Paulaweh (Paloeweh, Rokatenda is the previously active dome) is a stratovolcano whose aerial portion forms the island of Palu’e north of Flores Island in the eastern part of Indonesia. The volcano rises about 3,000 m from the ocean floor. Its aerial portion is 875 m above sea level. The island is about 8 km in diameter.
Palu’e island is located 15 km off the north coast of neighboring Flores Island. It is conically shaped with an area of 72 km2. The north slope from the sea is gradual and steep on the south side. Deep ravines run down all sides of the volcano to the sea, forming steeply inclined ridges and channels for pyroclastic flows from collapsing summit domes. There is not a lot of the primary forest left, partly due to volcanic activity and partly due to forest clearing for agriculture. The island is part of the arid tropics. There are clearly separated dry and rainy seasons, with the rains from Dec – Mar. Average rainfall on the island is 180 cm.
Total population on the island is generally reported as 10,000, though Encyclopedia.com reports 14,000, 2,500 of which have been permanently evacuated to coastal Flores. About a third of the island is not suitable for settlement or agriculture. Inhabitants have their own distinct language, Sara Lu’a. Roughly 215,000 live within 50 km of the island, to the south on Flores Island.
Humans have lived on Flores Island for at least 100,000 years, with at least two waves of pygmy Homo floresiensis (aka Flores Man or Hobbit), living on the island until the arrival of modern humans some 50,000 years ago. Western explorers discovered Flores in 1511 – 1512. Locals had myths of their origin in a place far to the west. Palu’e’s lack of water was noted by western ships in the mid-19th Century. Trade, piracy and slavery were common in the region.
Local religion is nominally Catholic today, with over 90% identifying.
There are perhaps 40 villages on the island, varying in size from 30 – 500 people. They are connected to the administrative center by a network of footpaths. These villages were traditionally fortified and positioned on easily defensible ridges at higher locations due to a history of interdomain warfare, slave raids and piracy.
Agriculture mainly grows tubers, mung beans and maize with a number of minor crops. The series of 2012 – 2013 eruptions did significant damage to local crops. There is a ancestral prohibition against growing rice, likely due to land shortages. Domestic animals include goats, sheep, dogs, and poultry. Pigs are kept for ceremonial exchange and sacrifice. Offshore fishing has significantly reduced local fish populations offshore. Volcanic vents are used as outdoor ovens. Most men on the island are seasonal migrants, taking boats on the north side of Flores Island during the dry season.
There is no obvious source of fresh water on Palu’e Island. Villagers learned to tap steam at geothermal sites on the island, run it through a series of bamboo pipes to cool it sufficiently to collect water at the end of the distribution system.
The island does not have a tourist industry, though villagers may host visitors. Visitors usually hire a fishing boat on the N coast of Flores. Local tradition requires a live chicken be sacrificed, cooked, eaten on the slopes leading to the volcano before visiting it. It also requires the chicken heart to be offered to the volcano to appease it when visiting.
The aerial portion of Paluweh is a group of domes. Up until the 2012 – 2013 series of eruptions, the active dome was Rokatenda. The summit region of the volcano is irregular with overlapping craters and several lava domes. The more recent the dome, the less vegetation it has. There are multiple flank vents along a NW-trending fissure. As far as I can find, the volcano has not yet suffered a flank collapse, though the extruded domes do regularly collapse. The active volcano has volcanic fissures, fumaroles and solfataras.
Eruptive products are generally andesites trending to basaltic and pico basaltic andesites. Chemically, they are relatively low in silicates.
Paluweh was a partly barren, partly wooded summit immediately following the 1928 eruption. There were several small cones in the crater region Rokatenda. The collapsed portion of the dome was south of the summit. There were an additional six ancient craters and three older lava domes found. Three new craters were formed south of the summit and around an old cone. Following the Sept 9 eruption, a fourth crater was found in 1930.
