulated stresses inside the magma chamber by the surrounding crust dominates magma chamber evolution, reducing the ability for the magma chamber overpressure to exceed the eruption threshold. At shallower depths (larger values of θ2), where magma chambers cool more rapidly and are generally more short‐lived, there is a clear trend of increasing cumulative mass erupted with an increased unloading rate. In addition, the unloading effects can outcompete those of higher rates of magmatic injection, extending the region of sensitivity to increases in erupted mass. Exceedingly rapid magmatic recharge will overprint the effects of unloading unless unloading occurs at a critical rate (i.e., the unloading timescale is less than or on the order of the magmatic injection timescale, θ3≤1.) As already described in Section 3.1, the unloading‐related increase in the erupted mass illustrated in Figure 6 is in part driven by additional unloading‐triggered eruption events. These eruptions are driven by two underlying mechanisms: first, overpressure is generated through the decompression‐induced volumetric expansion of the magma. Second, unloading alters the pressure budget required for the magma chamber to erupt, and as the overburden pressure decreases, a lower absolute pressure in the magma chamber is required to exceed the overpressure criteria for eruptions (Equation 8; Figure 4). These combined effects result in eruptions that would not occur in the absence of unloading, leading to a greater cumulative mass erupted over the lifetime of the magma chamber. As demonstrated in Figure 3 and Figure S5 in Supporting Information S1, these additional unloading‐triggered eruptions are observed from magma chambers that are volatile‐undersaturated and oversaturated alike. For magma chambers initially below the threshold of volatile saturation, these eruptions tend to occur while the volatiles are completely dissolved in the melt, since the magma is less compressible without exsolved fluid bubbles (see Figure S9 in Supporting Information S1). An additional source of the observed increase in cumulative erupted mass with unloading rate (Figure 6) is the quantity of larger‐sized eruptions, resulting from increased magma compressibility, as decompression promotes the release of volatiles from the melt. In Section 3.2, we show that with sufficient unloading rates, magma chambers initially below saturation conditions begin to exsolve volatiles earlier due to decompression. A regime diagram demonstrating the extent of sensitivity of initially volatile‐undersaturated magma chambers to an expedited transition to volatile saturation is provided in Figure S10 of the Supporting Information S1. Interestingly, the magma chambers that are most sensitive to earlier volatile saturation with unloading are cases that fail to erupt due to insufficient magmatic recharge (compare Figure 6 and Figure S10 in Supporting Information S1); these magma chambers are just short of the requirements to erupt but with slight perturbations in magmatic recharge could be pushed beyond the threshold. Magma chambers subjected to unloading will experience volatile saturation earlier in their evolution, resulting in the potential for more eruptions while the magma is volatile‐saturated (and hence more large eruption events). Nevertheless, there is still modest sensitivity to an earlier onset of volatile saturation with sufficient rates of unloading for magma chambers that do erupt, resulting in a greater cumulative mass erupted even without a change in the total number of eruptions (Figure 7). Regardless of whether the chamber is initially volatile‐saturated or becomes saturated later, with ice unloading (even if only active for a short duration with respect to the eruptible lifetime of the magma chamber), the volatile solubility is permanently reduced, resulting in increased magma compressibility and larger eruptions. Such long‐term effects, where even for short durations of ice removal (∼300 years), the removal of 1 km of ice results in an increased cumulative erupted mass long after the ice load is completely gone, are the most consequential in terms of the broader climate implications. Essentially, even if modern anthropogenic warming were curtailed immediately, the unloading that WARS subglacial volcanoes already experienced will still affect their behavior for hundreds to thousands of years to follow. Eruptions Before and After Volatile Exsolution. Revisiting the magma chamber evolution from Figure 5, where the onset of volatile exsolution is expedited by 150 years