sign in sign up
<<
Home , Lava, Lava Fork, Nass River, Volcanism of Canada

JOURNAL OF MAPS 2020, VOL. 16, NO. 2, 363–375 https://doi.org/10.1080/17445647.2020.1758809

Science Physical of Tseax , , Canada1 Yannick Le Moignea,b, Glyn -Jonesa, Kelly Russellc and Steve Quaned aCentre for Natural Hazards Research, Department of Sciences, Simon Fraser University, Burnaby, ; bLaboratoire et Volcans, Université Clermont-Auvergne, Clermont-Ferrand, France; cDepartment of Earth, Ocean and Atmospheric Sciences, University of British Columbia, , Canada; dDepartment of , Quest University Canada, Squamish, Canada

ABSTRACT ARTICLE HISTORY Tseax volcano erupted ∼ 250 years ago in NW British Columbia, Canada producing Received 23 October 2019 deposits and flows. Field mapping has defined the stratigraphy of Tseax and the lava Revised 16 April 2020 flow morphologies. Aerial photogrammetry and bathymetry surveys were used to create a Accepted 17 April 2020 high resolution digital elevation model of the volcano to facilitate mapping and estimates KEYWORDS ∼ 6 3 of erupted material volumes. Tseax volcano ( 10.4 ± 0.7 × 10 m ) comprises an outer Tseax volcano; lava flow; breached spatter rampart and an inner conical tephra cone. Tseax is associated with a mapping; DEM; bathymetry 32 km long and 0.49 ± 0.08 km3 -to- lava flow field covering ∼ 36 km2 and divided into 4 distinct lava flows with heterogeneous surface morphologies. We present a volcanological map of Tseax volcano at a scale of 1:22,500. This will serve as supporting information for further research on the eruptive history of Tseax volcano and the lava flow field emplacement.

1. Introduction volcanic field supported by detailed descriptions of the volcanic and deposits. The aim of this Tseax volcano (pronounced ‘See-Ax’, Wil Ksi Baxhl detailed map is to provide the context to help with Mihl in the Nisga’a language) also known as Aiyansh understanding the eruptive dynamics of Tseax volcano volcano (Sutherland-Brown, 1969), Aiyansh River vol- (i.e. sequence and duration of the volcanic events). cano or Tseax River Cone (Hanson, 1924; Hickson & Edwards, 2001) is a volcano located in northwestern British Columbia (55.11085° N, 128.89944° W; see 2. Regional and local geology Main map). The volcano is associated with a 32 km long lava flow and forms the Nisga’a Memorial Lava Tseax volcano lies in the Northern Cordillera Volcanic Bed Park (Anhluut’ukwsim Laxmihl Angwinga’a- Belt (NCVP; Main map; Edwards & Russell, 1999, sanskwhl Nisga’a). The eruption of Tseax volcano, 2000). The NCVP is an approximately 1200-km-long occurred in the late eighteenth century and is the and 400-km-wide volcanic province extending from second youngest volcanic eruption in Canada after northwestern British Columbia, Canada, to eastern (Elliot et al., 1981; Hickson & Edwards, , USA. in the NCVP is attributed to 2001). This eruption is believed to be responsible for continental rifting of the due the destruction of 3 villages and the deaths of up to to changes in the relative motion of the Pacific and 2000 people (Barbeau, 1935; Higgins, 2009; Nisga’a North American plates. More than 100 Neogene to Tribal Council & B.C. Parks Committee, 1997; Suther- Quaternary volcanic centres (ranging from individual land-Brown, 1969) making it one of Canada’s deadliest tephra cones and lava flows to stratovolcanoes) with natural disasters (Hickson & Edwards, 2001; Higgins, a dominantly basaltic and high alkali compositions 2009). have been mapped (Edwards & Russell, 1999, 2000). Tseax volcano and its associated lava flow field was Tseax volcano is the southernmost volcanic centre of initially mapped by Sutherland-Brown (1969) who the NCVP located near the towns of Gitwinksihlkw provided a preliminary morphological map of the and Gitlaxt’aamiks, about 60 km north of Terrace lava flow. Recently, new mapping has been conducted and ∼1200 km north of Vancouver, British Columbia on Tseax volcano including extensive fieldwork, bathy- (Figure 1, Main map). The region surrounding Tseax metric surveys, high resolution mapping aerial pho- volcano is part of the Coast Belt plutonic complex tography and photogrammetry. Here we present a (Edwards & Russell, 1999; Woodsworth et al., 1991) 1:22,500 volcanological map (Main map) of the traversed by the which is 3rd the largest

CONTACT Yannick Le Moigne [email protected] Centre for Natural Hazards Research, Department of Earth Sciences, Simon Fraser University, Burnaby, B.C., Canada; Laboratoire Magmas et Volcans, Université Clermont-Auvergne, Clermont-Ferrand, France 1 Supplemental data for this article can be accessed https://doi.org/10.1080/17445647.2020.1758809 © 2020 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group on behalf of Journal of Maps This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 364 Y. LE MOIGNE ET AL.

Figure 1. Local geology around Tseax volcano from the regional map by the Geological Survey of Canada. Tseax lava field is shown as a single unit and represented in purple, HA. The locations of the 3 former Nisga’a villages are shown by white squares. Modified from Evenchick et al. (2008). river in British Columbia (Roden, 1967). Tributary to volumes and stratigraphy of volcanic products. The the Nass River, Tseax River flows in a narrow 15 km- volcanological map of Tseax volcano and the lava long V-shaped valley that separates the Vetter peak field was produced on ESRI ArcGIS 10.7 from field- massif from the Hazelton Mountain Ranges (Mt work observations, bathymetry, aerial photos and satel- Hoeft, Mt Phillipa; Figure 1). Tseax is located at lite imagery (Figures A.1, A.2, A3). ∼600 m above in the steep sided Crater A dataset of aerial photos was acquired (see Appen- Creek valley, a 5 km-long E-W tributary valley to dix A.1 for details) and was used to build a 3D model of Tseax River valley. The surrounding area is heavily the Tseax lava flow field using the Agisoft PhotoScan vegetated and most of the mountain slopes are covered software. It is an advanced image-based solution for by a thick and dense rainforest. creating professional quality three-dimensional con- Tseax volcano rests on the sedimentary Bowser tent from images (Agisoft LLC, 2019). PhotoScan is group (dated at -, JKBu and JKBCR on widely use in geology (e.g. Bemis et al., 2014; Gwinner Figure 1; Evenchick et al., 2008; van der Heyden et al., et al., 2000; James & Robson, 2012) to reconstruct and 2000). These rocks are grey sandstones intercalated by visualize objects or topographic surfaces in 3 dimen- dark grey and black siltstones, mudstones and conglom- sions. The 3D model supported the investigation of erates. Some plutonic rocks (ETg, Figure 1), the different surface morphologies and the calculation mostly and granodiorites belonging to the of the lava volumes. Coast Belt plutonic complex, are present near Tseax vol- The calculation of the lava flow volumes of Tseax are cano. A few occurrences of tonalites and monzogranites challenging as the base of the flow field was never have also been documented by Carter (1981). Pleisto- observed. The volumes of the lava flows were calculated cene volcanic products including basaltic and andesitic by multiplying the thickness with the area. The thick- deposits (Pb, Figure 1; Evenchick et al., 2008; van der ness of the lava flows in the Nass Valley and in the Heyden et al., 2000) are found within 50 km around Tseax River and Crater Creek Valleys were estimated Tseax volcano. They are columnar-jointed and differently. The calculation details are in Appendix basaltic assumed to be 175-50 ka old and not A.4. The pyroclastic edifices observed are tephra associated to the volcanic history of Tseax. cones and a spatter rampart. We simplified the tephra cones as truncated cones with a summit circular crater and the spatter rampart as a semi-elliptical truncated 3. Methodology cone with a semi-cylindrical inner. The calculation Fieldwork was carried out during the summers of 2016, details are in Appendix A.4. 2017 and 2019 focussed on the Tseax volcanic field The lava flow field is partially submerged at the with the goal of mapping the extent, distribution, northern end of and Melita Lake (see JOURNAL OF MAPS 365

