John P. Lockwood

The 1986 Lake Nyos gas disaster in Cameroon, West Africa

The sudden, catastrophic release of gas from Lake Nyos on 21 August 1986 caused the deaths of at least 1700 people in the northwest area of Cameroon, West Africa. Chemical, isotopic, geologic, and medical evidence support the hypotheses that (i) the bulk of gas released was carbon dioxide that had been stored in the lakes hypolimnion, (ii) the victims exposed to the gas cloud died of carbon dioxide asphyxiation, (iii) the carbon dioxide was derived from magmatic sources, and (iv) there was no significant, direct volcanic activity involved. The limnological nature of the gas release suggests that hazardous lakes may be identified and monitored and that the danger of future incidents can be reduced.

Geophysical observations of Kilauea Volcano, Hawaii, 2. Constraints on the magma supply during November 1975–September 1977

Abstract Following a 22-month hiatus in eruptive activity, Kilauea volcano extruded roughly 35 × 10 6 m 3 of tholeiitic basalt from vents along its middle east rift zone during 13 September–1 October, 1977. The lengthy prelude to this eruption began with a magnitude 7.2 earthquake on 29 November, 1975, and included rapid summit deflation episodes in June, July, and August 1976 and February 1977. Synthesis of seismic, geodetic, gravimetric, and electrical self-potential observations suggests the following model for this atypical Kilauea eruptive cycle. Rapid summit deflation initiated by the November 1975 earthquake reflected substantial migration of magma from beneath the summit region of Kilauea into the east and southwest rift zones. Simultaneous leveling and microgravity observations suggest that 40–90 × 10 6 m 3 of void space was created within the summit magma chamber as a result of the earthquake. If this volume was filled by magma from depth before the east rift zone intrusive event of June 1976, the average rate of supply was 6–13 × 10 6 m 3 /month, a rate that is consistent with the value of 9 × 10 6 m 3 /month suggested from observations of long-duration Kilauea eruptions. Essentially zero net vertical change was recorded at the summit during the 15-month period beginning with the June 1976 intrusion and ending with the September 1977 eruption. This fact suggests that most magma supplied from depth during this interval was eventually delivered to the east rift zone, at least in part during four rapid summit deflation episodes. Microearthquake epicenters migrated downrift to the middle east rift zone for the first time during the later stages of the February 1977 intrusion, an occurrence presumably reflecting movement of magma into the eventual eruptive zone. This observation was confirmed by tilt surveys in May 1977 that revealed a major inflation center roughly 30 km east of the summit in an area of anomalous steaming and forest kill first noted in March 1976.

Origin of Comb Layering and Orbicular Structure, Sierra Nevada Batholith, California

A new descriptive term, comb layering , is proposed to replace the informal term Willow Lake-type layering , first introduced by Poldervaart and Taubeneck (1959) to describe layering in granitoid rocks in which constituent crystals are oriented approximately perpendicular to individual layers. The term schlieren layering is proposed to describe the “normal” layering of granitic rocks defined by alternating layers enriched or depleted in the normal mafic minerals. In such layers, elongate minerals commonly lie in the plane of the layering. Comb layering is widespread in plutonic rocks of California and is commonly associated with orbicular diorites. Evidence from a detailed study of three localities in the Sierra Nevada indicates that comb layering forms chiefly in overturned troughs along overhanging walls of plutons or along walls of dikes or pipes that cut country rocks adjoining plutons. Orbicular rocks associated with the comb layering are formed by a nucleus surrounded by multiple comb layers. The growth direction in comb layers can be determined by the branching and widening of plagioclase and hornblende crystals and is invariably toward the parent pluton. Field data indicate that comb layering cannot have formed from silicate magma, and further suggest that the layers have been deposited by large volumes of aqueous fluids that migrated upward along contacts between magma and wallrock or along the interface between magma and previously solidified melt. Comb layering and orbicules are largely restricted to structural traps into which upwardly migrating, solute-rich water was channeled owing to its low density. The comb layers grew on the solid walls of fluid-filled channels, whereas orbicules formed by precipitation of comb layers on hobbling inclusions suspended within the upward-flowing fluid.

