|
Volcaniclastic Rocks of the Orton-Bradley Formation, Banks Peninsula, New Zealand. Discussion and Geologic History
|
|
Chapter 3: Terminology and Stratigraphy Chapter 5: Physical Volcanology Chapter 6: Interpretations and Lithofacies Analysis Chapter 7: Discussion and Geological History Acknowledgements and References
|
Chapter 7
Discussion and Geologic History The deposition of the Mt. Bradley Volcaniclastic Member has been dated between 9.5 to 9.1 Ma (Sewell, Weaver and Thiele 1988; Sewell and Weaver 1990). The deposition of the Tablelands Volcaniclastic Member occurred at some time after this (probably between 9.1 to 8.9 Ma approximately). Eruptions from the Lyttelton Volcano had ceased and the main activity, during this period, centred on the southeastern breach in the Lyttelton crater wall. Lavas from the earlier Lyttelton eruptions formed the walls of the breach, into which the Orton Bradley Formation rocks were erupted. It is highly likely that the eruptions took place over a much shorter time period, ranging from months to years, with long time breaks in which the epiclastic beds were deposited. No epiclastic beds are noted in the Tablelands Volcaniclastic Member and it is possible that this deposit was the result of two eruptive cycles, one for each vent, each of which was active for a relatively short period of time. 7.1 Eruptive/Depositional Styles and Mechanisms In order to establish the mechanisms responsible for the formation of the Mt. Bradley Volcaniclastic Member and the Tablelands Volcaniclastic Member, it was necessary to research several widely different topics. Work on volcaniclastic facies by Lajoie (1979) and Suthern (1983) outlined several possible mechanisms: subaqueous pyroclastic flows, base surges, subaqueous gravity flows, shallow water and subaerial reworking. Papers on the relevant epiclastic processes were read, particularly those relating to epiclastic processes associated active volcanism. The deposits noted in fluvial and shallow water environments by White (1990), Lewis and Ekdale (1990), Pirrie (1989), Smith (1986, 1987, 1987), Lewis (1983) and Monroe (1981) are distinctly different from the deposits of the two volcaniclastic members, so shallow water and subaerial reworking was discarded as the primary depositional mechanism. Turbidites (Lewis 1983, Collinson and Thompson 1982, Cas et al 1981,) and subaqueous pyroclastic flows (Cas and Wright 1987, Fisher and Schmincke 1984, Fiske 1963) were also examined and discarded as possible depositional mechanisms. The structures observed differed widely from those that would be expected and there was no evidence to support the theory that the volcaniclastic members were deposited subaqueously. The evidence both from field and laboratory work indicates that most of the Mt. Bradley Volcaniclastic Member and all of the Tablelands Volcaniclastic Member is composed predominantly of base surge deposits formed by phreatomagmatic volcanism from the vent on the northern flanks of Mt. Bradley. The following two sections cover the mechanisms involved in these eruptive and depositional processes in general and relate the Mt. Bradley and Tablelands Volcaniclastic Members to deposits of similar origins. 7.1.1 Phreatomagmatic Volcanism Phreatomagmatic eruptions are the result of the explosive interaction of magma and water. The water component can be surface water or ground water. Phreatomagmatic volcanism involves fuel-coolant interaction in which magma is the fuel and water is the coolant (Sheridan and Wohletz 1983). The fuel is required to be at a temperature in excess of the boiling point of the coolant, with magma having temperatures often in excess of 1000° C this requirement is easily met. On contact the water at the interface is vaporised, often explosively, causing fragmentation and quenching of the melt contact. This fragmentation increases the surface area of the water/melt interface, which in turn increases the vaporisation/ fragmentation rate, and increases the surface area of the water/melt interface. Thus the water/melt interaction increases due to this positive feedback system until the energy released exceeds the confining pressure and an explosive eruption occurs (Fig. 7.1). The extent of the explosive reaction is largely determined by the water:magma ratio, Sheridan and Wohletz (1983) gave the ratio of 0.3 as being the optimum ratio for the maximum degree of fragmentation and distribution. If less water is involved then the predominant driving force behind the eruption is the degassing