PUBLICATIONS

Geochemistry, Geophysics, Geosystems

RESEARCH ARTICLE The insular shelves of the Faial-Pico Ridge ( archipelago): 10.1002/2015GC005733 A morphological record of its evolution

Key Points: R. Quartau1,2, J. Madeira2,3, N. C. Mitchell4, F. Tempera5, P. F. Silva2,6, and F. Brandao~ 7  The Pico insular shelf is dominated by volcanic progradation 1Divis~ao de Geologia e Georecursos Marinhos, Instituto Portugues^ do Mar e da Atmosfera I.P., Lisboa, , 2Instituto  Prograding lava deltas that reach the Dom Luiz, Faculdade de Ciencias^ da Universidade de Lisboa, Lisboa, Portugal, 3Departamento de Geologia da Faculdade shelf edge trigger slope instability de Ciencias,^ Universidade de Lisboa, Lisboa, Portugal, 4School of Earth, Atmospheric and Environmental Sciences, Univer-  Improved understanding of Faial-Pico 5 ridge geological evolution is sity of Manchester, Manchester, UK, Marine and Environmental Sciences Centre and Institute of Marine Research, Depar- achieved tamento de Oceanografia e Pescas, Universidade dos Ac¸ores, , Portugal, 6Area Departamental de Fısica, Instituto Superior de Engenharia de Lisboa, Lisboa, Portugal, 7Estrutura de Miss~ao para a Extens~ao da Plataforma Continental, Pac¸o Supporting Information: de Arcos, Portugal  Supporting Information S1

Shelves surrounding reefless volcanic ocean islands are formed by surf erosion of their slopes Correspondence to: Abstract R. Quartau, during changing sea levels. Posterosional lava flows, if abundant, can cross the coastal cliffs and fill partially [email protected] or completely the accommodation space left by erosion. In this study, multibeam bathymetry, high- resolution seismic reflection profiles, and sediment samples are used to characterize the morphology of the Citation: insular shelves adjacent to . The data show offshore fresh lava flow morphologies, as well as an Quartau, R., J. , N. C. Mitchell, irregular basement beneath shelf sedimentary bodies and reduced shelf width adjacent to older volcanic F. Tempera, P. F. Silva, and F. Brand~ao (2015), The insular shelves of the edifices in Pico. These observations suggest that these shelves have been significantly filled by volcanic pro- Faial-Pico Ridge (Azores archipelago): gradation and can thus be classified as ‘‘rejuvenated.’’ Despite the general volcanic infilling of the shelves A morphological record of its around Pico, most of their edges are below the depth of the Last Glacial Maximum, revealing that at least evolution, Geochem. Geophys. Geosyst., 16, 1401–1420, doi:10.1002/ parts of the island have subsided after the shelves formed by surf erosion. Prograding lava deltas reached 2015GC005733. the shelf edge in some areas triggering small slope failures, locally decreasing the shelf width and depth of their edges. These areas can represent a significant risk for the local population; hence, their identification Received 20 JAN 2015 can be useful for hazard assessment and contribute to wiser land use planning. Shelf and subaerial geomor- Accepted 11 APR 2015 phology, magnetic anomalies and crustal structure data of the two islands were also interpreted to recon- Accepted article online 15 APR 2015 struct the long-term combined onshore and offshore evolution of the Faial-Pico ridge. The subaerial Published online 12 MAY 2015 Corrected 26 JUN 2015 emergence of this ridge is apparently older than previously thought, i.e., before 850 ka.

This article was corrected on 26 JUN 2015. See the end of the full text for details. 1. Introduction Volcanic ocean islands grow above sea level by volcanic construction occurring faster than the effects of destructive geological processes. Their present-day morphology results from the action imprinted by vol- canic, erosional, depositional, tectonic, isostatic, eustatic, and mass-wasting processes [Ramalho et al., 2013]. Although surveying island submarine flanks with sonar, seismic, and submersibles has become common, most of the information discerned in the literature about island geological evolution is still based on onshore studies. Furthermore, the majority of the offshore surveys have been focused on their submarine slopes to study mass-wasting deposits and volcanic geomorphology [e.g., Glass et al., 2007; Holcomb and Searle, 1991; Krastel et al., 2001; Le Friant et al., 2011; Llanes et al., 2009; Mitchell et al., 2002; Mitchell, 2003; Moore et al., 1994]. This concentration of effort in the deeper parts of volcanic islands is also a consequence of the difficulty in performing nearshore surveys, because it is dangerous and time consuming to survey shallow-water coastal areas. Consequently, little is known about the shelves surrounding these islands. How- ever, the morphology that results from the imprint of the physical processes over these nearshore features can provide valuable insights on the geological evolution of the islands [Babonneau et al., 2013; Casalbore et al., 2015; Chiocci et al., 2013; Fletcher et al., 2008; Kennedy et al., 2002; Mitchell et al., 2012a, 2008; Quartau and Mitchell, 2013; Quartau et al., 2010, 2012, 2014; Romagnoli, 2013]. The shelves surrounding volcanic ocean islands are formed by a competition between wave erosion, which

VC 2015. American Geophysical Union. forms and enlarges them, and volcanic progradation which narrows them [see Quartau et al., 2010, Figures All Rights Reserved. 1 and 2]. In general, the shelves of the Azorean islands dominated by wave erosion tend to be wider and