Main map). In order to fully map the extent of the vol- of black tephra (, and ballistics). The summit canic products, we collected bathymetry data for these defines a circular crater ∼30 m deep and 80 m in diam- two using a Lowrance Elite 5Ti chart plotter with eter. The pyroclastics around the crater are oxidized a 10 Hz internal high-sensitivity GPS receiver with (brownish colours). The flanks of the cone are between DGPS and WAAS correction. 30 and 35°. On the and west flanks, several active The GIS processing of the field and remote data scarps are present. Other small concentric scarps (less allowed for the production of the main map. Figure 2 than 3 m high) are located on the south and west shows an example of the datasets used for mapping. flanks. This may indicate diverse collapse events during construction of the cone. The volume of the cone is esti- mated at 2.8±0.4×106 m3 (Equation A.1) and the 4. Description and volumes of the volume of ejected tephra is ∼6.5 × 106 m3 (Equation pyroclastic deposits and lava flow A.3). morphologies The following sections provide brief descriptions of the 4.1.3. Satellite cone and other pyroclastic features volcanic products within the Tseax volcanic field. All About 150–200 m north of Melita Lake, there is a ∼20 products from Tseax volcano (lava flows and pyroclas- m high tephra cone, called Satellite cone. This cone tic deposits) are Fe-rich, Mg-poor basanite to trachyba- has a diameter of ∼50–55 m. A small crater (4 m – salt (45.8 47.4 wt% SiO2, 14.0–16.0 wt% FeOT, deep, 7 m in diameter) occupies its summit. This – – 4.2 4.6 wt% MgO, 5.6 6.3 wt% Na2O+K2O; Gallo, small cone is asymmetric and the vent opens towards 2018; Hanson, 1924; Higgins, 2009; Le Moigne et al., the Northwest. The summit is topped by a small layer 2018; Sutherland-Brown, 1969). of oxidized agglutinated scoria and accretionnary lapilli. Descending towards the base, the cone is pro- gressively covered by black tephra. We estimate that 4.1. Tseax volcano and other pyroclastic vents the Satellite cone has a volume of 1.7 ± 0.1 × 104 m3 Tseax volcano (Wil Ksi Baxhl Mihl) has been described (Equation A.1). as two imbricated yet asymmetric volcanic cones: an Southwest of the Satellite cone, there is an alignment outer horseshoe shaped cone and an inner tephra of 3–4 small tephra cones a few metres high. This cone cone. Our mapping shows that there is, in fact, one alignment may represent a volcanic fissure oriented main tephra cone partially surrounded by a spatter ram- N10-20°. The small cones are completely covered by part (Figure 3). A Satellite cone is located north of Tseax black and iridescent tephra. Nevertheless, red oxidized volcano. A few other pyroclastic vents present near these tephra are consistently found stratigraphically below main cones are also described below (Figure 3). the black tephra. The limit of the distribution of the red oxidized tephra was not found, however, they are 4.1.1. Spatter rampart only present on and in close proximity to the Satellite The spatter rampart sits on the southern slopes of Cra- cone. This suggests that the oxidized red fountain ter Creek Valley and has a horseshoe shape that opens deposits are only linked to the building of the Satellite northwards (maximum opening is ∼280 m). This cone. Assuming this hypothesis, the volume of ejected horseshoe feature consists of two walls ∼15–25 m red tephra by the Satellite cone is ∼2×104 m3 high (the base is not visible) and ∼75–125 m long. (Equation A.3). The eastern spatter wall is oriented N170-180° and As a result, the total volume of pyroclastic deposits the south-western wall is oriented N150-160°. The (pyroclastic edifices and ejected tephra) at Tseax is spatter ramparts are not fully visible as they are par- ∼10.4 ± 0.7 × 106 m3. tially collapsed and covered by dense forest, especially along the southernmost parts. The spatter edifice is 4.2. The Tseax lavas 550–600 m wide and 360 m long maximum with a ∼320 m wide . The estimated volume is ∼1.1 The large-scale elements of the lava flow field are most ± 0.3 × 106 m3 (Equation A.2). The spatter walls consist visible from aerial and satellite images. We used field of reddish to brownish and black to grey agglutinated mapping in combination with aerial photos to describe spatter, scoriaceous spatter and scoria. The spatters and understand the structural organization of the lava are generally more coalesced at the bottom of the ram- field. part forming thick and dense layers. The outermost deposits are completely covered by forest. 4.2.1. The Tseax lava flow field The term ‘lava flow field’ is used to designate all 4.1.2. Tephra cone effusive products at Tseax. Due to its young age The main tephra cone is located towards the northern (∼250 years), the Tseax lava flow field has undergone end of the eastern spatter wall. This cone is ∼65–75 m minimal post-emplacement erosion. Nevertheless, high and ∼350–400 m in diameter, consisting mainly most of the lava flow field surface is covered by a few 366 Y. LE MOIGNE ET AL.

Figure 2. (a) Panoramic photograph of the lava flow in the Nass Valley; looking south across the Nass River above Gitwink- sihlkw. The extent of the lava flow is delimited by the white dashed line. The field of view is about 13 km. (b) Aerial photograph used for mapping. This example shows an area with multiple major breakouts on large lava-rises and rafted plates (also called platy lava, Keszthelyi et al., 2004; Stevenson et al., 2012). Lr: Lava-rises; Sl: slabby pāhoehoe. (c) Subset of the high resolution DEM (60 cm) of Tseax lava flow. The black lines show the lava flows margins. Note the interpolated holes due to water or forests. a-b topographic profile trace shown in (d). Sl-Rb: slabby-to-rubbly pāhoehoe; Hm: hummocky pāhoehoe. (d) Topographic profile across the lava flow. The lava thickness is ∼12 m according to Purssell (1993). Surface features are discussed in the text. JOURNAL OF MAPS 367

Figure 3. Images of different volcanic features in the Tseax vent area. (a) Oblique 3D view (AgiSoft PhotoScan point cloud) of Tseax volcano and surrounding area. Tseax volcano consists of a tephra cone and an outer horseshoe-shaped spatter rampart. The Satellite cone is located north of Lower Melita Lake. The symbol ‘∢’ represents the point of view of the photograph in (d). (b) Tephra cone crater (∼80 m in diameter and ∼30 m deep). (c) The Satellite cone is about 20 m high. (d) View of the eastern spatter wall from the tephra cone. (e) Close up view of the eastern spatter wall (∼9 m high). centimetres of green to yellow lichens and mosses. Fur- outline of the Tseax lava flow field is nevertheless easily thermore, large portions of the lava flows are already identified from field observations and aerial or satellite covered by dense rainforest, e.g. the vast majority of imagery. the lava in Tseax River Valley. Moreover, in some The lava flow field originates from the vent area in places the lava underlies rivers, streams (e.g. Tseax Crater Creek Valley, where it forms a dam (made of River, Vetter Creek) and a few small lakes or ponds a few small lava islands) separating Lower and Upper such as in the Tseax River Valley. The Melita Lake (T’aam Baxhl Mihl). 368 Y. LE MOIGNE ET AL.