of the magma, forming Strombolian scoria cones. While excessive amounts of water result in Surtseyan eruptions and lahars. The magma rise rate is important here as low rates may cause temperatures to rise slowly, evaporating the water (if only small volumes are available) before the magma front reaches the water contact. Other controlling factors also come in to play when determining the type and extent of a phreatomagmatic eruption. The nature of the water component and the extent of its mixing with the rising magma, generally determines the eruptive style, with deep ground water contacts forming maars, while subaqueous eruptions can range from pillow lavas (if confining pressures are great enough) to explosive Surtseyan eruptions. The Mt. Bradley Volcaniclastic Member and the Tablelands Volcaniclastic Member deposits have few features indicating the presence of large volumes of water during the eruption of the bulk of the pyroclastic deposits. This and the general range of the distribution of the deposits, indicates the more explosive fuel-coolant interaction was the mechanism responsible for these phreatomagmatic eruptions. 7.1.2 Base Surges Base surges are the eruption style most commonly associated phreatomagmatic volcanism, surges are described by Cas and Wright (1988) and Fisher (1979) as being turbulent, highly expanded, low particle concentration flows. They are topographically controlled, forming the thickest deposits in depressions while still mantling the topography (either radial or directional depending on the topography). Wohletz and Sheridan (1979) describe surge deposits as systematically varying away from the vent as the surge deflates, moving from a sections dominated by sandwave facies beds to massive facies beds to planar facies beds. Near the vent much of the particle transport is due to tractional processes because most of the material is lean phase fluidised. This results in the formation of the various `sedimentary' features characteristic of the sandwave facies bed. The massive facies beds are formed by the `freezing' of densely fluidised material, while the outer most planar facies beds are due to nonfluidised, inertial flows stopping due to frictional drag. There are variations in the deposits depending on the temperature of the surge and the moisture content, with cooler, wet surges having increased cohesion between both other particles in the surge and the ground surface over which it is travelling. Recent work by Sohn and Chough (1989,1990), arrived at a different model for base surges, this model is not exclusive of the above model, but refers to drier base surges than those covered by Wohletz and Sheridan (1979) and Fisher (1979). As the turbulent, highly concentrated surge cloud spreads out from the vent, unstratified, poorly sorted, generally massive lapilli tuff units are deposited by the `freezing' out of the high concentration underflow. As the surge moves further from the vent, it becomes more diluted due to sediment loss and the incorporation and expansion of ambient air. Decreased turbulence results in segregation of grainsize within the surge cloud. Stratification becomes more apparent, with bedding thickness and grainsize decreasing down current (Fig. 7.2). Saltation and interaction between the basal load and the topography may result in undulatory bedding and various other `sedimentary' bedding structures. This process continues until the surge cloud has become very dilute and well segregated, with thin, fine grained tuff units being deposited. Should the surge cloud cool enough for the steam to condense, then depositional features more in line with `wet' surge deposits will be seen, though this is only likely in the more distil sections. It should be noted that the surge cloud does not move as a single discrete body, but rather as a series of pulses. The base surge deposits of the Mt. Bradley Volcaniclastic Member appear to be relatively `dry', ie it is most likely that steam rather than water was the fluidising medium rather than liquid water, as there are no vesiculated tuffs or accretionary lapilli, and the overall sections fit in more closely with Sohn and Chough theory, than that proposed by Wohletz and Sheridan. The structure and bedforms of the Tablelands Volcaniclastic Member, being wetter, appear to fit the Wohletz and Sheridan model. 