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1401 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

present erosional surfaces that are in profile sharply angular with the submarine slopes and are normally covered by widespread sea level highstand sediments [Quartau et al., 2010, 2012, 2014]. The shelves that are dominated by volcanic progradation are narrow, mainly underlain by rocky outcrops of submarine lava flows and are overlain by thinner and more localized sedimentary bodies. The morphology that results from the interplay between these two main processes can be used as a proxy for the relative age of the adjacent subaerial volcanoes, allowing us to reconstruct their extents prior to wave erosion, measure the vertical movements of the island, and estimate the timing of posterosional volcanism. The guidelines on performing this geomorphological analysis have been detailed in Quartau et al. [2014], so only a basic explanation is provided in section 3. In this work, high-resolution seismic reflection, multibeam bathymetry, magnetic, and sediment data are used to characterize the morphology of the shelf surrounding Faial and Pico islands, with a clear focus on Pico. We show how volcanic progradation can be inferred using morphological analysis of the shelves and cliffs. The results suggest that volcanic progradation has been the main process in determining the shelf morphology surrounding Pico Island. A conceptual model of how insular shelves dominated by volcanic progradation evolve is put forward and the term ‘‘rejuvenated shelves’’ is proposed. In addition, we suggest a dimensionless variable as a way to distinguish shelves dominated by wave erosion from those that are rejuvenated. Locally voluminous lava deltas almost entirely fill the erosional shelves and have apparently triggered small mass-wasting events at the shelf edge. Although giant landslides on the flanks of volcanic islands are infrequent, for example, occurring every 100 ka in the Hawaiian archipelago [Garcia et al., 2007] and 125–170 ka in the Canary archipelago [Krastel et al., 2001], smaller failures of the upper slopes are able to generate local but significant tsunamis [Casalbore et al., 2011; Kelfoun et al., 2010]. Small-scale failures caused by lava progradation [Bosman et al., 2014; Chiocci et al., 2008; Poland and Orr, 2014; Sansone and Smith, 2006] probably pose a frequent threat to local populations but one that remains poorly evaluated. Mapping of such areas can be more efficient for hazard assessment than time-consuming techniques such as satellite and ground-based sensors for estimating the subsidence of coastlines. Finally, the combination of the offshore morphology with the onshore geomorphology allowed us to propose a new model for the Faial-Pico Ridge evolution in the Azores archipelago.

2. Regional Setting The Azores archipelago (Figure 1) is located in the central North , on an irregular triangular- shaped plateau limited by the 2000 m bathymetric contour [Lourenc¸o et al., 1998]. This plateau is the result of magmatic activity related to a hot spot close to the triple junction between the Eurasian (Eu), Nubian (Nu), and North American (NA) plates [e.g., Cannat et al., 1999; Gente et al., 2003; Laughton and Whit- marsh, 1974; Schilling, 1975]. The plateau extends beyond the Mid-Atlantic Ridge (MAR) to the NA plate, where the of islands (Flores and Corvo) lie. East of the MAR, the plateau is bounded in the north by the Terceira Rift (TR) and in the south by the inactive East Azores Fracture Zone (EAFZ). Linking the MAR to the western tip of the Gloria Fault (GF) lies a complex structure composed of alternating volcanic edifices and tectonic basins that progressively rotates from E-W to WNW-ESE and NW-SE. Over this - tectonic structure which is the expression of the right-transtensional shear zone that constitutes the Eu-Nu boundary [Hipolito et al., 2013; Lourenc¸o et al., 1998; Marques et al., 2013] lie the central (Terceira, , S~ao Jorge, Pico, and Faial Islands) and eastern groups of islands (S~ao Miguel and Santa Maria Islands).

2.1. Local Geological Setting The following details of Pico Island’s geology are taken from several studies [Madeira, 1998; Madeira and Brum da Silveira, 2003; Nunes, 1999]. Although there is no consensus about Pico’s volcanic stratigraphy, we will follow the scheme proposed by Nunes [1999] because it incorporates the most recent and detailed mapping. This volcano-stratigraphy is based on field relations, geomorphological and weathering criteria, and a limited number of radiometric ages. Pico Island corresponds to the westward subaerial expression of an elongated submarine ridge of around 83 km length (Figure 1) [Lourenc¸o et al., 1998; Mitchell et al., 2012b; Stretch et al., 2006]. Above sea level, Pico Island is 46 km in length, has a maximum width of 15.8 km, and rises to an altitude of 2351 m at Ponta do Pico (Figure 2). The subaerial part of the island can be divided into three different physiographic regions [Madeira, 1998]: a tall stratovolcano, Pico Volcano, which dominates the western part of the island; a shield-

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1402 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

Figure 1. Regional setting of the Azores archipelago within the North American (NA), Eurasian (Eu), and Nubian (Nu) triple junction. The top right inset depicts the location of the Azores archipelago in the North Atlantic Ocean. Tectonic structures: MAR, Mid-Atlantic Ridge; NAFZ, North Azores Fracture Zone; PAFZ, Princess Alice Fracture Zone; EAFZ, East Azores Fracture Zone; GF, Gloria Fault; Terceira Rift (TR). Islands: C, Corvo; F, Flores; SJ, S~ao Jorge; Gr, Graciosa; T, Terceira; SM, S~ao Miguel; SMa, Santa Maria. Faial and Pico Islands are highlighted in black. Solid lines correspond to major structures/faults. Bathymetry of the Azores archipelago from Lourenc¸o et al. [1998].