Figure 4. (a) Overview map of Tseax lava field. Bathymetry of (b) Lower Melita Lake and (c) Lava Lake. Bathymetry on Lower Melita Lake does not show any particular volcanic or lava features. On the northern shores of Lava Lake, we can identify the termination of Flow 3 at ∼7 m depth and the termination of the ow field at ∼27 m depth. Coordinate system: WGS 1984 / UTM Zone 9N (metres).

Investigation of the Upper Melita lake by canoe River. A 4–7 m high lava cliff about ∼5 km in length and paddle board has shown that the lava does not is one of the most notable and spectacular features of go more than ∼150 m to the east. Bathymetry the Tseax lava flow field, forcing the Nass River to shows that Lower Melita lake is very shallow (∼2m form a narrow canyon near Gitwinksihlkw (Figure 2 deep on average), with the deepest part (reaching (a)). The western margin of the lava field is entirely 3.5 m) found near north-western shores (Figure 4 covered by rainforest, but easily mapped in the field (b)). The chaotic floor of the lake suggests that it con- and from aerial photos. The southern margins in the sists of lava. The lava-filled Crater Creek Valley is Nass Valley corresponds to Ksi Ts’oohl Ts’ap creek about 5 km long on a slope of ∼4°. Based on the topo- and marshlands. A significant part of the lava flow graphic slopes on both sides of Crater Creek Valley, field in the Nass Valley is covered by recent forest we estimate that the average depth of the lava in growth in the area where Vetter Creek disappears this valley is 30.5 ± 5.5 m. At the bottom of Crater beneath the lava. The depth of the lava in the Nass Creek Valley, the lava flow debouches in the Tseax Valley is difficult to estimate, however, geotechnical River Valley. There, the lava creates a significant drilling indicates a lava depth of 12 m near the road dam on Lava lake. Bathymetry measurements per- bridge to Gitwinksihlkw (Figure 2(a); Purssell, 1993). formed on Lava lake show that the lava ends 250 m Therefore, we consider an average thickness of 12 ± south of the shore at a depth of about 25–30 m 2 m for the lava in the Nass Valley as some parts of (Figure 4(c)). The lava descends northwards along the flow may have been emplaced on irregular topo- the narrow Tseax River Valley (average slope of graphy or may have undergone different magnitude ∼1.5° on 15 km), reaching about 22 km in length of inflations. with an estimated average lava depth of 21 ± 3 In total, the Tseax lava flow field covers ∼36 km2. m. The lava then spreads out towards the southwest Using our thickness estimates in Crater Creek, Tseax in the Nass Valley flood plain. There the lava forms River and Nass Valleys, we calculate a volume of a ∼11 km long and up to 3.8 km wide lava plain 0.49 ± 0.08 km3 for the entire lava flow field, most of (slope of 0.1–0.2°). Eastwards, the lava flow is bor- the volume being in the Nass Valley. This correlates dered by Tseax River and Spencer Lake. The Nass with the 0.5 km3 previously estimated by Hanson River marks the northern margin of the lava flow (1924) and Higgins (2009) and the 0.455 km3 esti- with no lava found on the northern side of the Nass mated by Sutherland-Brown (1969). JOURNAL OF MAPS 369

4.2.2. The 4 Tseax lava flows 4.3. Subaqueous lavas – bathymetry We define ‘lava flow’ as an individual outpouring of Melita Lake (T’aam Baxhl Mihl, Figure 4(b)), near lava from the vent, likely corresponding to an indi- Tseax volcano, was formed after damming of Crater vidual eruptive event. The margins of a lava flow Creek by the lava flow. Melita Lake consist of two are generally steep and easily identifiable in the lakes, Lower and Upper Melita Lake (∼0.05 km2 and field. Despite being emplaced during a single erup- ∼0.2 km2, respectively). Lower Melita lake is separated tion, the Tseax lava flow field can be divided into 4 from Upper Melita Lake by a series of small and heavily distinct lava flows, named Flow 1, Flow 2, Flow 3 vegetated lava islands. The shore of Lower Melita Lake and Flow 4 (see Main map). The main geometrical consists of large blocks of pāhoehoe, in some places characteristics of the lava flows are reported in completely covered by clay and silt. Field observations Table 1. suggest all the lava exposed along the shoreline and on Flow 1 is the longest, most voluminous and strati- the islands separating Lower and Upper Melita Lakes graphically earliest lava flow. It is a -poor are from the same flow unit. Petrographic analysis of pāhoehoe lava 31.6 km in length with an average a lava block sampled along the northern shore of thickness of 12 ± 2 m in the Nass Valley. Flow 1 Melita Lake shows a phenocryst-poor lava, similar to forms the entire lava plain in the Nass Valley, 2.8 Flow 1 and Flow 2. The deepest part of the lake is km wide on average and ∼12 km long. We suggest found near the western shores (∼3.5 m deep). The east- that the thickness of Flow 1 increases to 14 ± 2 m in ern part of the lake is shallower although two deep Tseax River and Crater Creek Valleys due to narrow- spots are located close to the central and northeast ing. This results in a volume of 0.41 ± 0.07 km3. Flow shores. The bathymetry results show a relatively chao- 1 thus accounts for ∼84% of the total flow field tic floor with no particular volcanic features. Collapsed volume. lava blocks and agglutinated tree logs are present along Flow 2 is a 21.2 km long phenocryst-poor pāhoehoe the western and northern shores of the lake. lava flow. Located along Crater Creek Valley, it ends Lava Lake (Sii Tax, Figure 4(c)) is located in Tseax near the mouth of Tseax River Valley where it forms River Valley at the bottom of Crater Creek Valley. The two distinct ∼2 km long lava lobes. We estimate a Tseax River debouches from the northwestern extre- thickness of 7 ± 1 m. Despite the length, we estimate mity of the lake. While Lava Lake already existed before a volume of 0.06 ± 0.01 km3, representing ∼13% of the eruption of Tseax (D. Nyce, personal communi- the total flow field volume. cation), it’s depth increased due to damming of Flow 3 is a 7.2 km long phenocryst-rich ‘a‘ā lava Tseax River by Flows 1, 2 and 3. Pressure of flow descending from the volcano in Crater Creek Val- Flow 3 form the northern shores of Lava Lake. We ley and terminating in Tseax River Valley near Ross estimate that Flow 3 ends ∼20–50 m from the lake. We estimate an average thickness of 3 ± 1 m in northern shore at a depth of about 7 m. The flow Tseax River Valley and 4.5 ± 1 m in Crater Creek Val- field terminates ∼250–270 m from the northern shores ley. Therefore, the volume of this flow is 0.009 ± at ∼27 m depth. 0.002 km3, accounting for < 2% of the total flow field volume. Flow 4 is a phenocryst-rich ‘a‘ā lava flow and the 4.4. Lava flow surface morphologies stratigraphically uppermost flow of the Tseax lava flow field. It is 5.3 km long and located only in Cra- Lavas of the Tseax flow field can be classified according ter Creek Valley. This ‘a‘ā flow ends where Crater to their surface morphologies. From field observations, Creek Valley debouches in Tseax River Valley but we distinguished pāhoehoe and ‘a‘ā flows with numer- does not reach Lava Lake. We estimate a thickness ous distinctive surface morphologies. In this section, of 4.5 ± 1 m which gives a volume of 0.006 ± we present short definitions of these surface morpho- 0.001 km3. This corresponds to ∼1% of the total logical features as well as their location. We refer the flow field volume. reader to the Main map for a better understanding. Figure 5 shows photographs of representative surface morphologies seen on the Tseax lava flow field. Table 1. Type (Pāhoehoe or ‘A‘ā), length, surface area and The main categories of surface morphologies are volume of the Tseax lava flow field and the 4 identified ‘a‘ā (Flow 3 and Flow 4) and pāhoehoe lava flows individual lava flows. (Flow 1 and Flow 2, see Main map). We identified Length Surface Area Volume hummocky (Hm), lava-rises (Lr), slabby (Sl), hum- Flow # Type (km) (km2) (km3) Error (km3) mocky-to-lava-rises (Hm-Lr) and slabby-to-rubbly Flow 1 Pāhoehoe 31.6 26.97 0.41 0.07 (Sl-Rb)pāhoehoe surfaces. The general characteristics Flow 2 Pāhoehoe 21.2 6.74 0.06 0.01 Flow 3 ‘A‘ā 7.2 1.11 0.009 0.002 of the lavas are described in details in Appendix Flow 4 ‘A‘ā 5.3 1.22 0.006 0.001 B. The area occupied by these surface morphologies Total 31.6 34.06 0.49 0.08 are indicated in Table 2. 370 Y. LE MOIGNE ET AL.