the highest unloading rate for the removal of 1‐km ice (3 m/yr). At this rate, the ice load is completely removed by 300 years (indicated by yellow star). Eruptions that occur while H2O and CO2 are still dissolved in the melt are notably smaller than those post volatile exsolution because of the profound increase in magma compressibility with subtle increases in the exsolved fluid volume fraction. When the magma chamber is subjected to the fast‐unloading scenario, it erupts two additional times while the chamber is volatile‐saturated than the equivalent magma chamber without unloading. Given these unloading effects on the total mass erupted, we observe a stronger sensitivity of magma chambers that are initially saturated (i.e., Figures 6b–6d) subject to unloading compared to that of magma chambers that are initially volatile‐undersaturated (compare with Figure S7 in Supporting Information S1). While the magma chamber is volatile‐saturated and thus more compressible, a larger amount of mass must be evacuated in a single eruption to restore the magma chamber pressure to lithostatic (consider the effect of these larger eruptions compounded over the entirety of the eruptible lifetime of an initially volatile‐saturated magma chamber, as compared to that for a magma chamber that only becomes saturated later in its evolution). Addition to Enthalpy Budget An increase in mass erupted from a subglacial magma chamber means more heat is introduced near the ice sheet. The enthalpy of an eruption is calculated as the sum of the sensible heat from the erupted material, Qsensible, and the latent heat of crystallization and volatile exsolution, QLatent,x and QLatent,e respectively: (14) Herupt=Qsensible+QLatent,x+QLatent,e, where (15) Qsensible=ρcTerVerupt, (16) QLatent,x=ρxεxVeruptLx, and (17) QLatent,e=ρgεgVeruptLe. Lx and Le are provided in Table 1. Here, Ter, εx, and εg are the temperature, volume fraction of crystals, and exsolved fluid volume fraction, respectively, inside the magma chamber at the midpoint of the eruption. Figure S8 in Supporting Information S1 presents a regime diagram showing how the total erupted enthalpy from magma chamber deviates from the zero‐unloading scenario at various rates of unloading. Unloading‐triggered volcanism will transfer additional enthalpy directly to the ice sheet base and/or surface once the erupted material penetrates through the ice. We can consider a theoretical average additional heat supply to the ice due to the volcanic response to unloading over time as follows: (18) q=Heruptwithunloading−Heruptwithoutunloadingeruptiblelifetimeofmagmachamber where the eruptible lifetime of the magma chamber is the duration of time over which the magma remains below εx=0.5. Figure 8 demonstrates this additional heat input to the ice sheet for many magma chambers with initial volatile contents of 1 wt. % H2O and 5,000 ppm CO2 (volatile‐saturated throughout their entire evolution). While additional heat released over time due to unloading could result from either a larger erupted mass from the magma chamber without changes in its eruptible lifetime, or from the same (or less) erupted mass over a condensed eruptible lifetime, we find that unloading causes negligible changes in the eruptible lifetime of magma chambers that are initially volatile‐saturated (see Figure S11 in Supporting Information S1). Hence, the sensitivity of magma chambers to this additional source of heat released over time in Figure 8 is primarily due to the increase in the total mass erupted with unloading. The highest rates of additional heat released due to unloading in Figure 8 are exhibited from magma chamber simulations subject to the intermediate unloading rate (0.3 m/yr). At even higher unloading rates, the 1‐km ice load is removed rapidly with respect to the eruptible duration of the magma chamber simulation; once unloading “shuts off”, the volumetric expansion‐driven overpressure is gone. This means that only the unloading effects of greater magma compressibility and reduced pressure threshold for eruptions can increase the cumulative mass erupted in the longer term. At more moderate unloading rates, magma chambers subjected to the removal of a finite ice will experience the overpressure effect from volumetric expansion under decompression for a larger fraction of their eruptible lifetimes, which may compound the total heat released through eruptions over time. It is important to note that while modern ice loss rates may exceed those tested in this paper, modern subglacial magma chambers have also been