Figure 2. Simplified geologic map of Pico Island (modified from Nunes [1999]) draped over a shaded relief image of the subaerial topogra- phy [from Instituto Geografico do Exercito, 2001a, 2001b, 2001c, 2001d, 2001e, 2001f, 2001g]. Pico Volcano, Topo Volcano, and the Achada Plateau constitute the three physiographic regions that coincide, respectively, with the Montanha, Topo-Lajes, and S~ao Roque- Vol- canic Complexes. Straight comb-like black lines represent fault scarps of the graben (see section 2.2 for details): 2 is the Lagoa do Capit~ao fault and 3 the Topo fault. Dashed black lines are inferred faults. Delineation of landslide scars in red is from: 4 (S~ao Mateus landslide; adapted from [Madeira, 1998]), 5 (Topo landslide; adapted from Hildenbrand et al. [2012a]), 6 and 7 [adapted from Costa et al., 2014]. Names with the dates correspond to historical eruptions. The top right inset represents the Faial-Pico channel and the . RVC, CVC, AF, and CapVC correspond, respectively, to Ribeirinha Volcanic Complex, Caldeira Volcanic Complex, Almoxarife Formation, and Capelo Volcanic Complex. Straight comb-like red lines represent fault scarps of the graben structure in the island. Lines offshore represent 50, 100, 200, and 500 m bathymetry contours (digitized from Instituto Hidrografico [1999]).

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1403 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

like volcanic structure, Topo volcano, located on the south-central part of the island; and the Achada Pla- teau, a 29 km long, WNW-ESE to NW-SE volcanic ridge. These regions roughly coincide with the three main volcanic complexes described by Nunes [1999]: Montanha Volcanic Complex (MVC), Topo-Lajes Volcanic Complex (TLVC), and S~ao Roque-Piedade Volcanic Complex (SRPVC) (Figure 2). The oldest part of the island is the Topo volcano, which forms a promontory on its southern-central part (Figure 2). It is formed of ankaramitic and basaltic lava flows of the TLVC and the only published K/Ar date provided an age of 250 6 40 ka [Chovelon, 1982]. The still reaches an elevation of 1020 m, although modified by landslides, displaced by faulting, and covered by younger lavas [Hildenbrand et al., 2012a, 2013; Madeira, 1998; Madeira and Brum da Silveira, 2003; Mitchell et al., 2012a, 2013]. The Achada Plateau (Figure 2) can be divided in two sections: the easternmost section trending WNW-ESE (from the eastern tip of the island to around number 1 in Figure 2) and the westernmost section trending NW-SE (from around number 1 in Figure 2 to the western end of SRPVC). It is composed of numerous Hawaiian/strombolian scoria cones and related lava flows of the SRPVC, located along the WNW-ESE to NW- SE trending axis of the plateau (Figure 2). Lavas issuing from these cones are mostly aa-flows, which cas- caded down the steep northern and southern slopes of the plateau. The oldest subaerial lavas have a K/Ar age of 230 ka [Chovelon, 1982] but the present-day morphology is dominated by recent flows. Some of these flows have passed over cliffs cut in older units and originated lava deltas, such as the event of 1562– 1564 A.D. eruption of Ponta do Misterio (Figure 2). Volcanic activity has jumped along the Achada Plateau ridge without any well-defined migration pattern [Nunes, 1999; Nunes et al., 2006]. The basaltic stratovolcano of Pico Mountain (Figure 2) rises 2351 m above sea level. MVC volcanism is basal- tic and dominantly Hawaiian-strombolian, with occasional littoral surtseyan eruptions. Besides summit erup- tions, abundant flank eruptions have occurred on Pico Mountain, producing scoria and spatter cones, and eruptive fissures and associated lava flows. The single K/Ar date of this volcanic complex is from Costa et al. [2014] and points out to an age of 53 ka. The very fresh morphology of Pico Mountain and the radiocarbon dates indicate intense activity during the Holocene [Madeira, 1998; Nunes, 1999]. Since the settlement in the fifteenth century, there have been two eruptions (Figure 2), in 1718 A.D. (Santa Luzia and S~ao Jo~ao eruptive centers) and 1720 A.D. (Silveira eruption) and a submarine eruption off the NW coast in 1963 [Madeira, 1998, 2005; Nunes, 1999; Weston, 1964; Zbyszewski, 1963]. Pico Island is bordered to the west by a narrow and shallow channel that separates it from Faial Island; hence, it is called the Faial-Pico Channel. This channel (top right inset in Figure 2) is in fact the submarine link between the two islands. It constitutes an 8 km wide shelf whose depth varies from 200 m at the N and S entrances to an average of 80 m in the middle. East Faial is formed of volcanic products erupted by the Ribeirinha shield volcano that rose above sea level at 850–800 ka [Feraud, 1980] and was reactivated 400 ka ago in its NE flank [Hildenbrand et al., 2012b]. The west flank of this volcano is overlapped by the Caldeira stratovolcano since 120 ka [Hildenbrand et al., 2012b]. Two basaltic fissural volcanic systems developed on SE Faial (Almoxarife Formation at 30 6 20 ka [Feraud et al., 1980]) and on the west tip of the island (Capelo Peninsula over the last 8 ka [Di Chiara et al., 2014]). The later was the site of the only two historical eruptions in 1672–1674 and 1957–1958 [Madeira et al., 1995].