The 10-month long eruption sequence Oct 2012 – Aug 2013 built the new Rerombola dome. There were at least three 2 – 3 month long high extrusion events that were ended by explosive events. Initial extrusion was fed by a degassed magma that built a dense, viscous plug in the conduit. Gasses from this plug escaped around the plug, forming multiple craters around the new dome. Degassed magma was forced out of the vent and unloaded deeper magma with much higher volatile content that forced the system into an explosive eruptive phase.
Paluweh has been historically active with at least 10 eruptions since a VEI 3 in 1650. There was an eruption of an unknown size in 1831. There were VEI 3 eruptions in 1928, 1972, and 2012. VEI 2 eruptions took place 1963 – 1966, 1973, and 1984. The only VEI 1 took place in 1985, though it did spread 3 cm of ash on villages and crops on the N side of the island. The known location of all these was the Rokatenda dome. Three of these eruptions do not have a listed location.
The Aug – Sept 1928 VEI 3 eruption was accompanied by an earthquake and tsunami. Population of the island was 266 at the time. The tsunami created by this eruption ran up 5 – 10 m on the island, killing 128 people and destroying the SW part of the island. It also formed a new dome in the summit region. A NYT headline Aug 10, 1928 claimed that the eruption killed 1,000 and wiped out six villages on Palu’e Island. Though not mentioned, the close proximity of Flores to the south of Paluweh and the 1,000 dead indicate the 5 – 10 km tsunami also hit the north Flores coast.
Increased seismic activity was reported in Apr 2009 and again in Jan 2012. Local residents observed a small zone of dead vegetation and new fumarole activity in the summit region June 2012. Lava extrusion began in this region, the beginning of the 2012 eruption. Extrusion continued over the weeks to follow, building a new lava dome, this one named Rerombola (The Friendly One). Photovolcanica observed dome extrusion and multiple flank collapses from it during a visit Nov 30 – Dec 2, 2012. The new dome was 150 – 200 m high by this time. It is on the west flank of Rokatenda and NE of the other suspected recent dome. Dome collapses at this time were rocky avalanches with minimal ash, indicating relatively low gas content in the extruded magma at least at the beginning. This changed in early Dec when a larger collapse generated a pyroclastic flow to the W that did not make it out of the summit region. Weak ash venting took place almost continuously during the Nov 30 – Dec 2 observation.
Several villages in the vicinity of the summit were evacuated as the new dome extrusion process began. Indonesian authorities set up a 3 km exclusion zone and evacuated some residents of the island. By the time of the deadly Aug 10 eruption, 10,000 still remained on the island, with 2,000 of them still within the exclusion zone.
Activity continued to ramp up Jan – Aug 2013 as the new dome ejected ash plumes 2 – 3 km high and pyroclastic flows. There were also dome collapses, explosions, avalanches and pyroclastic flows. As activity increased, the pyroclastic flows worked their way down the slope of the volcano, eventually reaching the ocean. Several ash plumes per month were reported by the Darwin VAAC. These were accompanied by MODVOC satellite thermal observations.
An explosion on Feb 3 put a plume to over 13 km. This plume drifted over 500 km SE. This eruption also put pyroclastic flows to the S and SW, avalanches, and removed perhaps 25% of the S portion of the new dome.
One surprising thing about the Feb 2013 eruption was injection of significant amounts of SO2 into the stratosphere by what was a relatively small tropical eruption. This eruption also had a disconnect between reported visible plume height of 4 – 6 km and satellite SO2 aerosol measurements to 16 – 18 km. The SO2 aerosol cloud was significantly higher than the ash plume. The Darwin VAAC initially reported the ash cloud above 13 km, with a final warning of ash at 7 – 8 km a day or so later. The eruption took place at 2300 L, which was fortuitous as there was regular foot traffic to and from the active dome as indicated by the great photography of it.
Satellite data was taken by a suite of new UV ozone mapping and profiler suite launched in Oct 2011. Conclusions from these observations are that reported plume altitudes may be biased low and cannot always be used to infer eruption magnitude, SO2 plumes, and resulting stratospheric impact. This may be a problem for tropical eruptions, with small SO2 injections into the stratosphere being more common that apparent from existing volcanic activity databases.
