Report on Open University Field Workshop:
Socompa Debris Avalanche Deposit, northern Chile

Scott Bryan, Department of Geology & Geophysics, Yale University, PO Box 208109,
New Haven CT 0520-8109 USA.
Tel: +1 203 432 1269, Fax.: +1 203 432 3134. Email: scott.bryan@yale.edu



An informal six-day field trip to the Socompa debris avalanche deposit near the northern Argentina-Chile border took place during 22-28 November 2004, following the IAVCEI General Assembly in Pucón, Chile. The trip was organised and led by Professor Steve Self and Marie-Noelle Guilbald from the Open University (UK), with the field trip in part prompted by Masters' research Marie-Noelle had undertaken on the avalanche deposit.

Participants flew to Antofagasta, from where we drove to the BHP Billiton porphyry copper mine of La Escondida, about 2 hours east of Antofagasta; the debris avalanche deposit is another 2 hours drive further east from the mine. The mine obtains much of its water from aquifers beneath the debris avalanche deposit, and new roads built by the mine to access drill sites, pumping stations and permanent camps have greatly increased access to the deposit over the last few years. However, due to the increased exploration activity undertaken by the mine in the region, and because we had to traverse across the La Escondida mine, access in an out of the region would not have been possible without the assistance of BHP Billiton. We were able to stay at one of the La Escondida water exploration camps (Monturaqui) located in the central part of debris avalanche deposit that afforded us easy day trips to access all parts of the debris avalanche deposit.

Socompa Volcano provides one of the world's best exposed examples of a sector collapse derived debris avalanche deposit (van Wyk de Vries et al., 2001), thought to have occurred approximately 7000 ybp. The hyper-arid climate of the Atacama Desert of northern Chile has aided in the remarkable preservation of this deposit. The avalanche deposit had a run-out length of 40-50 km, covering an area of up to 500 km
2. The avalanche deposit volume is 36 km3, of which 11 km3 are large slide or 'toreva' blocks (volumes from Wadge et al., 1995), which remained stranded in the upper part of the amphitheatre. The collapse scar or amphitheatre is a wedge, 12 km wide at its mouth and ~300 m deep. The majority of the avalanche deposit is made up of basement material (Pliocene Salín Formation) that formed a pedestal upon which Socompa was constructed. The Salín Formation, where exposed, comprises weakly to moderately indurated, biotite-rich, felsic block and ash flow deposits, fluvial to sheetflood-type sedimentary breccias/sandstones and salar (evaporitic lacustrine and salt) deposits. The remaining avalanche deposit volume is composed of a veneer of volcanic rocks from Socompa (Ramirez, 1988; Wadge et al., 1995). Although the summit of Socompa is >6000 m ASL, one of the intriguing aspects of the debris avalanche deposit is that the most distal components of the deposit were sourced from some of the lowest pre-collapse elevations of the volcano (ie. had a vertical drop of ~500 m); previous workers have surmised that this occurred by efficient transfer of momentum from rear to front of the avalanche (Wadge et al., 1995).

Fig1

Figure 1. The Field party with the debris avalanche deposit and Socompa Volcano in the background.

Day 1 was spent getting a general overview of the deposit. First stop was the intriguing geomorphic feature known as the 'median scarp', a NE-trending ridge across the deposit where there is an abrupt change or step-up in deposit elevation (and thickness?) of up to 30 m. The median scarp marks the lowest point in the run-out of the avalanche deposit; volcanoward from the median scarp, material flowed downslope, whereas outwards from the scarp, avalanche material flowed either along flat-ground or up-slope. Within the median scarp were some exposures of intact blocks with a preserved internal stratigraphy of the basement Salín Formation. Blocks and block margins were strongly deformed and sheared, and both small to medium-scale compressional and extensional structures could be observed cutting 'block facies' along parts of the median scarp and elsewhere in the deposit. The second stop was at a site where a thin veneer of the debris avalanche deposit mantled a pre-existing mafic lava (termed the “lions' paw” lava), which formed part of a series of mafic lavas that probably erupted along a NNE-trending fault system at the foot of the stratocone chain. Some debate centred on whether the lava unit was in situ or possibly a slide block, and the possible effects of topography on debris avalanche deposition - were megablocks trapped/deposited on the lee-side of large bed obstacles (ie. the lion's paw lava)? A continual theme for discussion over the entirety of the trip was the definition and use of 'block' and 'matrix' domains for debris avalanche deposits, and many polarised discussions resulted from outcrop interpretations as either block or matrix facies. The internal architecture of some hummocks exposed in road cuts was very complex and dominated by intensely brecciated and sheared domains surrounding more intact domains or block facies.

During our traverse across the debris avalanche deposit, we saw examples of young pumiceous deposits, either as extensive but distant plinian fall deposits blanketing the northern and eastern flanks of Socompa, or pumiceous flow deposits on top of the avalanche deposit. These deposits raised the question of whether any Plinian activity was associated with the avalanche, but the lack of age dating, and potential young age of the deposits made this a difficult question to answer. A prominent facies of the avalanche deposit is a monomict, crystal-rich and dominantly glassy dacitic lava breccia (blocks up to 10 m diameter) that in many exposures is characterised by prismatic jointing. Two alternate views on the origin of this facies were that it represented syn-avalanche lava/dome eruptions (ie. hot state of emplacement with breadcrusting formed at the site of deposition), or that it represented pre-existing dacite lava and domes entrained into the avalanche deposit (ie. cold state of emplacement, and breadcrusting developed during dome formation on the volcano summit). Outcrop discussion on the prismatic jointed glassy dacite was animated, with the aim of trying to identify criteria for a hot-state of emplacement for the dacite, and how many avalanche events may have taken place. That is, did the dacite breccia facies represent a discrete and younger avalanche event? Nevertheless, it was remarkable what a bipartite stratigraphy existed for the avalanche deposit, comprising essentially Salín Formation materials (reconstituted ignimbrite facies of van Wyk de Vries et al., 2001) and crystal-rich, glassy dacite monomict breccias as a thin veneer with little mixing between the two.