2.2. Tectonic Structure The island presents two main fault systems [Madeira, 1998; Madeira and Brum da Silveira, 2003]. The WNW- ESE to NW-SE trending normal dextral structures are mostly present on the Achada Plateau and are expressed by alternating volcanic alignments (scoria cones, scoria ramparts, and crater rows) and short fault scarps. This system forms a graben structure bounded in the north by the Lagoa do Capit~ao and in the south by the Topo fault zones (respectively, 2 and 3 in Figure 2). The western portion of the graben is com- pletely covered by Pico stratovolcano while to the east, it is partially filled by lava flows emitted by cones of the SRPVC installed along this fault system. The NNW-SSE trending conjugate faults are marked by volcanic alignments and observed at the outcrop scale. Two large gravitational (flank collapse) structures were observed; one on the south flank of Pico stratovolcano (4 in Figure 2) (S~ao Mateus landslide [Madeira, 1998; Nunes, 1999]) and another on the south flank of Topo shield volcano (5 in Figure 2) (Arrife landslide [Hilden- brand et al., 2012a, 2013; Madeira, 1998; Madeira and Brum da Silveira, 2003; Mitchell et al., 2012a, 2013]). Two further inferred landslides (6 and 7 in Figure 2) on the north side of the island are largely obscured by

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1404 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

Figure 3. (a) Magnetization intensity (A/m) of the Faial-Pico ridge on a shaded relief map of the islands and seafloor. (b) The light green polygon (4) corresponds to the body with high crustal velocities referred in Dias et al. [2007] and the three darker green polygons corre- spond to high-density bodies in Camacho et al. [2007] (1) and Nunes et al. [2006] (2 and 3). Red dashed lines in polygons 1 and 2 are inferred limits of the high-density bodies of Camacho et al. [2007] and Nunes et al. [2006] because the gravimetric surveys were land based.

younger volcanism but have been identified in multibeam sonar data [Costa et al., 2014]. GPS and InSAR data covering the entire island from the 2001–2006 period revealed that the average modern subsidence of the island is between 5.7 and 7.2 mm/a [Catalao~ et al., 2011]. Eastern Faial Island (top right inset in Figure 2) is dominated by a graben system composed of several nor- mal dextral faults, which form two oppositely verging domino structures [Madeira, 1998; Madeira and Brum da Silveira, 2003; Trippanera et al., 2014]. Figure 3 shows a magnetization map of the Faial-Pico ridge combined with a digital elevation model of the islands and the surrounding seafloor. This map represents a solution from a 3-D inversion of aeromagnetic data from the Azores platform [Luis et al., 1994; Miranda et al., 1991], compiled into a singular grid with 0.0058 step (latitude and longitude). Using Mirone software [Luis, 2007], data were reduced to the pole and inverted according the methodology of Parker and Huestis [1974]. The magnetization map results from a layer with constant thickness of 1 km below bathymetric surface. A band-pass filter (1–120 km) was applied

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1405 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

in order to ensure convergence. Most of the subaerial part of the islands is dominated by positive magnet- ization values, suggesting construction during the Brunhes Chron (<790 ka) [Quidelleur et al., 2003], which is consistent with their known recent geological history. However, a significant area, covering the entire channel and part of east Faial and west Pico, presents negative magnetization values. The negative magnet- ization suggests construction of this area during the Matuyama Chron (>790 ka) [Quidelleur et al., 2003]. According to gravity studies, a large high-density body (dark green area numbered 1 in Figure 3), located between 0.5 and 6 km below sea level, with a wall-like structure develops from the central caldera of Faial [Nunes et al., 2006]. No shipboard gravity data exist in the channel, but this body follows the direction of the graben system at Faial, which has a morphologic expression in the channel between these two islands [Quartau and Mitchell, 2013; Tempera, 2008], and most likely extends across to Pico Island where another high-density body can be found [Camacho et al., 2007] (dark green area numbered 2 in Figure 3). Dias et al. [2007] also refer to a body located NNE of Faial Island (light green area numbered 4 in Figure 3) exhibiting high values of P-velocity (6.3–7.4 km/s) located deep in the crust, below 6 km. It partially overlaps the north- ern border of the two high-density bodies mentioned by Camacho et al. [2007] and Nunes et al. [2006]. These three bodies are probably connected and may correspond to solidified magmatic intrusions related with the feeding structures of the Faial-Pico ridge during the Matuyama period.

2.3. Climate, Wave Regime, and Sedimentation Precipitation on Pico Island varies mainly with altitude. Annual precipitation ranges from 1101 to 1845 mm/a in the coastal area [Cruz and Silva, 2001] to more than 3600 mm/a above 800 m in the Pico Mountain and Achada Plateau [Agencia Estatal de Meteorologıa and Instituto de Meteorologia de Portugal, 2012]. In Pico, the cover by young and very permeable lava flows has resulted in a very low drainage den- sity. The few streams are ephemeral and surface drainage is only present during heavy precipitation epi- sodes, mostly in the rainy season (October–March). Information regarding oceanographic conditions in the islands of the central group is detailed in Quartau et al. [2012]. Records show that the Azores are struck on average every 7 years by violent storms (extratropi- cal storms or hurricanes [Andrade et al., 2008]). Even outside these stormy events, regular years show remarkable wave energy [Rusu and Guedes Soares, 2012]. The annual prevailing waves [see Quartau et al., 2012, Figure 3] hit the NW (29%) and W (24%) coastlines. They are also the highest, with average significant

wave height (Hs) of 2.9 and 3.1 m, respectively. Waves from the N are also important (16%) with Hs of 2.5 m.