Figure 5. Representative photographs of different surface morphologies. Note that the lava is covered by green to yellow moss and/or lichen. (a) ‘A‘ā lava flow at the bottom of Crater Creek Valley. (b)-(h) These photographs were taken on the Nass Valley lava plain. (b) Hummocky pāhoehoe (Hm). (c) A large lava-rise area (Lr). (d) Hummocky-to-lava-rises pāhoehoe (Hm-Lr). (e): Slabby pāhoehoe (Sl). (f) Slabby-to-rubbly pāhoehoe (Sl-Rb). (g) Lava-rise pit ∼2 m deep and ∼15 m in diameter. (h) A surface plate; note the disrupted slabs surrounding the plate. JOURNAL OF MAPS 371

Table 2. Area occupied by the different pāhoehoe surface was analysed in Reefmaster underwater mapping soft- morphologies in the Nass Valley. ware. Bathymetry and geological maps were produced 2 Surface morphology Area (km ) Area (%) using ESRI’s ArcGIS 10.7. The maps were edited and Hummocky Pāhoehoe (Hm) 5.5 22 structured using Adobe Illustrator CC 2015. Lava-rises (Lr) 4.6 18 Hummocky-to-lava-rises Pāhoehoe (Hm-Lr) 2.5 10 Slabby Pāhoehoe (Sl) 8.2 33 Slabby-to-rubbly Pāhoehoe (Sl-Rb) 3.7 15 Acknowledgements Rafted plates 0.5 2 Many thanks to Dr. Deana Nyce of the Wilp Wilxo’oskwhl Nisga’a Institute (WWNI) for her continued and enthusias- ffi ’ 5. Conclusions tic support. Mansel Gri ths and Harry Nyce (Nisga a Lisims Government) generously shared their extensive knowledge A 1:22,500 map of Tseax volcano and its immediate of Tseax and the Nisga’a Lava Beds. We are grateful to surroundings is presented here and provides the first Irene Squires (WWNI) for her generous hospitality, to Rachel Warwick, Rose Gallo, Alex Wilson and Sarah Aufrère detailed and comprehensive description of the volcanic for assistance during field work, and to Margaret Vanderberg products. The mapping and characterization of the vol- for her assistance in providing historical information and canic products was conducted over three field seasons field notes. We also thank G. B. M. Pedersen, R. Pierosan with the help of high-resolution aerial photographs, and A. Labetski for their constructive reviews. This work ’ photogrammetry and bathymetric surveys. The map was supported by the WWNI, the Nisga a Lisims Govern- ment and an NSERC Discovery grant to G. Williams- is designed to document and characterize the lava fea- fl Jones. This research was carried out as part of B.C. Parks / tures and the lava ow surface morphologies. NLG permit 108312. Tseax volcano consists of two imbricated volcanic structures (a central tephra cone and an outer spatter rampart) located at about 600 m a.s.l. in Crater Creek Disclosure statement Valley. The pyroclastic products cover a surface of > No potential conflict of interest was reported by the author 10 km2. Tseax volcano is associated with a 0.49 ± (s). 0.08 km3 and 32 km long lava flow field. We identified 4lavaflows, likely corresponding to 4 different eruptive events. Stratigraphically, we have from bottom to top: 2 Funding long and voluminous pāhoehoe lava flows (Flow 1 and This work was supported by NSERC Discovery [grant num- Flow 2, 0.41 ± 0.07 and 0.06 ± 0.01 km3,respectively) ber RGPIN-2016-03953]. and 2 shorter ‘a‘ā lava flows (Flow 3 and Flow 4, 0.009 3 ā ± 0.002 and 0.006 ± 0.001 km , respectively). The p hoe- References hoe flows show several distinctive surface morphologies namely: hummocky, lava-rises, slabby, hummocky-to- Agisoft LLC. (2019). Agisoft Metashape User Manual: lava-rises, slabby-to-rubbly pāhoehoe and rafted plates. Professional Edition, Version 1.5. Barbeau, M. (1935). Volcanoes on the Nass. Canadian Usually, we observe a complete transition between Geographical Journal, 10(5), 215–225. these surface morphologies in a lava lobe unit within a Bemis, S. P., Micklethwaite, S., Turner, D., James, M. R., flow. The main sharp disturbances are marked by Akciz, S., Thiele, S. T., & Bangash, H. A. (2014). major breakouts on the large lava-rises that disrupted Ground-based and UAV-based photogrammetry: A the smooth upper into slabby pāhoehoe. multi-scale, high-resolution mapping tool for structural fi geology and paleoseismology. Journal of Structural The volcanological map for the Tseax volcanic eld Geology, 69(PA), 163–178. https://doi.org/10.1016/j.jsg. represents fundamental data required to understand 2014.10.007 the sequence and duration of events that ultimately Carter, N. (1981). and Molybdenum depos- led to the Nisga’a fatalities. Furthermore, the described its west-central British Columbia (Tech. Rep.). Ministry of lava flow surface morphologies are critical information Energy, Mines and Petroleum Resources. for quantification of eruption parameters, key for Edwards, B. R., & Russell, J. K. (1999). Northern Cordilleran fl volcanic province: A northern Basin and range? Geology, numerical modelling of lava ow emplacement. This 27(3), 243–246. https://doi.org/10.1130/0091-7613 work is also an important contribution towards hazard (1999)027<0243:NCVPAN>2.3.CO;2 assessment as forestry and fishery industries, tourist Edwards, B. R., & Russell, J. K. (2000). Distribution, , attractions and ∼1000 people are located in close and origin of Neogene-quaternary in the proximity to the volcano. northern Cordilleran volcanic province, Canada. Bulletin of the Geological Society of America, 112(8), 1280–1295. https://doi.org/10.1130/0016-7606(2000)112<1280: DNAOON>2.0.CO;2 Software Elliot, R. L., Koch, R. D., & Robinson, S. W. (1981). Age of fl fi fl fi ows in the Blue River Valley, Brad eld Canal The Digital Elevation Model of the Tseax lava ow eld Quadrangle. U.S. Geological Survey of Alaska, was created with AgiSoft PhotoScan Professional Accomplishments during 1979. U.S. Geological Survey 1.2.4.2339 and CloudCompare v2.8. Bathymetric data Circular, 823-B, B115–B116. 372 Y. LE MOIGNE ET AL.

