LAHAR FORMATION BY CATASTROPHIC COLLAPSE OF A PYROCLASTIC DAM: HISTORY, VOLUME, AND DURATION OF THE 2360 BP SALAL LAKE, MOUNT MEAGER, BRITISH COLUMBIA, CANADA

Mount Meager is situated at the northern end of the Canadian portion of the Cascade Arc (Garibaldi Belt), and produced the youngest explosive eruption in Canada. The 2360 BP eruption occurred in an area of high relief and produced airfall, pyroclastic density currents, and block and ash flow deposits that inundated and partly-filled the Lillooet River valley immediately beneath the vent. A direct consequence of the 2360 BP eruption of Mount Meager was that the Lillooet River, which drains the Lillooet Icefield 20 km upstream of Mount Meager, was dammed at least twice by the rapid accumulation of block and ash flow deposits.

The first damming-event was caused by formation of a > 100 m high dam of welded to non-welded block and ash flow deposit. Uniquely, we demonstrate that the filling of a temporary lake (Salal Lake), and its subsequent draining, occurred during welding of the pyroclastic dam, such that the upper non- to incipiently-welded part of the dam was removed. During a hiatus in volcanism, outflow from the paleo-lake (estimated volume 0.2 km³) formed a river that used the top of the welded pyroclastic deposits (e.g., dam) as a natural spillway without significant erosion.

Subsequently volcanism waxed again, producing a second, largely non-welded block and ash flow dam on top of the remnants of the first dam (> 150 m total height). An additional 0.35 km³ of water was added to the lake behind the heightened dam, together with minor deltas where tributaries entered the lake. This time, however, the entire dam failed catastrophically removing most of the second dam and excavating a 2 km long canyon in the still hot, welded deposit beneath. This catastrophic dam failure and deluge completely drained Salal Lake and formed a lahar that travelled at least 50 km downstream, inundating the present site of the village of Pemberton.

We have modeled the duration of the second dam-building and failure event using field observations that constrain the size of the impounded lake and the volumes of water and sediment transported by lahars. Our calculations are also used to estimate the amount of compaction during welding of the pyroclastic dam, the maximum time available for welding of the first dam (e.g., prior to failure), and the effective viscosity of the pyroclastic deposits during welding.