Waves from the SW also present the highest Hs (3.1 m) but their frequency is much lower (8%). In the spe- cific case of Pico Island, sheltering effects probably occur in the western and north-eastern coasts, caused by the presence of Faial and S~ao Jorge Islands, respectively (Figure 1). The Azorean islands are very young geomorphologically due to their frequent and recent volcanism. There are no wide-open valleys, meandering streams, inundation plains, and all water courses are ephemeral in the archipelago [Louvat and Alle`gre, 1998]; thus, sediments on the shelf are mainly sourced from cliff erosion [Quartau et al., 2012]. The residence time of these sediments nearshore is also likely to be small due to the large wave energy to which these coasts are exposed. Because of that, the nearshore substrates of these islands, down to 30–50 m depth, are mostly rocky [Quartau et al., 2012]. Sediments are remobilized from the nearshore to the middle and outer shelf by downwelling currents that form during storms [Meireles et al., 2013], forming clinoform bodies [Quartau et al., 2012]. During the storms, these currents are so strong that fine sediments like silt or fine sand most likely cross the shelf edge and are deposited on the submarine slopes of the island. Measurements of these currents at 20 m water depth showed velocities up to 2 m/s

associated with Hs of 5 m [Youssef, 2005]. Such wave conditions occur on a yearly basis in the Azores and can easily reach twice this value during stronger storms [Instituto Hidrografico , 2005]. The Azorean shelves dominated by wave erosion have high coastal retreat rates (0.21 m/yr on average [Borges, 2003]) and con- tain extensive and thick (up to 50 m) aggrading to prograding sedimentary bodies, which are believed to have been formed in the last 6.5 ka [Quartau et al., 2012].

3. Data Sets and Methods 3.1. Acquisition and Processing of the Data Set High-resolution bathymetry of the outer shelf and slope of Faial and Pico islands and in Faial-Pico channel was acquired in 2003 down to 1600 m (acquisition and processing described in Mitchell et al. [2008]). This

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1406 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

Figure 4. Location of the data sets used for this study. The light blue and dark green areas offshore represent, respectively, the 2003 and 2004 multibeam bathymetric surveys. The dark blue represents the EMEPC multibeam bathymetric survey. The light green area represents the single-beam bathymetric surveys acquired in 2001-2002 and the orange the single-beam bathymetry from the EMODNET project. Red and blue lines depict, respectively, the CHIRP and boomer seismic profiles. The black dots represent the surface sediment sampling sites.

data set was supplemented with further surveying in 2004 in the western inner shelf area of Pico and inner shelf of Faial (acquisition and processing described in Tempera [2008]). Additional multibeam bathymetry was acquired in the scope of the Portuguese continental shelf extension program using a Simrad EM710 down to 1760 m (EMEPC, data courtesy). The remaining nearshore gaps around Faial and Pico were filled with single-beam bathymetry collected in 2001 (Quartau et al., 2002a) and in 2002 (Quar- tau et al., 2002b) concurrently with the seismic reflection survey. Offshore, the remaining gaps were filled with single-beam bathymetry from the EMODnet European project. A bathymetric mosaic compilation was produced from these five data sets with a cell size of 10 m for the multibeam data, 50 m for the near- shore single-beam data, and 500 m for the offshore single-beam data. Figure 4 shows the hydrographic surveys and Figures 5a and 5b the resulting bathymetry. High-resolution seismic reflection profiles were collected around Pico Island with a CHIRP sonar system (1.5–10 KHz, Datasonics CAP-6000W) and a boomer (EG&G 230-1) in July 2002, onboard the R/L Aguas Vivas and the R/V Arquipelago [Quartau et al., 2002b]. The seismic tracks (Figure 4) were designed for marine aggregate mapping purposes, so they were not placed optimally for studying the outer shelf of the island. The lines were acquired between 10 and 80 m water depth, seldom reaching the shelf edge. Survey lines were mainly normal to the shore separated by 700 m with a cross-line run around the island at approximately 40 m depth. Acquisition and processing parameters are described in Quartau et al. [2012]. Most of the subbottom analysis was carried out on the boomer data, whereas the CHIRP was mainly used for seabed classification because of limited penetration over the coarse sandy sedimentary seafloor [Quartau and Curado, 2002b; Quartau et al., 2003]. Vertical resolution of the boomer data is 1–2 m. The subaerial digital elevation model of the islands (Figures 5a and 5b) used in this work has a cell size of 10 m and was produced by the Portuguese military survey from 1:25000 topographic maps [Instituto Geo- grafico do Exercito, 2001a, 2001b, 2001c, 2001d, 2001e, 2001f, 2001g, 2001h, 2001i, 2001j]. Superficial sediment sampling was conducted in November 2003, aboard the R/V Arquipelago [Quartau et al., 2005]. Fifty-four stations were sampled using a box-corer (Figure 4). Grain size data were obtained using a dry sieving technique for material coarser than 21 Ø (2 mm) and a Coulter Counter LS-230 for finer- grained material (<2 mm). The dry sieving was carried out over intervals of 1 Ø (25Ø,24Ø,23Ø,22Ø,

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1407 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

