Magma ascent velocities, v (dH/dt; where H is depth and t is time), can be determined from decompression rates (dP/dt), and rates of cooling (dT/dt): , where ρ is magma density, P is pressure, T is temperature and g is the acceleration due to gravity. This equation for v provides a key to investigating the relationships between the initial ascent velocity of magmas and the depths of magma dehydration. Ascent velocities can be calculated using pressure and temperature (P–T) estimates from mineral–liquid thermobarometry and cooling rates inferred from crystal size distribution (CSD) theory. For recent Mt. Etna lava flows, both dP/dT and dT/dt have been characterized for the portion of the feeding system between the Moho (∼27 km) and 6 km based, respectively, on clinopyroxene thermobarometry and clinopyroxene CSDs. Deep-level (>6 km) magma ascent velocities range from practically zero (where clinopyroxene P–T estimates form a cluster, and so dP/dT ≈ 0), to about 10 m h^–1 for flows that yield very steep P–T trajectories. Many lava flows at Mt. Etna yield P–T paths that follow a hydrous (∼3% water) clinopyroxene saturation surface, which closely approximates the water content inferred from melt inclusions. Independent assessments of deep-level water contents have been obtained by means of a new geohygrometer and yield ascent rates of ∼1 m h^–1, in agreement with the slowest rates derived for magma effusion or vapor-driven ascent (∼0·001 to >0·2 m s^–1, or 3·6–720 m h^–1). Changes in P–T slope, as determined by pyroxene thermobarometry, indicate an upward acceleration of magma, which may be due to the onset of deep-level magma dehydration linked to the non-ideal behavior of water and CO₂ mixtures that induce a deep-level maximum of water loss at P ≈ 0·4 MPa and T ≈ 1200°C for a CO₂ content >1000 ppm.

We have refined the clinopyroxene-based hygrometer published by Armienti et al. (2013) for a better quantitative understanding of the role of H₂O in the differentiation of Etnean magmas. The original calibration data set has been significantly improved by including several experimental clinopyroxene compositions that closely reproduce those found in natural Etnean products. To verify the accuracy of the model, some randomly selected experimental clinopyroxene compositions external to the calibration data set have been used as test data. Through a statistic algorithm based on the Mallows’ C_P criterion, we also check that all model parameters do not cause data overfitting, or systematic error. The application of the refined hygrometer to the Mt. Etna 2011–2013 lava fountains indicates that most of the decreases in H₂O content occur at P < 100 MPa, in agreement with melt inclusion data suggesting abundant H₂O degassing at shallow crustal levels during magma ascent in the conduit and eruption to the surface.

Among the abundant mantle xenoliths carried by the Cenozoic alkaline basalts of northern Victoria Land, Antarctica, we have studied a suite of clinopyroxene-rich cumulates collected at Browning Pass (Mt. Melbourne Volcanic Province), ranging in composition from wehrlites to clinopyroxenites. Clinopyroxenes belonging to both the Cr-diopside (wehrlites) and Al-augite series (ol-clinopyroxenites and clinopyroxenites) all show convex-upward REE patterns. Modal and cryptic metasomatism has variably affected the xenoliths, accounting for amphibole replacement of clinopyroxene and/or selective enrichment in incompatible elements. Chemical features, along with O–Sr–Nd isotopic data, indicate that both the parental magmas and the metasomatizing melts are related to the Cenozoic magmatic activity and imply the role of at least two mantle components with distinct isotopic fingerprints. The positive covariation between δ¹⁸O_olivine and the amount of modal olivine, and between δ¹⁸O_olivine and olivine Fo content, suggest that during the fractionation of olivine and pyroxene, the parent magma experienced a change in O-isotope composition; a low-δ¹⁸O melt component was not only added to the minerals during the metasomatic event but was also involved in the genesis of the parental melts. The Browning Pass cumulates are used to constrain the origin of the Antarctic Cenozoic magmatism from a heterogeneous mantle source whose depleted end-member is inferred to be the local lithospheric mantle, whereas the enriched end-member is represented by early metasomatic veins or domains emplaced into the depleted mantle during an amagmatic phase of rifting at the beginning of Ross Sea opening. Thermobarometric analysis of the process shows that the respective contribution to magma generation of the two end-members is related to the change of local thermal regime induced by an ‘edge effect’ in the mantle circulation at the lithospheric step between the thick East Antarctic craton and the thinned Ross Sea crust.