Evenchick, C. A., Mustard, P. S., Greig, C. J., McMechan, M. Geophysics, Geosystems, 5(11). https://doi.org/10.1029/ E., Ritcey, M. E., Smith, D. H., & Ferri, F. (2008). Geology, 2004GC000758 Nass River, British Columbia. https://doi.org/10.4095/ Le Moigne, Y., Williams-Jones, G., Kelfoun, K., Labazuy, P., 225792 Vigouroux, N., & Russell, K. (2018). Investigating Gallo, R. (2018). History and dynamics of explosive volcanism Canada’s deadliest volcanic eruption, from field obser- at Tseax Cone, British Columbia (Undergraduate vations to lava flow modeling. In IAVCEI cities on volca- Research). Retrieved April 28, 2020, from https://open. noes 10. Naples, . library.ubc.ca/collections/undergraduateresearch/52966/ Nisga’a Tribal Council, & B.C. Parks Committee. (1997). items/1.0366164 Anhluut’ukwsim Laxmihl Angwinga’asanskwhl Nisga’a: Gwinner, K., Hauber, E., Jaumann, R., & Neukum, G. (2000). Nisga’a Memorial Lava Bed Park Master Plan (Tech. Rep.). High-resolution, digital photogrammetric mapping: A Purssell, T. (1993). Gitwinksihlkw Bridge Foundation Design tool for earth science. Eos, Transactions American Summary Report (Tech. Rep.). Ministry of Transportation Geophysical Union, 81(44), 513–520. https://doi.org/10. and Highways, North West Region, Province of British 1029/00EO00364 Columbia, File M52-45-2998. Hanson, G. (1924). Reconnaissance between Roden, G. I. (1967). On river discharge into the northeastern and Stewart, B.C. (Tech. Rep. No. 2031). Canada Pacific Ocean and the Bering Sea. Journal of Geophysical Department of Mines, Geological Survey. Research, 72(22), 5613–5629. https://doi.org/10.1029/ Hickson, C., & Edwards, B. (2001). Volcanoes and volcanic jz072i022p05613 hazards. In G. Brooks (Ed.), A synthesis of geological Stevenson, J. A., Mitchell, N. C., Cassidy, M., & Pinkerton, H. hazards in Canada (pp. 145-181). Geological Survey of (2012). Widespread inflation and drainage of a pahoehoe Canada, Bulletin 548. https://doi.org/10.4324/ flow field: The Nesjahraun, Thorngingvellir, Iceland. 9781315830889 Bulletin of Volcanology, 74(1), 15–31. https://doi.org/10. Higgins, M. D. (2009). The Cascadia megathrust 1007/s00445-011-0482-z of 1700 may have rejuvenated an isolated basalt volcano in Sutherland-Brown, A. (1969). Aiyansh lava flow, British : Age and petrographic evidence. Journal Columbia. Canadian Journal of Earth Sciences, 6, 1460– of Volcanology and Geothermal Research, 179(1-2), 149– 1468. https://doi.org/10.1139/e69-149 156. https://doi.org/10.1016/j.jvolgeores.2008.10.016 van der Heyden, P., Woodsworth, G. J., & Snyder, L. D. James, M. R., & Robson, S. (2012). Straightforward recon- (2000). Reconnaissance geological mapping in southwest struction of 3D surfaces and topography with a camera: Nass River map area, British Columbia. Current Accuracy and geoscience application. Journal of Research 2000-A6 (Tech.Rep.). Geophysical Research: Earth Surface, 117(3), 1–17. Woodsworth, G., Anderson, R., & Armstrong, R. (1991). https://doi.org/10.1029/2011JF002289 Plutonic Regimes. In H. Gabrielse, & C. Yorath (Eds.), Keszthelyi, L., Thordarson, T., McEwen, A., Haack, H., Geology of the cordilleran orogen in Canada (pp. 491– Guilbaud, M. N., Self, S., & Rossi, M. J. (2004). Icelandic 531). Geological Survey of Canada, Geology of Canada analogs to Martian flood lavas. Geochemistry, No. 4. JOURNAL OF MAPS 373

Appendices

Figure A.1. (a) Locations of the aerial photographic datasets used to construct the DEM. The yellow star is the location of the only geotechnical drilling on the lava ow field. (b) Acquisition of a GCP on the lava flow. (c) Ultralight aircraft used to take aerial photo- graphs above Crater Creek Valley and Tseax volcano. (d) Locations of the flight lines and bathymetry surveys. 374 Y. LE MOIGNE ET AL.

Figure A.2. Topographic profiles perpendicular to the lava ow along Crater Creek and Tseax River valleys made to estimate the ow thickness. The map shows the location of the profiles. JOURNAL OF MAPS 375

Figure A.3. Schematic of the tephra cones and spatter rampart. Supplementary Materials

A- Methodology Fieldwork was carried out during the summers of 2016, 2017 and 2019 focussed on the Tseax with the goal of mapping the extent, distribution, volumes and stratigraphy of volcanic products. The volcanological map of Tseax volcano and the lava field was produced on ESRI ArcGIS 10.7 from fieldwork observations, bathymetry, aerial photos and satellite imagery (Figures A.1, A.2).

A.1- Aerial Photograph Datasets The location of the aerial photographs used are shown on Figure A.1. We used an ultralight aircraft to acquire high resolution aerial photographs of Tseax volcano and Crater Creek Valley (Figure A.1(c, d)). The aircraft flew at ~ 500 m above the ground at a constant speed of 110 km/h. RGB photographs were taken using a Nikon D800 camera fixed to one of the aircraft's wings. We used a 28 mm lens on the camera. The horizontal FOV and angular FOV were 65 and 46, respectively. The resolution of the photographs are 7960 × 4912 pixels and each photo thus captured a field of view of ~ 643 × 429 m corresponding to a ground resolution of ~ 0.08 m. Nevertheless, the aerial photographs are blurry due to vibrations of the wing to which the camera was fixed. The minimum size of the identifiable object on the ground is thus ~ 0.5 m.