Figure 5. (a) Submarine topography: shaded relief imagery derived from the bathymetric compilation. Subaerial topography: shaded relief based on the Instituto Geografico do Exercito [2001a, 2001b, 2001c, 2001d, 2001e, 2001f, 2001g, 2001h, 2001i, 2001j] topographic maps; (b) Interpreted submarine topography: alternating black and blue lines represent the shelf edge and corresponding top of the cliffs of the shelf sectors defined around Pico Island. Light green, dark green, magenta, and light blue lines represent gradient breaks in the Faial-Pico channel. Red, black, and dark blue lines represent the shelf edge of the main volcanic edifices in Faial. Letters A–G represent each individ- ual shelf segment or gradient breaks referred to in the text. Brown areas offshore represent submarine lava flow fields and green areas rep- resent rocky outcrops without any interpretable volcanic structure and boulder fields. The black boxes locate the areas represented in Figures S1_a1–S1_f1 of the supporting information. Pink dashed dot line represents the interpreted old south coastline of the western Achada Plateau. Solid and dashed black lines onshore and offshore are, respectively, faults and inferred faults. Red dashed lines offshore represents the limits of the debris avalanches caused by the onshore landslides.

and 21 Ø sieves). Both data sets were then merged to produce composite grain size distributions. Sediment texture analysis shows that most samples are coarse sands according to the Udden-Wentworth scale [Went- worth, 1922], mainly composed of volcaniclastic sand (see Figures S1_a1_i, S1_b1_i, S1_b1_iii, S1_c1_i, S1_c3_i and S1_d3_i in the supporting information). Percentage of carbonate skeletal particles is normally less than 10%, but can occasionally reach higher values, up to 90%.

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1408 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

3.2. Geomorphological Analysis of the Shelves Shelves with a sharp seafloor surface (covered or not by sedimentary bodies) are interpreted to have resulted from surf erosion during the last sea level rise or even from previous sea level cycles. Additional information regarding the island history can be obtained from the depth of the present-day erosional shelf edge if the island has not been uplifting. If the depth of the edge is shallower than 123 m (the maximum depth reached during the Last Glacial Maximum (LGM) in Bintanja et al. [2005]), these shelves have most likely started forming sometime during the rise of sea level after the LGM [Quartau et al., 2014]. Insular shelves in Pico are mostly formed by erosion of rock surfaces, since the predominant volcanism is effusive. Therefore, the depth of the erosional shelf edge should record the water level at that time [Quartau et al., 2010; Trenhaile, 2000, 2001]. Around Pico, the present depth of the shelf edge marks the beginning of the formation of the shelf and a sea level curve might be used to roughly infer the timing of that beginning. That timing marks the end of the main building phase of the volcanic edifice and the start of incision by wave action. If, on the contrary, the depth of the edge is deeper than 123 m depth, these shelf sectors were formed by erosion during the first lowstand after the main adjacent volcanism took place and have sub- sided since then [Quartau et al., 2014]. A minimum subsidence rate can then be estimated based on the age of the occurrence of the first lowstand after the end of the main volcanism. High cliffs backing the shelves, absence of submarine lava flows and well-developed mass-wasting features on their seaward edges can also be evidence of their old age. Such evidence implies that the volcanic edifice being incised by wave action was formed before the LGM. We use the term erosional shelves in the following text for those dominated by wave erosion.

Shelves covered by fresh submarine lava flow morphologies that do not show signs of erosion, with low adjacent coastlines and commonly irregular in plan view, imply recent volcanic progradation; i.e., after 6.5 ka, when sea level reached almost its present-day position in Iberia [Hernandez-Molina et al., 1994], at the same latitude as the Azores. Here sea level has not yet intersected these structures so surf has not destroyed their primary surface features. The range of depths at which the fresh submarine lava flows occur may con- strain the age of their progradation onto the shelf during the sea-level rise after LGM since their position suggests that sea level was already above that depth of emplacement [Quartau et al., 2014]. Erosional shelves in Faial and Terceira show evidence of volcanic progradation [Quartau et al., 2010, 2012, 2014]; how- ever, these lava flows are often isolated and occupy small areas when compared with the erosional features of those shelves. We use the term rejuvenated shelves in the text for those that were formed at the LGM or before, but are now dominated by volcanic progradation. Nevertheless, these shelves still preserve some of their old evidence of erosion (shelf edges at 2123 m or deeper), although showing widespread volcanic progradation (low coastlines and the seafloor mostly covered by fresh lava flows). Another sign of shelf reju- venation is the irregular basement of the sedimentary cover (assumed to have been deposited after 6.5 ka, [Quartau et al., 2012]) that can be seen in seismic profiles. Irregular basements suggest that lava flows have prograded onto the shelf when sea level was at a higher position than the irregularity left by progradation and has not dropped since then. Otherwise, surf erosion would have smoothed these irregular features, cre- ating sharp erosion surfaces (see examples of this interaction in Quartau et al. [2012, Figure 14]).

The bathymetry, seismic profiles, topography, and onshore geology were integrated in a GIS environment to facilitate interpretation of the nearshore submarine morphology of the islands. The islands were divided into different shelf sectors based on their distinctive morphologies, which often have a clear correspon- dence with the onshore geomorphology. Three graphs were produced for each shelf sector showing the variation of shelf width, shelf break depth, and cliff height. On erosional shelf sectors, there is normally a good correspondence between these three morphologic indicators, i.e., high cliffs are present where there are wider shelves and deeper shelf edges. On the other hand, rejuvenated shelf sectors also have good cor- respondences, with low or absent cliffs occurring where there are narrow shelves and shallow shelf edges, although still showing localized evidence of their old age (isolated stretches of shelf edges at 2123 m or deeper and high cliffs). Between these two extremes (erosional and rejuvenated shelves) intermediate com- binations of these characteristics exist. In order to produce the graphs of Figures S1_a1_i, S1_b1_i, S1_b1_iii, S1_c1_i, S1_c3_i, S1_d1_i, S1_d3_i, S1_e1_i, S1_e2_i and S1_f1_i (see supporting information) the bathymetry and topography were resampled to 50 m resolution. The depth of the shelf break and height of the cliffs (where present) in each sector was determined at 50 m intervals, respectively, from the bathymet- ric and topographic data. Profiles for each sector were constructed, at 50 m intervals as perpendicular as