experiencing a longer history of ice unloading, since the last glacial maximum (similar to the 30,000‐year timescale for the removal of 1‐km ice) and prior glacial cycles. Regime diagram for magma chamber simulations at depths 6, 8, and 10 km in the crust, indicating the additional heat released in Terajoules per year due to unloading at various rates. All magma chambers have initial volatile contents of 1 wt. % H2O and 5,000 ppm CO2. Black circles represent magma chambers that do not exist in either case (with or without ice unloading). The magma chamber circled in cyan and indicated with an arrow is discussed further in the text. We consider the additional heat supplied from unloading‐triggered volcanic effects on the magma chamber circled in cyan in Figure 8, which releases an additional 10 Terajoules per year at the intermediate unloading rate (0.3m/yr). We calculate the amount of ice that can be melted by an additional 10 Terajoules (1×1012J) per year to the ice sheet; the heat required to melt a mass, m, of ice is given by: (19) Q=m·ΔHf Taking the heat of fusion of water, ΔH f , to be 334 J/g, the heat from unloading‐triggered volcanism is capable of melting ∼3×109 kg of ice per year, or equivalently ∼3×106 m3 of ice per year. This estimate stands for a single subglacial magma chamber at a depth of 8 km within the crust, with an initial volume of 1.5 km3, subjected to the removal of 1‐km ice at the intermediate unloading rate (0.3 m/yr). Supporting Evidence From the Geologic Record and Consequences for West Antarctica While there is a timely need to understand the consequences of glacio‐volcanic interactions for the assessment of risks both to local communities and globally as sea‐level rises, it is difficult to probe recent subglacial volcanic activity (Iverson et al., 2017). Perhaps the most complete coupled records of volcanism and deglaciation are from the Southern Volcanic Zone of the Andes in South America, where the Patagonian ice sheet grew to its greatest thickness of 1,600 m or more atop several volcanic composite systems between about 35 and 18 ka, and retreated very rapidly between 18 and 15 ka (Mixon et al., 2021; Moreno‐Yaeger et al., 2024; Rawson et al., 2016; Singer et al., 2024; Watt et al., 2013). Figure 9 illustrates the increase in both cumulative volume erupted and the number of eruptions observed in the records from Calbuco, Mocho‐Choshuenco, and Puyehue‐Cordon Caulle volcanoes in the Andean arc following the local Last Glacial Maximum at 18 ka. These reconstructions that span the rapid, <3 Kyr, transition from glacial to post‐glacial conditions suggest that there is a link between the rapid unloading of ice from the landscape and upticks in the frequency and volume of eruptions. As we have shown, magma chambers release more mass and heat with sufficiently high rates of ice unloading. Given that erupted volume is proportional to mass erupted, the records from these Andean volcanoes may be an illustration of the unloading‐triggered behavior outlined in this study. Since the Andean records reflect arc magmas, which typically have higher initial volatile contents than rift magmas, they were likely volatile‐saturated during the deglaciation. Our results suggest that volatile‐saturated magma chambers are more sensitive to unloading‐induced perturbations in erupted mass than magma chambers with a delayed onset of or without volatile exsolution, and hence the response of Andean volcanoes could be magnified in comparison to volcanoes in rift settings like West Antarctica. A positive feedback between unloading and magma chamber evolution could have accelerated deglaciation in the Andes from the heat released via eruptions and a decrease in ice albedo from subaerial volcanic deposits, potentially amplifying the volcanic response, in turn. While thick alpine glaciers remain atop some Andean volcanoes, they are typically narrow features (<1 km wide). Therefore, it is unlikely that the unloading of alpine ice can generate the requisite stress changes for eruptions at magma chamber depth (Singer et al., 2024). However, ice unloading in West Antarctica occurs over length scales more influential at depth, and thus increased volcanic activity may warrant consideration in the hazard assessment of future Antarctic expeditions. Records of cumulative volume erupted over time from three composite arc front volcanoes in the Andean Southern Volcanic Zone (Singer et al., 2024). 