Arc Heat Flow and Magmatic Heat Budgets

Abstract We evaluate hydrothermal heat loss from 11 volcanic‐arc segments (∼6,000 km of arc length, ∼10% of the global total), motivated by the observation that much magmatic heat ultimately crosses the land surface as heated aqueous fluid. Heat loss takes place by volcanic eruption, geothermal heat conduction to the surface, fumarolic (vapor) discharge, thermal springs, and discharge of groundwater that has been heated by only a few degrees. Heat loss by extrusion of volcanic products ranges from 0.1 to ∼4 MW/km. For some arc segments, the hydrothermal heat loss is dominated by “slightly thermal” springs (maximum ∼10 MW/km arc length) or thermal springs (maximum ∼21 MW/km), but more commonly by hydrothermal steam vents and volcanic fumaroles (maximum ∼7 MW/km). Total hydrothermal heat‐loss rates range from <2.5 MW/km (Southwest Japan, Cascade Range, Northeast Japan, Kurils) to ≥8 MW/km (Ryukyu, Apennines, Taupo Volcanic Zone). We use these hydrothermal heat losses to estimate rates of magma supply, and combine these estimates of magma supply with existing estimates of eruption rates to obtain intrusion:extrusion ratios. Potential causal influences include state‐of‐stress and the abundance of silicic magma in the midcrust. Likely causes of along‐arc variations range from the near‐surface (0–5 km) hydraulic architecture (Cascade Range) to the nature of the subducting plate (Ryukyu vs. the rest of southwest Japan). Inferred intrusion: extrusion ratios are generally between 0.5 and 9. Whole‐arc comparisons between heat‐flow‐based intrusion rates and those based on volatile fluxes and petrologic models are complicated by the wide range of along‐arc behavior and the fact that we sometimes rely on volatile fluxes (e.g., SO 2 ) to help calculate hydrothermal heat losses, so that the data sets are not fully independent. However, reasonable agreement can be demonstrated in some examples of arc subsections.