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1409 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

Table 1. Shelf Width, Cliff Height, and Depth of the Shelf Edge of the Sectors Defined for Pico Islanda E Achada Plateau Topo Sector

Southern Subsector A1 Northern Subsector A2 Southern Subsector B1 Northern Subsector B2

Min Avg Max Min Avg Max Min Avg Max Min Avg Max Shelf width 138 577 1272 157 822 1283 120 985 2040 773 1293 1851 Cliff height 0 66 194 0 84 410 0 30 178 0 34 180 Depth of the edge 15 88 163 26 74 127 12 112 199 76 117 198 W Achada Plateau Pico Mountain

Southern Subsector C1 Northern Subsector C2 Southern Subsector D1 Northern Subsector D2

Min Avg Max Min Avg Max Min Avg Max Min Avg Max Shelf width 79 179 377 234 900 1612 78 250 684 425 1701 2335 Cliff height 0 16 42 0 32 157 0 16 32 0 8 21 Depth of the edge 5 21 59 16 93 156 6 25 78 22 118 157

aMin, Avg, and Max represent the minimum, average, and maximum values in meters.

possible to the coastline, from the shoreline to the shelf edge, to determine shelf width. Minimum, average, and maximum values of the three morphologic indicators of each sector are presented in Table 1. The detailed description and discussion of the different shelf sectors defined in this study can be found at the supporting information section outside the main article. A brief presentation and discussion of the main results is presented in the following section. Readers are invited to skim the supporting information for a comprehensive understanding of the main interpretations achieved in this study.

4. Morphology of the Faial and Pico Island Shelves: Results and Implications 4.1. Conceptual Model of the Development of Rejuvenated Shelves The distinct nearshore submarine morphologic features around Pico Island allowed their division into five main sectors (Figures 5a and 5b). Their morphologies are directly linked to the evolution of the five main volcanic constructions along Pico Island. Thus, from east to the west we have the eastern Achada Plateau (sectors A1 and A2), the Topo volcano (sectors B1 and B2), the western Achada Plateau (sectors C1 and C2), the Pico Volcano (sectors D1 and D2), and the channel Faial-Pico (sectors E1, E2, E3, and E4). From A to D, these shelf sectors have always a north and south counterpart because they result from erosion of more or less centered onshore volcanism. The analysis of the data (Figures S1_a1 to S1_d4 in the supporting infor- mation and Table 1) shows shelves that have been formed by wave erosion and whose accommodation space has been continually diminished by volcanic progradation, although some subsidence has occurred. The seafloor is either filled by lava flows (brown areas in Figure 5b and Figures S1_a1_i, S1_b1_i, S1_b1_iii, S1_c1_i, S1_c3_i, S1_d1_i, and S1_d3_i) or sedimentary bodies (all areas that are not covered by lava flows or other unstructured rocky outcrops are sedimentary) underlain by a commonly irregular basement reflec- tion which is typical of shelves dominated by recent volcanic progradation (Figures 6a and 6b and Figures S1_a2–S1_a5, S1_b2–S1_b4, S1_c2, S1_c4–S1_c5, S1_d2 and S1_d4 in the supporting information). Almost all the shelf sectors have edges at depths below 123 m (Table 1), which suggests that these shelves were not completely filled by recent volcanic progradation. Their early edges were preserved and reflect an age of wave incision during the LGM and have subsided quickly, or at an earlier lowstand and subsided more slowly. GPS subsidence rates (5.7–7.2 mm/a) [Catalao~ et al., 2011] favor the first hypothesis, although as seen by the depth of shelf edges from volcanic edifices of known age in Faial and Terceira, longer-term sub- sidence rates are normally much lower (below 1 mm/a) [Quartau et al., 2010, 2014]. Given also that the ages of the main volcanism of the adjacent edifices are significant older than the LGM, it is more likely that these shelves started forming at an earlier lowstand although significantly changed by post-LGM volcanic progradation. Oceanographic conditions, coastline lithology, and gradient of the shelf surface can be very different around ocean island volcanoes worldwide; hence, one should expect a large variety of shelf widening rates. Nevertheless, several studies have shown that long-term (up to several Ma) shelf widening rates can vary between 0.6 and 9 mm/a [Dickson, 2004; Le Friant et al., 2004; Llanes et al., 2009; Menard, 1983, 1986;

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1410 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

Figure 6. Seismic profiles showing irregular basements beneath sediments, interpreted as having resulted from recent lava progradation to the shelf (uninterpreted (a) and (c) and interpreted (b) and (d)). Vertical exaggeration is approximately 6X. Red solid lines represent the acoustic basement interpreted as the base of the sedimentary bodies. Profile lines are located in Figure S1_a1_i of the supporting information.

