Kikai Volcano / Mt. Iwodake (Iwojima)

Kikai Caldera was the site of the VEI 7 Kikai-Akahoya eruption, one of the largest eruptions in the last 10,000 years, paralleling the more famous but less powerful eruption of Krakatau in 1883 in that massive pyroclastic flows travelled across the sea, devastating mainland settlements. Kikai is the southernmost and most recent of a series of N-S aligned calderas in Kyushu (the most southerly of Japans main islands), which further include, from North to South, Aso, Kakuto, Aira (see Sakurajima) and Ata. Whilst the Aira and Ata Calderas together form Kagoshima bay on the south coast of Kyushu, Kikai is located about 50km offshore. Only the islands of Satsumo-Iwojima and Take-shima, formed from remains of the northern flank of the pre-caldera edifice and post-caldera volcanism, and the small Showa-Iwojima lava dome formed in 1934-1935, today lie above sea level, with the floor of the about 18km wide caldera being up to 500 meters deep. The 7.3ka (7300 year ago) "Akahoya" eruption was not the only Caldera-forming eruption in the history of Kikai as revealed by massive "Koabi" and "Nagase" ignimbrite deposits dated at 140ka and 95ka, respectively.

Postcaldera activity started with formation of the rhyolite lava dome Mt. Iwodake which remains active to the present day and is a site of intense fumarolic activity. Inamuradake scoria cone and associated basaltic lava flows, formed about 3000 years ago. Both form parts of Satsumo-Iwojima island. Showa-Iwojima is a monogeneic lava dome which rose from the caldera floor starting at a depth of 300 meters below sea level in 1934-1935. An up to 530m wide and 55m high island was formed about 2 km east of Iwojima. Several submarine structures which may also be lava domes emplaced in the post-caldera stage are found in the western part of the caldera, SE of Satsuma-Iwojima. It is estimated that about 45 cubic kilometers of magma (7.5 per 1000 years) has been erupted in the post-caldera stage. This is significantly higher than at other Japanese Quaternary volcanoes, confirming that a large magma chamber must remain under the caldera (Saito et al. 2001 (J. Volc. Geotherm. Res. 108, p.11-31; Ref. therein)).

Mt. Iwodake (Iwojima)

Satsuma-Iwojima island forms part of the rim of the largely submerged Kikai caldera. Two prominent post-caldera features have contributed to the present shape of the island, Mt. Iwodake, a rhyolitic lava dome, and the smaller Inamuradake scoria cone (see Kazahaya et al., 2002. Earth Planets Space 54, p.327-335; references therein). The latter was formed, together with associated basaltic flow fields, about 3000 years ago. Formation of Mt. Iwodake is thought to have started soon after caldera formation, with effusive dome building sometimes accompanied by more explosive eruptions forming pumice and pyroclastic flow deposits. The rhyolites of Iwodake are similar in composition to those erupted during the caldera-forming eruption and are thought to represent remnants of the same magma body. Petrological analysis suggests that the Iwodake Rhyolites were stored at a depth of only 3km before migrating to the surface, whilst the basalts erupted from Inamuradake were sourced from a depth of 3-5km, suggesting a layered magma chamber.

Iwodake remains active with weak ash emissions occasionally occuring, although the last significant magmatic activity is thought to lie about 1300 years back. The petrology of Satsuma-Iwojima is studied in detail in Saito et al. 2001 (J. Volc. Geotherm. Res. 108, p.11-31). Fumarolic activity is a constant feature of Iwodake in recent times, with fumarole temperatures reaching as high as 900'C. In the summit crater, fumarole chimney structures can sometimes be observed. Historical records suggests that sulfur was being deposited on Iwodake by fumarolic activity at least 800 years ago. From 1990-1998, on average 550 tonnes of sulfur dioxide were emitted per day, together with 16000 t of water and 150 t carbon dioxide (Kazahaya et al. 2002). Kazahaya suggested that continuous magma convection is occuring, with gas-rich bouyant magma rising in the conduit, degassing as it rises until it is relatively depleted in gas and more dense than the magma below. This results in it gradually sinking again as fresh gas-rich magma ascends. The system is probably driven by the underlying body of basaltic magma which provides both a heat source for the unusually hot rhyolitic magma of Iwodake, and also a source of gases, which probably migrate upwards from the basaltic magma into the convecting rhyolitic body above. Seismic analysis localized tremor probably associated with magma cycling to a depth of only 40 meters under the edifice, showing that the system is shallow with magma close to the surface (Ohminato and Ereditato, 1997. Jap. Geophys. Res. Lett. 24, p.2845-2848). Based on current degassing rates, it is estimated that gases derived from about 100 cubic km of basaltic magma have been emitted by Iwodake in the last 800 years.