The camera was set to take a photo every 2 seconds. 1044 photos were taken along 7 flight lines above the lava flow in Crater Creek Valley and Tseax Volcano. The forward overlap between two consecutive photos is 82 % and the side overlap is 83 % between photos of two adjacent flight lines. Due to excessively high wind speeds, we were unable to acquire aerial photographs from the ultralight of Tseax River and Nass Valleys. However, 23 high resolution digital air photos (in grey- scale) were acquired from the Government of British Columbia (Figure. A.1(a)). These photos were taken on August 12, 2001, from an airplane flying at ~ 5,800 m a.s.l. The focal length used was 153 mm and the scale is 1:30,000 with a resolution of 19,225 ~ 19,218 pixels. The ground resolution is ~ 0.46 m and each photo has a field of view of 8.84 × 8.84 km. Between two consecutive photos, there is a forward overlap of 80 % and a side overlap of 65 %. The 2019 Digital Globe satellite imagery was also used to help map the lava flow surface morphologies on ESRI ArcGIS 10.7.

A.2- DEM Processing During the 2016 field season, we deployed 18 Ground Control Points (GCPs); 12 of which were set up on the lava flow (Figure A.1(b)). The GCPs consisted of large rectangular white cotton sheets about 1.6 × 2.0 m marked with a large black cross in the middle (about 0.5 m in length) and placed directly on the lava flow surface. We also used 6 obvious road markers or concrete blocks as GCPs. All the GCPs were measured using a differential GPS (Leica GPS System 500; Figure 1(b)). A base station was located in Gitwinksihlkw and ran continuously for 65 h 44 min in order to correct the position of the GCP stations. Each GCP station was measured between 25 and 50 minutes, depending on their location with respect to the base station (the farther from the base, the longer the measurement). GCP positions were then corrected using the Canadian Spatial Reference System Precise Point Positioning (CSRS-PPP; https://webapp.geod.nrcan.gc.ca/geod/tools- outils/ppp.php?locale=en) tool provided by 's Canadian Geodetic Survey (NRC-CGS).

A 3D model of the Tseax lava flow field was built using the Agisoft PhotoScan software. It is an advanced image-based solution for creating professional quality three-dimensional content from images (Agisoft LLC, 2019). PhotoScan is widely use in geology (e.g., Bemis et al., 2014; Gwinner, Hauber, Jaumann, & Neukum, 2000; James & Robson, 2012) to reconstruct and visualize objects or topographic surfaces in 3 dimensions. All aerial images (the 1044 RGB photographs taken from the ultralight aircraft and the 23 grey-scale digital air photos) were loaded in PhotoScan in two separate chunks as PhotoScan is unable to merge colour and grey-scale images. As such, the two chunks were prepared separately. All the photos were georeferenced with the GCPs. From the chunk containing the 1044 RGB aerial photos, we obtained a sparse cloud of 1.2 × 105 points and a dense cloud of 5.2 × 108 points. In the chunks containing the grey-scale aerial photos, we obtained a sparse cloud of 3.4 × 104 points and a dense cloud of 1.5 × 108 points. The 2 dense point clouds were subsequently exported in CloudCompare software (Girardeau- Montaut, Telecom Paris Tech, & EDF, 2017) and merged. The dense point cloud was then exported back into PhotoScan to generate the 3D meshes and DEM. The final DEM has a resolution of ~ 0.6 m (Figure 2(c)). For mapping design, we completed our DEM with the SRTM DEM (30 m resolution).

A.3- Bathymetry The lava flow field is partially submerged at the northern end of Lava Lake and Melita Lake (see Main map). In order to fully map the extent of the volcanic products, we collected bathymetry data for these two lakes (Figure A.1(a, d)) using a Lowrance Elite 5Ti chart plotter with a 10 Hz internal high-sensitivity GPS receiver with DGPS and WAAS correction. The data collection unit is rigged to the deck of a Kahuna Bruddah 10’6” long in inflatable paddle board with a Lowrance StructureScan HD transducer mounted to the underwater tail fin. Sonar data collection for these lakes was done in high CHIRP mode at a frequency of 200 kHz. To collect the sonar data, we paddled across the lake in a systematic manner using the track on the chart plotter visual display coupled with visual cues on the shoreline. Track spacing and geometry is dependent on the size and shape of the water body. For Lava lake, most tracks were spaced approximately 10 m apart with a maximum spacing of 150 m in large, open water sections. Due to its smaller size, Melita lake had an average track spacing of ~ 10 m and a maximum track spacing of 25 m. Bathymetric data is analyzed by importing .sl2 format latitude (X), longitude (Y) and depth (Z) files from the chart plotter into Reefmaster underwater mapping software. Within Reefmaster, bathymetric contours are interpolated using the maximum track spacing during data collection (from 0 to 250 m) in conjunction with a smoothing factor (from 1 to 20). We interpolated contours for Lava lake with a maximum interpolation distance of 150 m and a smoothing factor of 5 to create 5 m (major) and 1 m (minor) bathymetric contours. Melita lake had a maximum interpolation distance of 25 m and a smoothing factor of 10 to create 0.5 m (major) and 0.125 m (minor) contours. Contour data from Reefmaster are exported as Shapefiles (.shp) for use in ArcGIS.

A.4 Volume Calculations

A.4.1- Lava Flows The calculation of the lava flow volumes of Tseax are challenging as the base of the flow field was never observed. The volumes of the lava flows were calculated multiplying the thickness with the area. The thickness of the lava flows in the Nass Valley and in the Tseax River and Crater Creek Valleys were estimated differently.

We only have a single geotechnical drilling on the entire lava flow field. It is located in the Nass Valley near the bridge that leads to the town of Gitwinksihlkw (Figure A.1(a)) and was provided by the Ministry of Transportation and Highways of British Columbia (Purssell, 1993). We assumed that the lava depth indicated by this drilling is representative of the flow thickness in the Nass Valley. We decided to add an error of ± 2 m in order to account for lavas that may have undergone different inflation processes.

Along Crater Creek and Tseax River valleys, we drawn 16 topographic profiles perpendicular to the flow (Figure A.2). In order to determine the lava thickness, we reconstructed a possible pre- eruptive topography by extending the slopes on both sides of the valleys. The volumes in these valleys were bracketed between the minimum and maximum estimated depth. The total lava flow field volume was calculated by adding the lava volumes in the Nass, Tseax River and Crater Creek valleys.

A.4.2- Pyroclastic Deposits and Ejected Tephra The pyroclastic edifices observed are tephra cones and a spatter rampart. We simplified the tephra cones as truncated cones with a summit circular crater and the spatter rampart as a semi-elliptical truncated cone with a semi-cylindrical inner. The volumes of the tephra cones, Vc, and the spatter rampart, Vs, are calculated by (each term is defined on Figure A.3):

1 = [ ( + + ) ] (A.1) 3 2 2 2 𝑉𝑉𝑐𝑐 𝜋𝜋 𝐻𝐻𝑐𝑐 𝑅𝑅𝑐𝑐 𝑅𝑅𝑐𝑐𝑟𝑟𝑐𝑐 𝑟𝑟𝑐𝑐 − ℎ𝑐𝑐𝑟𝑟𝑐𝑐 1 = ( + + ) (A.2) 2 3 𝑠𝑠2 𝑠𝑠 2 2 𝑠𝑠 𝜋𝜋𝑅𝑅 𝐻𝐻 𝑠𝑠1 𝑠𝑠1 𝑠𝑠1 𝑠𝑠2 𝑠𝑠 𝑠𝑠1 𝑠𝑠2 𝑉𝑉 � 𝑠𝑠1 𝑅𝑅 𝑅𝑅 𝑟𝑟 𝑟𝑟 − 𝜋𝜋𝐻𝐻 𝑟𝑟 𝑟𝑟 � The volume of ejected pyroclastic𝑅𝑅 materials from a tephra cone, Ve, is related to the diameter, 2hc, of the main crater by the relation established by Sato & Taniguchi (1997):

= 0.11 . (A.3) 0 42 𝑒𝑒 𝐷𝐷 𝑉𝑉 References

Agisoft LLC. (2019). Agisoft Metashape User Manual: Professional Edition, Version 1.5.