Sedimentary and Gravity-Slide Emplacement of Serpentinite

Large deposits of serpentinite in alpine-type orogenic areas have been formed by sedimentary processes ranging from the detrital accumulation of bedded serpentinite sandstone and shale to the emplacement of chaotic breccias (olistostromes) and gigantic slide blocks. Known occurrences of sedimentary serpentinite are listed, and eight deposits from the circum-Pacific, Caribbean, and Mediterranean areas are described in detail. Sedimentary serpentinites range in age from early Paleozoic to Quaternary, although most are Cretaceous or Tertiary. Most were deposited in eugeosynclinal environments, early in the geosynclinal cycle. Individual deposits range in thickness from a few centimeters to nearly 3 km, and several extend laterally for tens of kilometers. Graded bedding is common, and many deposits contain marine fossils. Serpentinite is the dominant rock constituent, and clasts foreign to the alpine ultramafic assemblage are rare. Chemical analyses often detrital serpentinites show that these rocks contain slightly more silica and alumina than do nondetrital serpentinites, due to contamination by aluminosilicate minerals and quartz during deposition. This and nine other criteria are potentially useful in the recognition of sedimentary serpentinites. Several features suggest that most sedimentary serpentinites were deposited very rapidly by submarine landslides, mudflows, or turbidity currents. The sources of this serpentinite debris are postulated to be upward-migrating serpentinite protrusions which penetrate the seafloor or Earths surface upslope from eventual depositional sites. Sedimentary serpentinites are much more abundant in alpine-type orogenic areas than is commonly thought, and many ultramafic masses presently regarded as igneous intrusions or tectonic protrusions may in fact be coeval with, instead of younger than, their enclosing sedimentary or metasedimentary rocks. In eugeosynclinal sequences such as the Franciscan Formation, some elongate bodies now regarded as serpentinite sills may be beds of ultramafic detritus whose sedimentary features have been masked by post-depositional shearing; isolated masses may be exotic slide blocks. A sedimentary origin can explain some of the most persistent and perplexing characteristics of many alpine serpentinites: their conformity with enclosing sedimentary rocks, their grossly planar shapes, and the absence of metamorphism along their contacts.

Absolute paleointensity from Hawaiian lavas younger than 35 ka

Abstract Paleointensity studies have been conducted in air and in argon atmosphere on nine lava flows with radiocarbon ages distributed between 3.3 and 28.2 ka from the Mauna Loa volcano in the big island of Hawaii. Determinations of paleointensity obtained at eight sites depict the same overall pattern as the previous results for the same period in Hawaii, although the overall average field intensity appears to be lower. Since the present results were determined at higher temperatures than in the previous studies, this discrepancy raises questions regarding the selection of low versus high-temperature segments that are usually made for absolute paleointensity. The virtual dipole moments are similar to those displayed by the worldwide data set obtained from dated lava flows. When averaged within finite time intervals, the worldwide values match nicely the variations of the Sint-200 synthetic record of relative paleointensity and confirm the overall decrease of the dipole field intensity during most of this period. The convergence between the existing records at Hawaii and the rest of the world does not favour the presence of persistent strong non-dipole components beneath Hawaii for this period.

The 1977 eruption of Kilauea volcano, Hawaii

Abstract Kilauea volcano began to erupt on September 13, 1977, after a 21.5-month period of quiescence. Harmonic tremor in the upper and central east rift zone and rapid deflation of the summit area occurred for 22 hours before the outbreak of surface activity. On the first night, spatter ramparts formed along a discontinuous, en-echelon, 5.5-km-long fissure system that trends N70°E between two prehistoric cones, Kalalua and Puu Kauka. Activity soon became concentrated at a central vent that erupted sporadically until September 23 and extruded flows that moved a maximum distance of 2.5 km to the east. On September 18, new spatter ramparts began forming west of Kalalua, extending to 7 km the length of the new vent system. A vent near the center of this latest fissure became the locus of sustained fountaining and continued to extrude spatter and short flows intermittently until September 20. The most voluminous phase of the eruption began late on September 25. A discontinuous spatter rampart formed along a 700-m segment near the center of the new, 7-km-long fissure system; within 24 hours activity became concentrated at the east end of this segment. One flow from the 35-m-high cone that formed at this site moved rapidly southeast and eventually reached an area 10 km from the vent and 700 m from the nearest house in the evacuated village of Kalapana. We estimate the total volume of material produced during this 18-day eruption to be 35 × 10 6 m 3 . Samples from active vents and flows are differentiated quartz-normative tholeiitic basalt, similar in composition to lavas erupted from Kilauea in 1955 and 1962. Plagioclase is the only significant phenocryst; augite, minor olivine, and rare orthopyroxene and opaque oxides accompany it as microphenocrysts. Sulfide globules occur in fresh glass and as inclusions in phenocrysts in early 1977 lavas; their absence in chemically-similar basalt from the later phases of the eruption suggests that more extensive intratelluric degassing occurred as the eruption proceeded. Bulk composition of lavas varied somewhat during the eruption, but the last basalt produced also is differentiated, suggesting that the magma withdrawn from the summit reservoir during the rapid deflation has not yet been erupted.