40Ar/39Ar ages from effusive lava flows and 14C ages for tephra deposits have been used together with field relations to estimate eruptive volumes and growth rates (original data from Singer et al. (2008), Mixon et al. (2021), and Moreno‐Yaeger et al. (2024)). Comparison with the global marine proxy record for global ice volume (Lisiecki & Raymo, 2005) indicates that following both Termination II (TII) and the local Last Glacial Maximum (LGM), cone growth rates markedly increase. Efforts that couple subglacial magmatic and/or volcanic evolution self‐consistently with a variable pace of deglaciation (as in Singer et al. (2024)) can help provide evidence for the existence of an unstable feedback loop and inform the risk of runaway ice retreat in West Antarctica. Recalling from 4.2, the additional heat supplied from ice unloading effects on a particular magma chamber was ∼3 ×109kg/yr; considering the more than 100 potentially active subglacial volcanoes of West Antarctica, the aggregate potential contribution to ice mass loss still pales in comparison, for example, to the estimated ∼168 ± 78 Gt per year lost through the Amundsen Sea Embayment since 2010 (Davison et al., 2023) when considered in isolation. Even so, melted ice along the base of the ice sheet could facilitate sliding and ice loss to the ocean, and given the unloading effects on volcanism unaccounted for in our model at present (e.g., enhanced melt supply, stress conditions that promote dike initiation), this contribution to ice melting may be magnified. Nonetheless, this additional source of heat from the decreasing overburden pressure considered in this model contributes to uncertainty in the mass balance of the ice sheet and may play a more important role in the ice sheet mass balance if there is in fact a positive feedback loop between deglaciation and volcanism. Given the precarious state of the West Antarctic Ice Sheet, this demonstration of additional heat input warrants consideration of glacio‐volcanic feedbacks in projections of ice loss from West Antarctica. Conclusions We model the evolution of a subglacial magma chamber under various ice unloading rates to understand the physical mechanisms linking deglaciation and volcanism, with special consideration of the West Antarctic Rift System. To develop physical intuition for these processes, we consider ice unloading for a wide range of potential magma chambers. We find that a critical rate of unloading increases the total erupted mass from WARS‐type (basaltic) subglacial magma chambers when simulating the removal of realistic ice loads (e.g., thickness on the km‐scale). When the rate of ice unloading can compete with the rates of other processes affecting the internal dynamics of the magma chamber (i.e., magmatic recharge, cooling, and viscous relaxation of the surrounding crust), magma chambers produce additional eruptions due to the volumetric expansion of compressible magma inside the chamber, as well as a lowered critical pressure threshold required to trigger eruptions. In some cases, a longer duration of unloading at a more modest rate leads to more of these unloading‐triggered eruptions than for short periods of extremely rapid unloading. Additionally, as the rate of unloading increases over magma chambers on the cusp of volatile saturation, the onset of the first exsolution of volatiles is expedited. Even after volatile saturation is reached, continued unloading increases the volume fraction of bubbles and hence, the compressibility of the magma. With increased magma compressibility, a larger amount of mass must be withdrawn during an eruption to relax the overpressure (Huppert & Woods, 2002; Townsend et al., 2019). Consequently, the total mass erupted from a magma chamber due to unloading is highly sensitive to the volume fraction of exsolved volatiles. Even long after ice unloading ceases, the compressibility of the magma remains permanently elevated due to the reduction in lithostatic pressure, resulting in larger eruptions compounding over the long‐term trajectory of the magma chamber. The additional heat associated with such unloading‐triggered eruptions is currently unaccounted for in models of the West Antarctic Ice Sheet, despite its potential to perturb the surface mass balance and/or basal sliding rate of an already critically vulnerable ice sheet. In the interest of elucidating hazards both in terms of fieldwork in West Antarctica and global sea‐level rise, more work must be done to determine whether other glacio‐volcanic feedbacks will amplify modern ice loss. Conflict of Interest The authors declare no conflicts of interest relevant to this study. Supporting information Supporting Information S1