The Aug 10 eruption was deadly, ejecting a plume over 4 km and pyroclastic flow that finally reached the ocean on the N side of the island for the first time. It was funneled by existing stream channels (ravines) toward the beach. 6 living near the shore were killed in this eruption. The old Volcano Café covered the eruption with a post in 2013 by Chryphia and Carl. There was 10 – 20 cm of pyroclastic ash at the site of the village afterwards. Subsequent August eruptions were not as vigorous but still produced plume a couple kilometers high and pyroclastic flows. This flow extended the size of the island by a few meters.
The following video was taken on Dec 16, 2012 and posted to YouTube by Aris Yanto Dec 20, 2012. Plumes from this time rose 1.5 – 3.7 km in separate eruptions from 14 – 18 Dec.
Grey plumes were observed above the lava dome on occasion Jan – Apr 2014 as seismicity decreased after Nov 2013. The lava dome did not appear to have changed shape since the end of the 2013 eruptive sequence. White steam plumes were observed in Jan 2016 and Nov 2015. The activity had shallow and deep volcanic earthquakes. The 3 km diameter exclusion zone was still in force at this time.
Tectonics of Flores Island and neighboring Palu’e are driven by the collision of the northward moving Indo-Australian Plate into the Eurasian Plate. North of the main collision, the Pacific Plate is also colliding generally traveling to the west. This complex collision zone created the curved Banda Arc, a set of inner and outer arcs. Flores is located on the Inner Banda Arc. The inner arc is generally back arc volcanic islands separated from the older accretionary outer arc by a spreading forearc basin. The arc is over 1,000 km long. The Indo-Australian Plate is subducting under the Eurasian Plate along this arc.
The string of islands is called the Lesser Sunda Island. Flores Island is bounded by at least three basins – the Flores Basin to the NW, the South Banda Basin to the NE and the Savu Basin to the S. Sumba Island is located S of the western portion of Flores and Timor is located S of the eastern tip of Flores. There is speculation that Sumba is a sliver of microplate while Timor is the result of accretionary thrusting between the Savu Basin and the Timor Trough to its south.
Being a back-arc volcanic island, Flores is quite volcanic, significantly more volcanic that its neighbors to the E or W.
This portion of the Banda Arc has two tectonic phases. The first was an extensional phase 20 – 4 Ma. The S-migrating Sumba microcontinent opened the Savu basin. The incoming Indo-Australian Plate started a collisional phase around 8 Ma. Eastern Indonesia underwent a counterclockwise rotation with increased velocity between Bali and the E end of the Timor Trough. Flores is the transition zone from subduction to collision, likely explaining it vigorous volcanic activity.
A 2019 paper by Purwandono, et al makes a case that there is a slab rollback underway beneath Flores Island. This rollback contributes to the southward migrating volcanic arc on Komodo and Flores Islands. The migration is most visible in western Flores, becoming less visible in central and absent in E Flores. Volcanic edifices on Flores are visible as clusters or single chains, with clusters appearing to the W and single chains to the E.
Paluweh is an example of an active, dangerous back-arc volcano. It has been quite active over the years, primarily building and destroying domes with the volcanic effects that come with that process (plumes and pyroclastic flows). The good news is that there is not a particularly large number of people living on the island. The bad news is that they are widely scattered and in turn, close enough for any explosive eruption to be an immediate threat. Worse, despite multiple attempts to permanently evacuate residents from the island, they do not seem to be particularly interested in leaving. We know from the 1928 eruption that the island is capable of an earthquake / eruption sequence that will produce a tsunami large enough to kill people on neighboring Flores Island. While Flores Island is not densely inhabited, it has enough people making their living in the sea to be close enough to the shoreline so that tsunamis are an issue. This is a volcano that should be treated with a great deal of respect.
Final lesson from this volcano is the surprising satellite data demonstrating SO2 aerosols were lofted significantly higher than ash during the Aug 2013 eruptions. It is a lesson in looking at volcanoes with newer, different, and occasionally better instruments and finding something surprising.