Recovery of Datable Charcoal beneath Young Lavas: Lessons from Hawaii

Field studies in Hawaii aimed at providing a radiocarbon-based chronology of prehistoric eruptive activity have led to a good understanding of the processes that govern the formation and preservation of charcoal beneath basaltic lava flows. Charcoal formation is a rate-dependent process controlled primarily by temperature and duration of heating, as well as by moisture content, density, and size of original woody material. Charcoal will form wherever wood buried by lava is raised to sufficiently high temperatures, but owing to the availability of oxygen it is commonly burned to ash soon after formation. Wherever oxygen circulation is sufficiently restricted, however, charcoal will be preserved, but where atmospheric oxygen circulates freely, charcoal will only be preserved at lower temperature, below that required for charcoal ignition or catalytic oxidation. These factors cause carbonized wood, especially that derived from living roots, to be commonly preserved beneath all parts of pahoehoe flows (where oxygen circulation is restricted), but only under margins of aa. Pratical guidelines are given for the recovery of datable charcoal beneath pahoehoe and aa. Although based on Hawaiian basaltic flows, the guidelines should be applicable to other areas.

The potential for catastrophic dam failure at Lake Nyos maar, Cameroon

The upper 40 m of Lake Nyos is bounded on the north by a narrow dam of poorly consolidated pyroclastic rocks, emplaced during the eruptive formation of the Lake Nyos maar a few hundred years ago. This 50-m-wide natural dam is structurally weak and is being eroded at an uncertain, but geologically alarming, rate. The eventual failure of the dam could cause a major flood (estimated peak discharge, 17000 m3/s) that would have a tragic impact on downstream areas as far as Nigeria, 108 km away. This serious hazard could be eliminated by lowering the lake level, either by controlled removal of the dam or by construction of a 680-m-long drainage tunnel about 65 m below the present lake surface. Either strategy would also lessen the lethal effects of future massive CO2 gas releases, such as the one that occurred in August 1986.

The Uwekahuna Ash Member of the Puna Basalt: product of violent phreatomagmatic eruptions at Kilauea volcano, Hawaii, between 2800 and 210014C years ago

Abstract Kilauea volcanos reputation for relatively gentle effusive eruptions belies a violent geologic past, including several large phreatic and phreatomagmatic eruptions that are recorded by Holocene pyroclastic deposits which mantle Kilaueas summit area and the southeast flank of adjacent Mauna Loa volcano. The most widespread of these deposits whose original distribution can be reconstructed is the Uwekahuna Ash Member of the Puna Basalt, a basaltic surge and fall deposit emplaced during two or more eruptive episodes separated by a few decades to several centuries. The first episode occurred between 2770 ± 70 and 2265 ± 50 14 C yr ago. It included two major pyroclastic surges, each preceded by unusually vigorous lava fountaining from a vent near the volcanos summit. Before the second eruptive episode, 2110 ± 120 14 C yr ago, plants had re-colonized the rainforest environment northeast of the summit, and at least two lava flows from Mauna Loa had buried parts of the first-episode deposits. The second episode also began with vigorous lava fountaining, followed by widespread lithic ashfall, a third major surge and finally a fourth fountaining event. Before the final pumice deposit could be significantly reworked, it was partly buried by picritic basalt flows that are unusual in Kilaueas summit area. In proximal areas, the Uwekahuna Ash Member is more than 1 m thick (locally > 5 m) and includes lithic blocks up to 0.8 m in diameter. Coarse, primarily lithic debris was deposited around the vent by laterally expanding surges; fallout deposits accumulated preferentially to the northeast under the influence of high-altitude counter-tradewinds. The area devastated by surges and originally buried by at least 15 cm of the Uwekahuna was about 420 km 2 . The bulk volume of the deposits was approximately 0.3 km 3 , including less than 0.1 km 3 of juvenile material. Juvenile constituents are olivine-tholeiitic basalts similar in major-element composition to typical Kilauea summit lava flows, but variations in both major elements and trace elements suggest that the eruptions tapped more than a single, uniform source region. We infer that the eruptions which produced the Uwekahuna were driven by water interacting with a fluctuating magma column. Magma withdrawal episodes may have been accompanied by large-volume submarine effusive eruptions and by summit collapse. The volume, extent and character of the Uwekahuna deposits underscore the hazards posed by relatively infrequent but potentially devastating explosive eruptions at Kilauea, as well as at other basaltic volcanoes.

Origin and age of the Lake Nyos maar, Cameroon

Abstract Lake Nyos occupies a young maar crater in the Precambrian granitic terrane of northwest Cameroon. The lake is partly surrounded by poorly consolidated, ultramafic nodule-bearing pyroclastic surge deposits that were explosively ejected from the Nyos crater at the time of its formation. Radiocarbon dates show that the maar probably formed about 400 years ago. Field evidence suggests that carbon dioxide could have been the principal volatile involved in the formation of the Nyos maar, and that the role of water may have been minor. The formation of the Nyos maar was preceded by a brief period of effusive basaltic volcanism, but the maar itself may have largely formed by cold, ‘dry’ explosive processes. Carbon dioxide may still be trapped interstitially in a diatreme inferred to underlie Lake Nyos; its gradual release into the waters of Lake Nyos may have set the stage for the tragic gas-release event of August 21, 1986. Only young maar lakes such as Nyos may pose a danger of future lethal gas releases.