The locality from which the C
14 age of 7 Ka was obtained from a suspected blast deposit (thought to be associated with the avalanche deposit) was examined, with general consensus being that it was not a blast deposit, but likely to be either sedimentary deposits (supporting this was the partial burial of the debris avalanche deposit by sediments and the presence of basement lithic fragments derived from the west of avalanche deposit and Socompa), or possibly a fine grained ignimbrite, whose relationship to the avalanche deposit remains unknown. Consequently, the 7 Ka age for the Socompa debris avalanche deposit is questionable, and along with other observations made during the trip such as the occurrence of a thin silicic ash fall bed mantling distal sections, suggest the avalanche deposit may be significantly older than previously assumed.

Day 2 focussed on the distal margins of the avalanche deposit. The frontal scarp of the deposit is a prominent feature at least 50 m high in places, and is generally steep to repose-sloped, with evidence of some surface modification. A prominent banding parallel to the northern margins of the avalanche deposit is apparent in aerial photographs and landsat imagery. The banding corresponds to lithological variations that may in part reflect a primary stratigraphy in the source material (Salín Formation). Relatively intact blocks of basement Salín Formation were exposed at various points along the distal margin of the avalanche deposit, surrounded by large matrix domains of 'reconstituted ignimbrite/block & ash flow deposits' of the Salín Formation. Beyond the frontal toe of the avalanche deposit were salar sedimentary sequences, variably deformed in response to the avalanche deposit, with deformation varying from back-tilting to recumbent folding.

Fig2

Figure 2. View of northwestern frontal scarp of the debris avalanche deposit approximately 30 m high; note some secondary slippage and modification of the steep deposit front. The avalanche has travelled up-slope to be deposited at this point, with basement rocks gently dipping eastward.



Day 3 involved a hike along the Antofagasta-Salta railway line that largely follows the margins of the debris avalanche deposit, but in a few places provides cross-sectional exposure through the deposit. Some basal contacts could be observed along hill slopes illustrating some incorporation of substrate material into the basal portions of the deposit, as well as directional features suggesting late-stage slumping of material down the hill sides (ie. back toward the volcano).

Day 4 examined the toreva blocks and general morphology of the central amphitheatre at the foot of Socompa, where the scarp walls are 2-400 m high. The toreva blocks preserve intact sections of the edifice stratigraphy, but become progressively disjointed away from the volcano, and show incipient faulting and slumping round their margins. Some show a back tilting toward the volcano. One toreva block we examined exposed a section of a proximal fall sequence grading upward from dacitic nonwelded/sintered fallout to densely welded spatter agglutinate. The afternoon of the final day was spent examining Salín Formation deposits at La Flexura, an isolated cliffed exposure within proximal regions of the avalanche deposit. Up to 150 m thickness of a gently folded and westward plunging, interbedded sequence of dacitic block and ash flow deposits, sedimentary breccias/conglomerates and sandstones and lesser salar deposits are exposed at La Flexura. It was unclear if the La Flexura section was in situ, or if it too represents a toreva block. Nevertheless, the La Flexura section provides an important glimpse of the materials that make up ~80% of the avalanche deposit. The folding has previously been considered important in the development of the avalanche, developed in response to loading by the Socompa stratocone (e.g., van Wyk de Vries et al., 2001), but being a product of transpression along a basement fault zone could not be ruled out. Evidence for recumbent folding described previously at La Flexura, remained elusive after field checking. Nevertheless, the gentle westward dip of the La Flexura strata suggested that initial detachment may have been bedding plane-controlled, and that La Flexura represents an avalanche detachment scarp.

Fig3

Figure 3. Section through one hummock in the central part of the Socompa avalanche deposit with reconstituted Salín Formation blocks in white cut by intensely sheared and brecciated zones, and overlain by monomict glassy dacite breccia. Some blocks at the top left exhibit prismatic jointing.



In conclusion, despite four full days in the field, we literally only scratched the surface of the complexity that exists in this impressive deposit. The basic lack of petrological/geochemical and age data for Volcan Socompa and the surrounding deposits were a limiting factor in trying to better understand the development of the avalanche deposit and in distinguishing between pre-, syn- and post-avalanche dacite. Steve and Marie led a fantastic and enlightening field trip on a deposit and region that requires much more investigation. We are very much indebted to Steve for organising a logistically difficult trip and to staff from La Escondida for escorting us to and from Monturaqui and providing us with some very necessary fuel supplies on the last day so that we could return to Antofagasta!

List of Participants: Steve Self, Marie-Noelle Guilbald, Anthony Lidster, Ralf Gertisser, Kirti Sharma, Luke Wooller, Anne Jay, Michael Branney, Tiffany Barry, Claus Siebe, Marina Belousov, Sacha Belousov, Shinji Takarada, Scott Bryan.


References

Ramirez CF (1988) The geology of Socompa volcano and its debris avalanche deposit, northern Chile. MSc dissertation, Open University, Milton Keynes, UK.

van Wyk de Vries B, Self S, Francis PW, Keszthelyi L (2001) A gravitational spreading origin for the Socompa debris avalanche. Journal of Volcanology and Geothermal Research, 105: 225-247.

Wadge G, Francis PW, Ramirez CF (1995) The Socompa collapse and avalanche event. Journal of Volcanology and Geothermal Research, 66: 309-336.




Last Update: May 10, 2005