The summit of Iwodake comprises several crater structures (together 400 x 140m wide) which are the result of explosive activity following emplacement of the dome. The central Oana crater is adjoined to the SW by the smaller Kintsuba crater. Significant explosive eruptions occurred around 1200 and 500-600 years ago and were associated with pyroclastic flows which descended the west flank of the dome. Smaller ash eruptions have occurred in recent times. In 1988, a series of small ash eruptions were observed on January 18. In 1998, on several occasions, further ash eruptions were recorded, accompanied with increased tremors and a small pit crater formed in the summit crater in 1998. The eruptions do not appear to have involved juvenile material. Further infrequent small ash eruptions were observed in 1999, 2001, 2002-2004 and 2013.

The main (Oana) crater has been a site of sulfur mining and roads were constructed for this purpose leading into the summit crater. More recently, mining of silicic minerals was performed in the Otanibira quarry on the west flank below the summit (until 1997).

Map Satsumo Iwojima island

Kikai-Akahoya Eruption

The sequence of eruptive events during the VEI 7 Kikai-Akahoya eruption has been reconstructed in detail based on analysis of deposits on the islands representing remnants of the caldera rim and on mainland Kyushu. (Maeno and Taniguchi, 2007. J. Volc. Geotherm. Res. 167, p.212-238).

Phase 1 of the eruption was a powerful Plinian eruption emplacing pumice and ash deposits (Units A1-A4) several meters thick near the vent and still a meter thick on the southern tip of Kyushu, 40km away. Units A1-A2 represent an intial Plinian phase terminating in a column collapse event (deposit A2). This was then followed by a more powerful Plinian eruption associated with vent enlargement (recognizable by high proportion of lithics from walls of conduit in deposit A3) and sustained high discharge rate resulting in an over 40km high ash cloud and depositing about 40 cubic km of tephra over a period of at least 28 hours.

Phase 2 involved collapse of the Plinian column and access of water to magma in the upper parts of the conduit as the eruption rate declined. Deposits (Unit B) appear to have been largely emplaced by a succession of pyroclastic flows associated with powerful explosive eruptions and are predominantly found on topographic lows on Satsuma Iwo-jima Island with a maximum thickness of about 3 meters.

Phase 3 is split into two sub-phases, the first of which involving further water-magma interaction and initiation of caldera collapse. This resulted in emplacement of Units C1 and C2 (each up to nearly 5m thick), in similar locations to Unit B deposits. Finally, the climax of the eruption occurred with catastrophic caldera collapse, a huge Plinian ash cloud and voluminous pyroclastic flows which reached mainland Kyushu. This resulted in emplacement of Unit C3a-b (10 cubic km; C3a up to 10m thick; C3b up to 3m thick) and C3c (20-35 cubic km; 30m thick on Takeshima), both of which are found on mainland Kyushu. Unit D represents the ash-fall deposit associated with the climactic eruption. It is spread over a huge area, being detectable and has a total volume of 100 cubic km. For a detailed analysis of the morphology and petrology of the deposits see Maeno and Taniguchi. Units C1 and C2 are not found on Takeshima, suggesting they must have been erupted somewhat laterally in order to reach mainland Kyushu much further away.

In total it is estimated that during the course of the entire eruption, 70-80 cubic km of magma (DRE) was erupted from a magma chamber about 3-7km below the surface with a diameter of about 5 km. The eruption was essentially rhyolitic (silicate levels 72-74%), except for the inclusion of small amounts of andesitic material (56%) in the final stages of the eruption, presumably from the bottom of the magma chamber.

The climactic pyroclastic flows of the Kikai-Akahoya Eruption extended as far as 100 km from the volcano and inundated the southern parts of the Osumi and Satsuma peninsulas in southern Kyushu, leaving deposits (Units C3a-c; predominantly Unit C3c which was lighter as ash and pumice rich and thus better able to travel over the sea) e.g. about 50 cm thick at 50 km from source. It is unlikely that anybody could have survived. Indeed, following the eruption, the Kyushu-Kaigara-Mon and Oshigata-mon styles of pottery attributed to the indiginous population of southern Kyushu are no longer found at archaeological sites. Pottery found in layers above the pyroclastic flow deposit are invariably of styles developed outside of the region (Machida, 1984. Geol. Survey Japan. Rep. 293, p.301-313). The flows also had a major impact on the vegetation in the area. The lucidophyllous forests such as Castanopsis and Lauraceae associated with bamboo grass bush were totally destroyed and replaced by Miscanthus dominated grassland (Sugiyama, 2002. Quaternary. Res. 41(4), p.311-316). Stone implements used for processing of tree nuts are virtually absent in soil layers immediately above the PF deposit and only gradually return, starting on northerly areas of the inundated zone (Kuwahata, 2002. Quaternary. Res. 41(4), p.317-330). Forest recovery was slow and took at least 600 years, in places far longer. Even about 2000 years after the eruption, settlement sizes in the area were smaller than pre-eruption settlements. The deposits from the eruption are so widespread that Japanese archaeologists use them as a stratigraphic marker, called the Akahoya layer.