Bemis, S. P., Micklethwaite, S., Turner, D., James, M. R., Akciz, S., T. Thiele, S., & Bangash, H. A. (2014). Ground-based and UAV-Based photogrammetry: A multi-scale, high-resolution mapping tool for structural geology and paleoseismology. Journal of Structural Geology, 69 (PA), 163{178. doi: 10.1016/j.jsg.2014.10.007

Girardeau-Montaut, D., Telecom Paris Tech, & EDF. (2017). Cloud Compare User manual.

Gwinner, K., Hauber, E., Jaumann, R., & Neukum, G. (2000). High-resolution, digital photogrammetric mapping: A tool for earth science. Eos, Transactions American Geophysical Union, 81 (44), 513-520. doi: 10.1029/00EO00364

Purssell, T. (1993). Gitwinksihlkw Bridge Foundation Design Summary Report (Tech. Rep.). Ministry of Transportation and Highways, North West Region, Province of British Columbia, File M52-45-2998.

Sato, H., & Taniguchi, H. (1997). Relationship between crater size and ejecta volume of recent magmatic and phreato-magmatic eruptions: Implications for energy partitioning. Geophysical Research Letters, 24 (3), 205-208. doi: 10.1029/96GL04004

B- Lava Flow Surface Morphologies Here, we present short definitions of these surface morphological features as well as their location. We refer the reader to the Main map and to the Figure 5 for a better understanding.

‘A‘ā lava flows on the Tseax lava flow field are Flow 3 and Flow 4 (see Main map). They are the stratigraphically uppermost and least voluminous flows (Table 1). ‘A‘ā lava flows present brecciated, rough and jagged surfaces made of loose angular to rounded lava fragments called clinkers (Figure 5(a); Harris Rowland, Villeneuve, & Thordarson, 2017; Lipman & Banks, 1987; Macdonald, 1953; Peterson & Tilling, 1980). The majority of clinkers are 0.1 - 0.5 m in diameter, with the largest up to 2 m in diameter. The clinkers form an upper crust ~ 1 to 2 m thick. When exposed, the core of ‘a‘ā lava flows are dense. The base of ‘a‘ā lava flows was never visible.

The terminus of Flow 3 displays a series of pressures ridges oriented perpendicularly to flow direction. These pressure ridges are 1-4 m high with a regular spacing of 15-20 m. The highest pressure ridges are found between 250 and 900 m south of Ross Lake. Flow 4 can be followed from Tseax cone to the mouth of Crater Creek Valley and shows multiple frontal margins of individual lava lobes, notably in the higher part of Crater Creek Valley. Channel levees a few metres high are sometimes preserved. A meandering network of channels can be seen in aerial photographs. A few pressure ridges are located near the termination of Flow 4.

Pāhoehoe lava flows. Pāhoehoe lavas are often described as having smooth and regular surfaces compared to rough ‘a‘ā lava flows (Harris et al., 2017; Macdonald, 1953; Self, Keszthelyi, & Throdarson, 1998). However, various surface morphologies on pāhoehoe lava flows can be differentiated implying different mechanisms of emplacement and flow regimes. The pāhoehoe flows at Tseax are Flow 1 and Flow 2. They are the longest, most voluminous and stratigraphically lowest of the lava flow field (see Main map, Table 1). We identified hummocky (Hm), lava-rises (Lr), slabby (Sl), hummocky-to-lava-rises (Hm-Lr) and slabby-to-rubbly (Sl-Rb) pāhoehoe surfaces. The general characteristics of the lavas are described below but we note that there is a continuum between these types. Furthermore, the characteristics of these pāhoehoe surface types may vary within the flow field.

Hummocky Pāhoehoe (Hm). We define hummocky pāhoehoe as very irregular pāhoehoe lava flow surfaces made of numerous hummocks called tumuli (Figure 5(b); Hon, Kauahikaua, Denlinger, & Mackay, 1994; Self et al., 1998; Swanson, 1973). Tumuli are ~ 10-200 m long, sinuous and inter-crossed mounds of 1-4 m high pāhoehoe lava flow lobes that failed to coalesce (e.g., aborted inflation) or that have undergone differential inflation processes (Hon et al.,1994). Due to inflation, most of the tumuli have a central and/or lateral cracks, called lava-inflation clefts (Walker, 1991), up to 2.5 m deep and 1.5 m large. The slopes of the tumuli flanks are steep at their base (sometimes more than 45°) and progressively flatten towards their summit. Some well developed tumuli have a flat topped surface. At Tseax, hummocky pāhoehoe are mainly found along the northern and western margins of Flow 1 in the Nass Valley.

Lava-rises (Lr). Lava-rises are large surfaces (several km2) of flat, almost perfectly horizontal pāhoehoe lava plates (Figure 5(c)). Lava-rises are the results of a fully developed inflated pāhoehoe flow and are only poorly disturbed by superficial cracks and collapsed pits called lava-rise pits (Walker, 1991). Superficial cracks are 10-400 m long, 0.1-2 m deep and up to 0.5 m wide linear features probably formed during cooling. Lava-rises are only present on Flow 1 and mainly found on the western part of the Nass Valley lava plain. Some lava-rise areas can be > 1.5 km wide without significant change in surface morphology. The total area of lava-rises is difficult to calculate as they generally progressively develop into slabby or hummocky pāhoehoe.

Hummocky-to-Lava-rise pāhoehoe (Hm-Lr). Hummocky-to-lava-rise pāhoehoe is the transition region where both hummocky and lava-rise coexist in nearly the same proportion (Figure 5(d)). The hummocks are smaller than in hummocky pāhoehoe. Hummocky-to-lava-rise pāhoehoe is found in the Nass Valley, mostly in the upper central part of the lava plain.

Slabby Pāhoehoe (Sl). Slabby pāhoehoe is formed by repetitive surface deformation of a large inflated lava flow via fragmentation of the upper crust (Guilbaud, Self, Thordarson, & Blake, 2005; Macdonald, 1953; Peterson & Tilling, 1980; Swanson, 1973). Slabby pāhoehoe is characterized by chaotic surfaces dominated by broken pāhoehoe lava crusts, called pāhoehoe slabs (Figure 5(e)). The slabs are dislocated pāhoehoe crust of 1-10 m2, tabular to curved, tilted up and sometimes piled and turned upwards. On average, individual slab thickness varies between 0.1-1 m. Some areas that underwent more compression are characterized by smaller and thinner slabs with occasionally the presence of pressure ridges. Flow 2 appears to consist essentially of slabby pāhoehoe. On Flow 1 in the Nass Valley lava plain, slabby pāhoehoe is also the principal surface morphology. Slabby pāhoehoe are mainly found in the central and western part of the Nass Valley lava plain.

