Predictions of volcanic eruptions have undergone a transition from empirical models to models based on magmatic dynamic properties, and the establishment of these prediction models is strongly dependent on the physicochemical conditions of magma transport and storage, including the composition, viscosity and temperature of the magma.
Identifying the storage environment of magma before eruptions is critical to modeling eruption predictions. Geophysical data show that there are significant differences in the depth of magma storage in different volcanoes, but as of now it is not clear what factors control this difference. A popular view is that when the density of rising magma and the density of the surrounding rock are equal, the magma can be stably stored at the corresponding depth, that is, the "natural buoyancy" depth. But this theory has only been tested in volcanoes in the mid-ocean ridge (Hooft and Detrick, 1993), and whether it can also be applied to arc magma is debatable. At the same time, the existing research has shown that the storage depth of magma is affected by the regional stress state and/or magmatic channel structure in the crust, and the magma is stored on the boundary of whether the crust is rheological, which is not only constrained by the nature of the crust, but also controlled by the strain rate. Although crustal rheology is an important factor for long-term stable magma reservoirs, it is still a question worth pondering whether different arc magma storage depths are the only influencing factors on a global scale. In addition to external control factors, some studies have proposed that the water content in magma is used as an intrinsic factor to control the storage depth of magma. In the process of magma rise, volatile emissions and crystal crystal crystals in the magma will increase the viscosity of the melt to a certain extent, resulting in the magma's rise being inhibited or forced to stop, this theory has been confirmed in the mid-ocean ridge volcano and some island arc volcanoes, but whether it is suitable for global arc magma needs to be further verified.
In response to the above questions, Dr. Daniel J. Rasmussen of the National Museum of Natural History and his collaborators have counted the water content data of melt inclusions in a total of 62 medium-base island arc magmas published in the past 20 years, and combined with geophysical observations, revealed and verified that the water content of arc magma controlled the storage depth before magma eruption, and the relevant results were recently published in Science.
Rasmussen et al. selected the water content data in 3856 melt inclusions, converted them into the saturation depth of water, and compared them with the current geophysical results, and found that the water saturation depth obtained by the petrological method and the magma storage depth obtained by the geophysical method have good consistency within the error range (Figure 1). Model simulations suggest that this consistency may be caused by changes in viscosity as the magma rises.
Figure 1 Arc magma storage depth (geophysical estimates) vs. Saturation depth of water (melt inclusion estimate). Natural buoyancy depth represents the depth at which the surrounding rock density and magma density are equal (Rasmussen et al., 2022)
The water content at the depth of magma storage may represent the water content in the magma at the time of generation in the source area, i.e. the magma produced has not undergone significant outgassing (mantle control mode) before reaching the storage location through the crust; it may also be the result of some degree of degassing of the magma produced in the source area before reaching the storage location (Crustal control mode) (Plank et al., 2013). Distinguishing between these two patterns is important for understanding the effect of water content in arc magma on the storage depth before magma eruptions. To distinguish between the two modes, the ratio of non-volatile trace elements in the melt inclusion is used to determine the relationship with the water content in the melt inclusion. The results showed that the H2O/Ce in the melt inclusions had a good correlation with Ba/La and Nb/Ce, respectively (Figure 3), which showed that the melt inclusions did not experience significant degassing after formation, that is, the mantle control mode. On this basis, the authors simulated the ascent process of magma using rhyolite-MELTS, and found that when the magma reached the depth of water saturation, the rapid increase in viscosity caused by volatilization and crystallization hindered the further rise of the magma, and this depth was well consistent with the storage position of the magma within the margin of error (Figure 4), which showed that the viscosity during the magma ascent controlled the storage depth of the magma.
Fig. 2 Schematic diagram of crustal control and mantle control mode of deep water content in magma storage (Adapted from Plank et al., 2013)
Fig. 3 H2O/Ce and non-volatile trace elements Ba/La and Nb/Ce of melt inclusions in the arc magma of the central and eastern islands of Aleutian (Rasmussen et al., 2022)
Fig. 4 Relationship between magma storage depth and viscosity increase during magma ascent, with the "knee" point being the point with the fastest increase in viscosity (Rasmussen et al., 2022)
This finding suggests that island arc magma has buoyancy at its storage location and experiences a "viscosity stagnation" phase before eruption, while magma with more water content produces outgassing and crystallization at a deeper location, resulting in greater magma storage depths. These results shed light on the storage environment of magma, will help to build more accurate predictive models of volcanic eruptions, and also lead us to further ponder whether the above laws also apply to other tectonic backgrounds, such as continental subduction zones, rift valleys, etc. Exploring the pre-eruption magmatic storage environment in a variety of geological contexts will help build more comprehensive models of eruptive predictions.
Key references
Hooft E E, Detrick R S. The role of density in the accumulation of basaltic melts at mid-ocean ridges [J]. Geophysical Research Letters, 1993, 20(6): 423-426.
Plank T, Kelley K A, Zimmer M M, et al. Why do mafic arc magmas contain ～ 4 wt% water on average? [J]. Earth and Planetary Science Letters, 2013, 364: 168-179.
Editor: Chen Feifei
Proofreader: Liu Qi county