7.1.3 Comparisons With Similar Deposits In order gain a greater understanding of the Mt. Bradley Volcaniclastic Member and the Tablelands Volcaniclastic Member, comparisons were made with other phreatomagmatic deposits in New Zealand and around the world. There are several basaltic phreatomagmatic deposits in the Taupo Volcanic Zone, each with different characteristics. The Kaiapo deposit was described by Smith (1990) and is believed to have been erupted into a proto Lake Taupo. The deposit is spread over a relatively small area (<1km), with the deposits being distinctly different from those on Banks Peninsula. The lithofacies making up the Kaiapo deposit indicate greater volumes of water were involved in the eruptions of the base surges, than those of both the volcaniclastic members studied in this thesis. The nearby Pekanui Stream deposits are though to be partially the result of dry base surge eruptions (Sutton 1990), and are interbedded with scoria deposits. The dry base surge units are similar to some of the facies present in the Mt. Bradley Volcaniclastic Member, though it has no scoria beds. Other phreatomagmatic deposits in the vicinity were also noted (Wilson and Smith 1985). These are made up of matrix supported lapilli tuff, with little structure being visible, making determination of depositional mechanisms difficult. Another deposit, near Ohakune, is also a mixture of strombolian and phreatomagmatic deposits (Houghton and Hackett 1984). The phreatomagmatic deposits are airfall deposits and were deposited less than 1km from source. These deposits are quite distinct from those of the volcaniclastic members studied, especially as neither volcaniclastic member has any pure scoria deposits. The nature of this deposit does, however, indicate that the water source of the two volcaniclastic member is relatively continuous, giving a relatively constant supply. Other deposits from around the world were also compared, but these tended to be either wetter (Waters and Fisher 1971, Leys 1983, Kokelaar 1983 and 1986, Moore 1985, Kokelaar and Moore 1987) or drier (Houghton and Schmincke 1989) than the eruptions responsible for the volcaniclastic members of the Orton Bradley Formation. However the deposits associated with the Songaksan (Chough and Sohn 1990) and the Suwolbong (Sohn and Chough 1989) tuff rings in Korea, are very similar to those of the Mt. Bradley Volcaniclastic Member. The lithofacies noted in these deposits and their lateral facies sequences compare favourably with those of the Mt. Bradley Volcaniclastic Member, so the model derived from that research was used in the interpretation that member. It should also be noted that the deposits of the Songaksan tuff ring are attributed to the explosive interaction between magma and ground water in an aquifer underlying the deposits. 7.2 Geological History Outline of the eruptive history of the Orton Bradley Formation: (1) 9.5-9.1 Ma (2) 9.1-8.6 Ma (3) 8.6-8.3 Ma Erosion formed a deep, southwest-draining breach in the Lyttelton crater wall and excavated the interior of the crater (Fig. 7.3, 1a). Following this, from 9.7 to 9.5 Ma, the lavas in the lower Orton Bradley Formation, were erupted from a vent at the northern end of Orton Bradley Park and another in the vicinity of the Head of the Bay. These eruptions were hawaiian, forming major lava sheets which flowed down to east, into the breach in the Lyttelton crater wall, forming the foundations for the Mt. Bradley Volcaniclastic Member (Fig. 7.3, 1b). After cessation of volcanism from these vents, approximately 9.5 Ma, the eruption of the Mt. Bradley Volcaniclastic Member began from a vent within the breach of the Lyttelton crater wall (Fig. 7.3, 1c). Magma rising up to the vent on the northern flanks of Mt. Bradley probably came in to contact with a body of surface water. The early eruptions were probably of the 'wet' base surge type, forming a small tuff cone. This tuff cone probably only covered a limited area, with deposits being limited to the near vent area. A small crater lake formed within the crater of the tuff cone, providing the water component for the phreatomagmatic eruptions. It should be noted the this early eruptive phase is poorly exposed, with most of the interpretation being supposition based on the limited evidence available. Further activity within the crater, probably prior to a explosive eruption, tend to overflow the crater lake, discharging a series of water saturated fine tuff flows that formed units 1a-1d and unit 2a. Base surge eruptions were also taking place during this phase and their deposits are interbedded with the fine tuff deposits. This phase indicates a change in the eruptive style of the Mt. Bradley