Map Satsumo Iwojima island

Caldera-forming eruptions in marine environments may be associated with devastating tsunamis such as those at Santorini and Krakatau. Sediment analysis in Tachibana Bay on the western coast of Kyushu (220 km from the volcano) suggested that a significant tsunami may have been generated by the Akahoya eruption (Okamura et al. 2005, Joint Meeting Earth Planetary Sci. Abstract J027-P025). This would have undoubtedly taken its toll on coastal communities outside of the pyroclastic flow zone. However, no evidence for tsunami inundation of mainland Kyushu could be found at several sites studied along the south coast, although it is possible that traces of the tsunami were destroyed by the climactic pyroclastic flow (Maeno et al. 2006. Earth Planets Space, 58, p.1013-1024). Mathematical modelling by the authors suggest that large tsunamis would have been expected unless caldera collapse occured at a relatively slow speed.

Evidence exists at archaeological sites on mainland Kyushu that at least 2 massive earthquakes accompanied the eruption (Naruo and Kobayashi, 2002. Quat. Res. 41, p.287-299). However, whether these caused a significant death toll prior to arrival of the climactic pyroclastic flows is not known.

Showa Iwo-jima

Showa-Iwojima lava dome is the result of an initially underwater eruption starting in 1934. The eruption is described in detail by Maeno and Taniguchi (Bull. Volcanol. 68, p.673-688 (2006)), based on the accounts of Tanakadate and more recent geological studies.

In September 1934, pumice was noticed rising to the surface and tremors were felt on the islands. By December 8th, a pyroclastic cone became visible on the surface in an area where the sea-floor lay at a depth of 300 meters, and a small island started to form around it. Frequent explosive eruptions (at times every 1-2 minutes) were observed with cauliflower-shaped dark ash clouds emerging from the otherwise steam-rich white plume. On December 30, the pyroclastic cone was destroyed by a larger explosion, marking the start of an ephusive phase with interspersed series of phreatomagmatic explosive eruptions generating classic "cocks-tail" jets of dark ash. In January, effusion was accompanied by growth of a new pyroclastic cone, reaching a height of about 13 meters. From late January to March, new silicic lava effused and formed a dome. Small explosive eruptions occured in March as the cone widened and parts of its rim collapsed. Subsequently, the remnants of the cone were largely buried as a final effusive phase occured, covering much of the surface of the island. The eruption ceased by the end of March. A further small islet was formed 50m north of the main island, starting on February 10th, yet this was soon eroded by wave action. Indeed, the main island, which had dimensions of about 530 x 150 m and rose up to 55m out of the sea has also been partially eroded at its edges and also apparently subsided (and presumably shrunk slightly as the lava cooled) since its highest point dropped by about 30 m within three months of the eruption and is today only about 20m above sea level.

Average effusion between 21 January and 26 March could be calculated by estimating the difference in the size of the island between the two dates. Maeno and Taniguchi calculated an effusion rate of 100000 cubic meters per day for this period. The dome has a relatively flat "axisymmetric" profile with no significant spines and a low relief, in contrast to the spiny and steep-sided domes seen at e.g. Mount Unzen, or indeed at the intensely studied Soufriere Hills Volcano. This is due to the relatively high temperature and lower viscosity of the lava compared to e.g. Unzen. In some ways it can be considered as a small silicic equivalent of a shield volcano. Viewing lava composition, one sees a gradual transition from dacite lavas to rhyolite lavas (by mid-January) which eventually covered all but parts of the perimeter of the island. Silicate content ranges from 67% at the perimeter of Showa-Iwojima to 73% in the center of the island where the two effusive vents which emplaced the large flows shortly before the end of the eruption were located.

Visitor Information

Kikai caldera can be reached by ferry from Kagoshima. The ferry terminal for the island connection lies south of the Sakurajima ferry terminal in the center of town. The same ferry line also services Suwanosejima volcanic island further south. Satsumo-Iwojima and Take-shima islands are both inhabited and a few guest houses and stores are present. Overseas visitors should not expect to find many english-speakers on the islands, although with a bit of patience and sign language it is normally possible to make simple arrangements, since the Japanese are generally polite and helpful.

Whilst a disused road leads up to the summit of Iwo-Dake, access may be restricted. Entering the crater of a strongly degassing volcano may be lethal if gas concentrations reach critical levels. Four tourists were indeed killed in 1997 at another Japenese volcano, Adatara (which also was historically used as a sulfur mine), due to hydrogen sulphide gas poisoning.

Photovolcanica Full Index