Stratigraphy of the 1902 and 1929 nuée-ardente deposits, Mt. Pelée, Martinique

Abstract

Mild nuee ardentes in 1902–1903 and 1929–1930 at Mt. Pelee formed high-aspect ratio (H.A.R.) deposits (i.e. deposits with a high thickness/extent ratio) in the Riviere Blanche channel. Several violent nuees ardentes from May to August, 1902, yielded low aspect ratio (L.A.R.) deposits unevenly blanketing an area of 58 km2 on the southwestern flank of the volcano. Three closely spaced nuees ardentes on August 30 additionally affected 56 km2 on the southeastern flank. The L.A.R. deposits are divided into eight stratigraphic units (U1-U8), either compound or consisting of a single flow unit. The deposits of the largest nuees ardentes, on May 8, 20 and August 30, 1902 are identified as U1, U3 and U8, respectively, and can be traced over most of the area. Individual L.A.R. flow units are lenticular and show large, erratic lateral variations of thickness and grain-size. Most flow units consist of three distinct beds: Bed 1 is a few centimeters thick, lenticular, fines-depleted layer. Bed 2 is the thickest layer and contains more silt-sized ash than bed 1. It is normally graded as a whole, though reverse grading may occur in the lower part. Bed 2 generally consists of two parts: bed 2a, nonbedded, and bed 2b, finer-grained and laminated with cross-bedding and duned structures. Bed 3 is a thin, silty, capping layer which contains accretionary lapilli. The distribution of the L.A.R. flow units, especially of the coarse-grained facies, suggests that they were formed from overpressured blast-flows expanding radially from the vent at a high rate, in the same manner as the laterally directed blast of Mount St. Helens on May 18, 1980. The transport system of the major 1902 nuees ardentes was an inflated, relatively low-concentration, turbulent cloud as suggested by several lines of evidence: (1) the transported material was able to cross ridges up to 300 m high; (2) many scour-and-fill and other erosional features can be observed in the deposits at any distance from the vent; (3) survivors within the devastated area on May 8 did not experience a dense, low-profile cloud but instead a choking, hot ash-laden cloud; (4) many human corpses in St. Pierre remained where they fell instead of having floated away. If the current flow-surge terminology were to be used, the 1902 blast-flows would be better considered as pyroclastic surges. The normally graded, multi-layered structure of the flow units is also consistent with deposits primarily formed from relatively low-concentration flows.