Volcaniclastic Member with decreasing water:magma ratio, either because the vent area was becoming progressively drier or the magma flow rates or volumes were increasing. This resulted in the eruptions becoming more powerful and forming more widespread deposits. The low angle of bedding, the distribution and relatively coarse nature of deposits, indicate that the later eruptions formed a tuff ring (Wohletz and Sheridan 1983). The drier nature of the vent is supported by the lack of accretionary lapilli, the erosive bases of deposits and the downwarping of beds due to ballistics instead of penetration (Fisher and Schmincke 1984, Walker 1980). The upper most pyroclastic beds are characteristic of the Sohn and Chough (1989) model for relatively dry base surge deposits. Though the levels of water entering the vent still varied, the eruptive cycle tend to show a general trend of drying, with the lower deposits having a higher water:magma ratio than the upper flows. However there is nothing to indicate that the vent dried out completely at any stage, as no signs of lava flows or strombolian deposits were found within the Mt. Bradley Volcaniclastic Member. After the deposition of the pyroclastic units from the Mt. Bradley Volcaniclastic Member, hawaiite lava flows from hawaiian style eruptions from vents in the vicinity of Mt. Herbert, partially sealed the breach to the southeast of Mt. Bradley (Fig. 7.3, 1d). This blockage prevented the drainage of the crater area, allowed the formation of a shallow freshwater lake, centred roughly on the northeastern flanks of Mt. Bradley and draining to the southeast (Fig. 7.3, 1e). This lake formed a relatively narrow band, orientated in a southeasterly direction, confined to the east by Lyttelton Volcanic Group lavas and to the west by a combination of the basement high, the Allandale rhyolite and the Lyttelton lavas. The presence of this lake, the large and the relatively high amounts of plant fossils in the lower mudstone deposits, indicates a relatively long break in the eruptive cycle of the area. The infilling of this lake with epiclastic and some pyroclastic (as it appears that some small scale volcanic activity occurred during the later part of this period) material caused it to dry up. The rate of infilling is difficult to determine, but it may have been relatively rapid as there would have been large amounts of poorly consolidated, unstable volcaniclastic deposits with little vegetation binding them together. Sewell (1985) believed that the lake was present for approximately 200ka and this appears to be a reasonable estimate. This was followed by the eruption of some thin basaltic lava flows, which covered the lacustrine sediments and further filling the breach in the Lyttelton crater walls (Fig. 7.3, 1f). The area then moved into another phase of phreatomagmatic volcanism with the eruption of the Tablelands Volcaniclastic Member (Fig. 7.3, 1g). These eruptions formed tuff cones around the earlier vent and a new vent that formed to the southeast. The eruptions were of the 'wet' base surge variety, forming relatively localised deposits. Water supply to the vents appears to be relatively constant, with little variation between the individual units, and is thought to be derived from water saturated deposits of the Mt. Bradley Volcaniclastic Member. The steeper bedding angles (up to 20° ) tend to indicate the formation of a tuff cone (Wohletz and Sheridan 1983), rather than the lowerangle tuff ring attributed to the Mt. Bradley Volcaniclastic Member. Eruptions from the tuff cone on the northeastern flanks of Mt. Herbert were similar to those of the vent on the northern flanks of Mt. Bradley. This tuff cone was also steep sided, with generally finer deposits covering a more limited area, generally less than 1km from source. There is some tuff supported scoria beds and evidence of lava ponding, indicating some limited dry or relatively dry vent activity. The deposition of volcaniclastic deposits was followed by a vigorous hawaiian style eruptive phase (Fig. 7.3, 2). From 9.1 to 8.6, eruptions continued along the southern flank of the Lyttelton Volcano with three eruptive vents, one positioned below what is now Mt. Herbert and the others to the southwest and to the northeast respectively. From these vents the upper flows of the Orton Bradley Formation and the Herbert Peak Hawaiites were deposited (Fig. 7.3, 3). This phase completely sealed the breach in the crater wall, to form what is today the highest point on the peninsula. |