Activity was monitored during 1-9 September using detailed field observations combined with satellite and aerial remote sensing data. Activity was generally similar to that reported in August. On 6 September a large eruption began. In the preceding days activity had fluctuated. On 1 September, the only activity observed was a small white gas cloud at 0944. Gas clouds were emitted from 0748 until 0942 on 2 September. These predominantly white and gray clouds rose only 200 m above the crater before dissipating. The only exception was a period of ten minutes when brown and dark gray clouds issued from the crater. The sole emission the following day was a small white gas cloud at 1506. On 4 and 5 September small gas emissions were observed from the fumarole on the S side of the cone.

Activity on 6 September was first noted at 0702 when large white and gray gas clouds rose from the whole crater. At 0704 part of the gas column began to sink and move down the upper flanks, obscuring the E-flank ice walls. The gray and brown gas cloud was densest on the S side of the crater and appeared to be expanding as it rose. At 0711, the whiter part of the cloud rose upward while the dark gray portion dropped ash on the N side of the cone. Wind speeds at the summit appeared to increase, and the 400-m-high column began to be pushed N. At 0716 more gas descended the flanks. At 0735 observers on the edge of the easternmost lava flow could smell sulfur.

The main gas emission continued to be from the S side of the crater and at 0740 another cloud descended over halfway down the flanks. At 0743 a large white and dark gray gas cloud emerged from the crater. Ash fell from it onto the upper and mid-slopes. Another large gray, white, and brown plume filled the whole crater at 0746 and billowing to 400 m. At 0749 the plume color changed to brown, yellow, and dark gray. Ash was blown N. New gas clouds emerged from the crater on average every 30 seconds. At 0824 the cloud color returned to white and light gray for a few minutes before it once again became brown, gray, and yellow. The brown portion seemed to contain the ash. Gas once again descended the upper slopes at 0846. Winds at the summit began to pull the top of the plumes apart and by 0854 they were almost flat across the crater.

There was a reduction in gas emission at 1143. Gas continued to periodically descend the upper slopes and ashfall appeared to be mainly on the N slopes. At 1155 a gas cloud descended to mid-slope. The interval between gas emissions grew during the afternoon. After three hours of white- and gray-colored gas clouds, yellow, white, and brown clouds emerged again at 1604. This marked renewal of activity was similar to that in the early morning. Gas originated mainly from the southern fumarole and occasionally descended the upper slopes. Gas clouds rose 500 m and formed a cumulo-like mass. At 1737 there was a big gas release, part of which descended the cone slope while the main cloud rose and curled N over the crater. After this the intensity of the activity from the cone diminished and gas clouds became light gray.

On 7 September a faint brown haze was noted over Sabancaya at 0630. Dust in the atmosphere obscured viewing. Gas clouds were observed at 0643, 0704, 0719, and 1210. Visibility improved around mid-day, and ashfall was observed on the S side of the cone at 1243. At 1652 a small gas cloud descended the upper slopes. From 1740 until dark, gas emissions were continuous, but none were seen the following day. On 9 September observers on a morning flight around the volcano observed light emissions from fumaroles on the N and S crater rims. Fresh sulfur deposits existed on the crater walls. The crater itself was deeper than the year before and the floor could not be seen. Recent ash eruptions had covered the ice walls on the E side.

Geological Summary. Sabancaya, located in the saddle NE of Ampato and SE of Hualca Hualca volcanoes, is the youngest of these volcanic centers and the only one to have erupted in historical time. The oldest of the three, Nevado Hualca Hualca, is of probable late-Pliocene to early Pleistocene age. The name Sabancaya (meaning "tongue of fire" in the Quechua language) first appeared in records in 1595 CE, suggesting activity prior to that date. Holocene activity has consisted of Plinian eruptions followed by emission of voluminous andesitic and dacitic lava flows, which form an extensive apron around the volcano on all sides but the south. Records of observed eruptions date back to 1750 CE.

Information Contacts: Mark Bulmer, Frederick Engle, and Andrew Johnston, Center for Earth and Planetary Studies, National Air and Space Museum, Smithsonian Institution, Washington DC 20560-0315 USA; Guido Salas, Departamento Academico de Geoloia y Geofisica, Universidad Nacional de San Augustin, Arequipa, Perú; Elian Perea, Universidad Nacional de San Augustin, Arequipa, Perú.