LAUSR

We welcome comments on our report of Archean eclogitefacies oceanic crust and deep subduction. However, Zou et al. (1)’s claims concerning our geobarometry and petrography (2) are demonstrably misleading. The garnet-clinopyroxene (Grt-Cpx) geobarometer (3) was calibrated at 2 GPa to 7 GPa and 900 °C to 1,550 °C, and, with other experiments, is extrapolatable to eclogitic rocks below 700 °C (3), not a minimum of 1,200 °C as misreported by ref. 1, who incorrectly redraw figure 4 of Beyer et al. (3), omitting original calibration data at low temperature and low AlCpx(the tetrahedral aluminum content in Cpx) in the range of our samples (the complete original figure 4 of Beyer is reproduced as our Fig. 1A). Even if AlCpx is low (e.g., <0.05 per formular unit (p.f.u.)), the geobarometer yields accurate results (Fig. 1B). The rare earth element–based Grt-Cpx geothermobarometer (4) was developed for Mg-rich samples (e.g., Mg of Grt > 0.54), yielding reliable results on many lower Mg rocks (e.g., Mg of Grt < 0.40) (4–7), establishing reliability for our samples. Zou et al. (1) claim they tested the above geobarometers using “published metamorphic studies” (their figure 1C and D), but no information about their tested data or validity of other pressure estimates is provided. Zou et al. (1) claim that our exsolution textures indicate cooling from high temperature, “inconsistent with eclogitefacies metamorphism.” This is misleading. Numerous studies (2) show similar textures formed during decreasing pressure–temperature (P-T) conditions from eclogite and/ or granulite facies (8). Exsolution textures themselves are not indicative of eclogite-facies metamorphism, but are consistent with and support our calculations, and statement “implying decreased pressure and temperature due to exhumation from HP–UHP (high pressure–ultrahigh pressure) metamorphic conditions (2).” We considered but did not use pseudosections, because the peak mineral assemblage is too simple to provide useful constraints on P-T conditions. More importantly, the major minerals (Grt, Cpx) of Shangying garnet clinopyroxenite are compositionally zoned, which prominently affects the effective bulk-rock composition usable for calculating peak stage pseudosections. However, ref. 1 constructs a P-T pseudosection for the system NCKFMASHTO using our raw bulk composition, but with thermodynamic dataset, Fe value, water content, and model conditions following literature, unrelated and not applicable to our case. Even if their pseudosection