Slabby-to-Rubbly Pāhoehoe (Sl-Rb). We define slabby-to-rubbly pāhoehoe as a lava flow surface morphology composed of broken pāhoehoe slabs (~ 60 %) with a significant amount of lava rubble (~ 40 %; Figure 5(f)). The name rubbly pāhoehoe has been proposed by Keszthelyi & Thordarson (2000) to describe lava flows where the flow top is composed of small and broken pāhoehoe crust instead of ‘a‘ā clinker. They are mainly found on alkali lavas when the flux is high enough to decouple the upper crust from the hot lava core (Keszthelyi, Thordarson, & Self, 2001). Rubbly pāhoehoe is sometimes referred to as a transitional state from pāhoehoe to ‘a‘ā (Keszthelyi & Thordarson, 2000). The slabby-to-rubbly pāhoehoe morphology is principally found at the mouth of Tseax River Valley and uniquely on Flow 1. The pāhoehoe slabs are similar to those observed on slabby pāhoehoe but are not randomly distributed. Near the front of the unit, the largest slabs are up to 1.5 m thick and ~ 5 m2. Moving upstream, the slabs become smaller and thinner, i.e., 0.10-0.40 m thick, and less than 3 m2, eventually fragmenting into rubble. Rubble is loose blocks of pāhoehoe crust, generally smaller than 0.5 m in diameter (rare rubble up to 1.5 m in diameter were observed). While a few dense rubble were found (probably from the lava flow interior), the majority of the rubble are vesicular and from the fragmented upper crust.

The slabby-to-rubbly pāhoehoe is associated with pressure ridges oriented perpendicularly to flow direction. The ridges are 10-150 m long and frequently coalesced together. The pressure ridges are up to 3.5 m high at the front of the unit and progressively decrease to 0.5-1 m high in the most upstream part of the unit. The wavelength of the pressure ridges ranges from 10 to 25 m on average with the longest wavelength located near the front of the unit.

Grooves are found on slabby and slabby-to-rubbly pāhoehoe flow surfaces. Grooves are long linear fractures that are formed after a flow thick enough to move over a barrier to the flow advance or topographic highs (Guilbaud et al., 2005).

Rafted Plates and Major Breakouts. Rafted plates are large rounded to polygonal plates of undisturbed pāhoehoe at crust surfaces surrounded by slabby pāhoehoe (Figure 5(h)). Large pāhoehoe lava surfaces composed of rafted plates are sometimes called platy lava (Keszthelyi et al., 2004; Stevenson et al., 2012). The edges of the surface plates are sometimes curved up and progressively broken into large and thick pāhoehoe slabs. 465 surface plates of more than 25 m2 were mapped. The largest surface plates are best visible from aerial photographs (the average area is 1,116 m2, the largest plate is ~ 70,000 m2). On both flows, they form a patchwork of 5 to 25 plates of rafted pāhoehoe crust (visually similar to a polar pack ice seen from above) surrounded by disrupted slabby pāhoehoe.

The majority of the rafted plates on Flow 1 are found downstream of major breakouts. We define major breakouts as a sudden disruption of large lava-rises into slabby pāhoehoe with rafted plates located directly downstream. A major breakout is characterized by piling up of large pāhoehoe slabs forming linear mounds up to 5 m high. At least 23 major breakouts > 100 m-long were identified from aerial photographs.

Lava-rise pits. Lava-rises and hummocky-to-lava-rises pāhoehoe are disturbed by collapsed pits that are mainly collapsed roofs of lava tubes (Figure 5(g); Walker, 1991). They form when lava drains downstream and reflect part of the system in the Nass Valley. Some of these pits are also related to different amounts of lava inflation (Walker, 1991) and are 2 to 5 m deep, circular to polygonal pits up to 9950 m2. The Main map only shows the largest collapsed lava tubes (i.e., < 250 m2). Large polygonal broken slabs of the pāhoehoe upper crust (~ 0.5 - 1 m thick) form the floor of the collapsed pits.

References

Guilbaud, M.-n., Self, S., Thordarson, T., & Blake, S. (2005). Morphology, surface structures, and emplacement of lavas produced by , A.D. 1783-1784. Geological Society of America Special Papers, 396 (07), 81-102. doi: 10.1130/ 2005.2396(07)

Harris, A. J., Rowland, S. K., Villeneuve, N., & Thordarson, T. (2017). Phoehoe, aa, and block lava: an illustrated history of the nomenclature. Bulletin of Volcanology, 79 (1), 7. doi: 10.1007/s00445-016-1075-7

Hon, K. E. N., Kauahikaua, J. I. M., Denlinger, R., & Mackay, K. (1994). Emplacement and inflation of pahoehoe sheet flows: Observations and measurements of active lava flows on Kilauea Volcano, . Geological Society of America Bulletin, 106, 351-370. doi: 10.1130/0016-7606(1994)106<0351

Keszthelyi, L., & Thordarson, T. (2000). Rubbly pahoehoe: a previously undescribed but widespread lava type transitional between aa and pahoehoe. In Meeting of the geological society of America abstracts with program.

Keszthelyi, L., Thordarson, T., & Self, S. (2001). Rubbly Pahoehoe: Implication for Eruptions and their Atmospheric Effects. In American geophysical union fall meeting abstracts.

Keszthelyi, L., Thordarson, T., McEwen, A., Haack, H., Guilbaud, M. N., Self, S., & Rossi, M. J. (2004). Icelandic analogs to Martian flood lavas. Geochemistry, Geophysics, Geosystems, 5 (11). doi: 10.1029/2004GC000758

Lipman, P. W., & Banks, N. G. (1987). A'a flow dynamics, 1984 eruption. US Geological Survey Professional Paper, 1350(May), 1527-1567.

Macdonald, G. A. (1953). Pahoehoe, aa, and block lava. American Journal of Science, 251, 169- 191. doi: 10.2475/ajs.251.3.169

Peterson, D. W., & Tilling, R. I. (1980). Transition of basaltic lava from pahoehoe to aa, Kilauea Volcano, Hawaii: Field observations and key factors. Journal of Volcanology and Geothermal Research, 7 (3-4), 271-293. doi: 10.1016/0377-0273(80)90033-5

Self, S., Keszthelyi, L., & Thordarson, T. (1998). The importance of Pahoehoe. Annual Review of Earth and Planetary Sciences, 26, 81-110.

Stevenson, J. A., Mitchell, N. C., Cassidy, M., & Pinkerton, H. (2012). Widespread inflation and drainage of a pahoehoe flow field: The Nesjahraun, Thorngingvellir, Iceland. Bulletin of Volcanology, 74 (1), 15-31. doi: 10.1007/s00445-011-0482-z

Swanson, D. A. (1973). Pahoehoe flows from the 1969-1971 eruption, Kilauea volcano, Hawaii. Bulletin of the Geological Society of America, 84 (2), 615-626. doi: 10.1130/0016-7606(1973)84<615:PFFTMU>2.0.CO;2

Walker, G. P. L. (1991). Structure and origin by injection of lava under surface crust, of tumuli, lava rises, lava rise pits, and inflation clefts in Hawaii. Bulletin of Volcanology, 53 (4), 612-634. doi: 10.1007/BF00298155