Formation of the Waiho Loop terminal moraine, New Zealand DAVID ALEXANDER, 1 * TIMOTHY DAVIES 2 and JAMES SHULMEISTER 1 1 University of Queensland, School of Geography, Planning and Environmental Management, Brisbane, St Lucia QLD 4072, Queensland, Australia 2 University of Canterbury, Department of Geological Sciences, Christchurch, New Zealand Received 5 September 2013; Revised 10 March 2014; Accepted 17 March 2014 ABSTRACT: We present an analysis of the formation and evolution of the Waiho Loop terminal moraine. Recent work has shown that the Loop comprises mainly rock avalanche material, and suggested its formation was associated with a rock-avalanche-driven glacier advance. New evidence from shallow seismic studies between the range front and the Loop suggests (i) that the presence of a basal trough critically influences glacier behaviour and moraine formation; and (ii) that the volume of the Waiho Loop is significantly greater than previously thought. A one-dimensional dynamic ice flow model is used to test two rock-avalanche-based scenarios for the formation of the Loop: first, that a rock avalanche caused a significant advance of the glacier terminus from a location within the confined mountain valley to the Loop; and second, that the rock avalanche occurred while the glacier was retreating with its terminus close to the position of the Loop. It is shown that this terminal moraine was not the result of a glacier advance. Copyright # 2014 John Wiley & Sons, Ltd. KEYWORDS: debris covered glacier; Franz Josef Glacier; numerical modelling; overdeepened trough; rock avalanche debris. Introduction Franz Josef Glacier, on the west of the South Island, New Zealand (Fig. 1), is one of the most intensively studied glaciers worldwide. The prominent, steep-sided arcuate piedmont terminal moraine of the Franz Josef Glacier, known as the Waiho Loop, has been of particular interest since it was identified as evidence of a Southern Hemisphere Youn- ger Dryas event (e.g. Mercer, 1988; Denton and Hendy, 1994). The moraine has been interpreted to indicate inter- hemispheric connectivity and synchronicity of climate change (Denton and Hendy, 1994; Broecker, 2000) or not (Singer et al., 1998). It has also been inferred to indicate a major Lateglacial cooling event in New Zealand (Anderson and Mackintosh, 2006). More recent research has indicated that the moraine may be Early Holocene rather than Lateglacial (10.5  0.2 ka; Barrows et al., 2007), although this interpretation is contro- versial (e.g. Applegate et al., 2008) and more recently the production rate for 10 Be in the Southern Hemisphere has been revised (Putnam et al., 2010b) and true ages are likely to be 15% older. More significantly, Tovar et al. (2008) and Shulmeister et al. (2009) suggested that the Loop moraine may not reflect climate variation at all. Attention was drawn to its composition dominated by rock-avalanche debris, and it was shown that it was the result of a landslide sourced from the upper Franz Josef catchment, because its lithology is dominated by angular quartzo-feldspathic sandstone found there (Tovar et al., 2008). They suggested that 0.1 km 3 of debris deposited onto the glacier ablation zone could have caused the glacier to advance 3 km from a position close to Canavan’s Knob down-valley to the Waiho Loop location (Fig. 2), where the 60-m-high terminal moraine was constructed. This was challenged by Vacco et al. (2010a, b), whose modelling suggested that the advance of the Franz Josef Glacier could not be a result of rock avalanche because there was no widespread, hummocky, glacial stagnation deposit up-valley of the Loop; however, hummocky moraine has been noted in river cuts within about 500 m of the inner margin of the Waiho Loop (Shulmeister et al., 2010a) so this ablation moraine is present but is mostly buried by subse- quent outwash deposits. Additionally, Vacco et al. (2010a) concluded that a rock avalanche of 4–20 times the estimated debris volume of Tovar et al. (2008) is required to construct the terminal moraine, as their model showed that only a small percentage (5–25%) of material from the rock ava- lanche will end up in the Loop moraine deposit. This paper presents a detailed explanation for the formation and preservation of the Waiho Loop. This uses newly acquired shallow seismic reflection data to clarify the topographic context of the Loop. We also use an ice flow model to investigate the effects of rock avalanche debris on the behaviour of the Franz Josef Glacier and the development of the Loop, and we explain how it has survived erosion by the powerful Waiho River. Approach Seismic surveys To investigate the subaerial topography into which the Waiho Loop was emplaced, shallow seismic reflection surveys were carried out to map subsurface structures upstream of, and across, the Waiho Loop (Figs 3 and 4). The surveys were carried out along one line (NW) west of the Tatare River and another (N) east of the Tatare. The latter continued on the downstream side of the Loop, but access difficulties pre- vented the former from doing so. The energy source used for acquiring seismic data was an accelerated weight drop of 25 kg, at an interval of 13.2 m. A StratVisor NZ (San Jose, CA, USA) receiver was used and data were processed using Landmark Geophysical’s ProMAX processing software. The data from these two surveys are shown in Fig. 3 and interpreted in Fig. 5. The clear reflectors in both sections, sloping upward towards the Waiho Loop, indicate the presence of an over- deepening west of the range front whose distal end coincides with the location of the Loop. This is supported by the presence of a surface outcrop of Greenland Group bedrock immediately north of the Loop (Fig. 5). The northward tectonic motion of the Australasian Plate relative to the Pacific Plate east of the range front Alpine fault takes place at about 30 m per millennium (e.g. Townend et al., 2012), so  Correspondence: David Alexander, as above. E-mail: d.alexander4@uq.edu.au Copyright # 2014 John Wiley & Sons, Ltd. JOURNAL OF QUATERNARY SCIENCE (2014) 29(4) 361–369 ISSN 0267-8179. DOI: 10.1002/jqs.2707