Centimeter-High Antidunes within Pyroclastic Deposits: Are They Products of Surge or Density Current?

Distinguishing pyroclastic surge vs. density-current deposits in the field is challenging, in particular when 10’s - 100’s m thick pyroclastic outcrops with “even” distributions of sedimentary structures (e.g., dunes/antidunes of similar size) are examined. The criteria commonly used to distinguish the two include the size and type of pyroclasts, and the height of individual sedimentary structures in a 1-D vertical section. However, we argue that these criteria ignore large-scale (m-10’m high) geometry of bounding surfaces and lateral (proximal-distal) facies change within the outcrops, and are of limited use. Instead, correct identification of surge vs. density-current deposits must be based on large (10's m high, 100's m wide) dip-oriented cross-section, similar to the ones used for architectural element analysis methodology for siliciclastic deposits. To illustrate this point, we compare two Holocene pyroclastic outcrops from Niijima and Izu-Oshima islands, 100 and 170 km SSW of Tokyo respectively. The Niijima outcrops were products of AD 886 eruptions, and form 10's-100 m high coastal cliffs made of pumice of coarse-sand and gravel size, with abundant antidunes. These cliffs were formerly divided into dm-to-m-thick vertically stacked antidune beds, and each bed was attributed to single surge event. As opposed to this layer-cake model, we have revealed that tens of these beds are bundled to form parallel wavy geometry: each wavelet measuring 1-3 m high and 50-150 m wide. These wavelets form a 10-20 m thick vertically aggradational unit, which is bounded at its top and base by regional unconformities. The entire coastal cliffs are divided into several unconformity-bounded units with similar internal wavy geometry, and we interpret each unit as sediment-wave (giant antidune) deposits formed by a single pyroclastic density-current event. The early to middle Holocene outcrops of Izu-Oshima was traditionally interpreted as exemplinary subaerial ash-fall deposits, even though giant wavy structures similar to the ones in the Niijima outcrops have long been recognized. Here, we have found abundant antidunes and marine trace fossils, and interpret them as formed by subaqueous pyroclastic density currents. In both outcrops, antidunes have three- to four-fold compound occurrence (ranks 1 to 4), where small antidunes are nesting within larger antidunes. In both outcrops, the largest (rank 1) antidunes correspond to sediment waves (m-10 m high) bounded at its top and base by unconformities. Whereas rank 4 (cm thick) and rank 3 (dm thick) antidunes are the most common size of the antidunes in Izu-Oshima and Niijima outcrops respectively, they were formerly interpreted as local "surge" deposits. However, a line of evidence suggests that they are of density-current origin; (1) nested architectgure of the antidunes, (2) these antidunes fill spaces between abundant lag deposits made of dm-m long lava fragments above the basal unconformities, (3) some of the bed boundaries downlap onto basal unconformities, (4) bed boundaries have high angles, occasionally exceeding angle of repose, and (5) abundant subaqueous indicators such as marine trace fossils and water-escape structures.