Fault zone processes during caldera collapse: Jangsan Caldera, Korea

Abstract

Caldera fault zones, identified in many modern and ancient volcanoes, have been the subject of geological and geophysical observations and of analog and numerical modeling. However, the physicochemical processes in fault zones during a caldera collapse are still poorly understood. Here, we present field observations from a caldera fault zone in the Cretaceous Jangsan Caldera, SE Korea. The fault zone is similar to 30 m wide and juxtaposes an intracaldera rhyolitic volcanic complex against older dacitic rocks; it consists of a minor fault, a main fault, and a series of fault-related intrusions (rhyolite and tuffisite). The main fault dips 90 degrees-88 degrees NW (vertically to steeply inward) and strikes N30 degrees-40 degrees E. A layer of pseudotachylyte (less than similar to 10 cm thick) with some injection veins occurs along the main fault. The pseudotachylyte contains rounded to sub-rounded clasts, commonly displaying fuzzy or embayed boundaries, as well as euhedral feldspar microlites ( < 10 mu m) and flow structures similar to those typically observed in tectonic pseudotachylytes. The tuffisite not only intrudes sharply into the pseudotachylyte and dacitic wall rocks but also displays a mingling structure with the pseudotachylyte. The contact between the rhyolitic intrusion and the tuffisite is nonplanar and complex. Based on the observed field relationships, we conclude that frictional melting and magmatic intrusion were near-coeval and that intrusion outlasted frictional melting by a short period. Although volcano-tectonic pseudotachylytes have rarely been identified in calderas (e.g., in the Glencoe Caldera, Scotland), our results suggest that frictional melting may be an important fault zone process during caldera collapse.