Quartau et al., 2010]. However, these rates are underestimated because the effects of shelf-narrowing proc- esses (mainly by volcanic progradation and shelf edge mass wasting), which can also differ greatly between islands, were not removed from these calculations. This is probably the reason, why, when looking at data for shorter periods (less than 100 ka), faster rates are reported. For instance, as in Faial (up to 60 mm/a [Quartau et al., 2010]), Terceira (up to 34 mm/a, calculated from maximum width of sector 4 in Quartau et al. [2014] and the 60 ka volcanism age from Hildenbrand et al. [2014]), or Vulcano islands (up to 34 mm/a, cal- culated from Romagnoli [2013, Figure 3.9]). The likelihood of some of these shelf-narrowing processes hav- ing occurred in such short periods is probably smaller, thus rates at this time-scale tend to approximate to net erosional values. On the other hand, rates tend to approximate over longer time scales because wave- eroded surfaces trend toward a state of equilibrium, as they become wider and more gently sloping [Quar- tau et al., 2010; Trenhaile, 2000]. Therefore, the best way of determining which process is dominating (wave erosion or volcanic progradation) when comparing different shelf sectors is to use a dimensionless variable. This can be achieved by comparing the maximum shelf width (max_W) against minimum shelf width (min_W) within the same shelf sector (i.e., formed by erosion after the main volcanic phase of the same vol- canic building). If a shelf sector is dominated by wave erosion, the ratio of max_W to min_W should be small. Shelf sectors that are rejuvenated should present high max_W to min_W ratios, if volcanic prograda- tion is voluminous and heterogeneous, i.e., not constant along the entire coastline. Table 2 shows this analy- sis for Faial and Terceira based on data from Quartau et al. [2012, 2014], where shelves are mostly dominated by wave erosion, and for Pico shelves. The ratio of max_W to min_W in Faial and Terceira is gen- erally between 1 and 2. The values higher than 2 on these islands are related to rejuvenated sectors (sectors A1 and A2 in Faial and sector 1 in Terceira) or changing wave exposure along the coastline (sector 3 in Ter- ceira). The ratios for Pico’s shelves are much higher (normally above 5 and up to 17). The sectors with lower ratios correspond to coastline segments where volcanic progradation were more or less homogeneous along-coastline, like sector B2 (ratio of 2). In shelf sectors where volcanic progradation is heterogeneous, such as sector B1 where the progradation of lava delta occurs on its east side, the ratio of max_W

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1411 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

Table 2. Maximum (Max) and Minimum (Min) Shelf Widths and Their Ratios for the Sectors Defined in Faial [Quartau et al., 2012], Terceira [Quartau et al., 2014], and Pico Island Faial A1 A2 B C D F G H Min 440 172 1739 1034 1401 1073 773 1239 Max 1174 727 3042 1434 3128 1827 1069 1728 Ratio (Max/Min) 3 4 212211 Terceira 1 2 3 4 5 6 NWTR SETR Min 820 300 644 1488 1546 2263 2658 3420 Max 3600 746 2092 2137 2635 5290 5697 6600 Ratio (Max/Min) 4 2 3 1 2 2 2 2 Pico A1 A2 B1 B2 C1 C2 D1 D2 Min 138 157 120 773 79 234 78 425 Max 1272 1283 2040 1851 377 1612 684 2335 Ratio (Max/Min) 9 8 17 25795

to min_W is high. Therefore, values higher than 5 appear to be good indicators of rejuvenation. The north- facing shelf sectors are generally wider than the south-facing sectors (Table 1), an observation that is related to the greater exposure to wave action (both in frequency and energy, see section 2.3). Nevertheless, the abnormally narrow shelves of subsectors C1 and D1 relatively to their north-facing counterparts is explained by the greater volcanic progradation to the south. Alternatively, they could also be young erosional shelves formed by surf erosion after recent and voluminous activity of the western Achada Plateau and Pico Vol- cano. The only exception is the southern Topo subsector B1 that is wider than its northern counterpart (B2); B2 is dominated by lava delta morphology, probably related with infilling of the landslide scar [Costa et al., 2014], which accounts for its smaller shelf width in comparison with B1. A conceptual model of how these rejuvenated shelves develop is proposed based on the most common examples found around Pico (Figure 7). The model is simplified by not including subaerial (explosive volcanism and stream erosion) and tectonic processes. However, it includes the two main processes (wave erosion and volcanic progradation) and their side effects (sedimentation, shelf width variation and minor slope failure at the shelf edge). The sequence starts with a volcanic island shelf formed by wave ero- sion during one or more glacial-interglacial sea level oscillations at a lowstand sea level position (SL1). From there it can evolve directly into i_a_3 or evolve first to a highstand sea level position with resultant cliff erosion and sediment deposition on the shelf (iii) and then on to a two end-member stage (iii_a_2 and iii_b_2): 1. From i to i_a_3. During the rise of sea level from the lowstand SL1 to the midstand SL2, significant vol- canic progradation occurs entirely filling the erosional shelf (i_a_1). The shelf edge is moved landward and upward (represented by the red star). As sea level continues to rise to the highstand SL3, wave ero- sion of the lava flow occurs and produces a sharp erosional surface above the level of SL2 (represented by the black dot in i_a_2 and i_a_3). When sea level reaches SL3, a sedimentary body starts to grow by erosion of the cliff and nearshore deposits of the lava delta and transport of the eroded material to the mid and outer shelf. If volcanic activity continues, the coastline is moved again offshore, further reducing the shelf width (i_a_3). From i to i_a_3 the shelf edge is moved landward and upward (now represented by the green star). 2. From iii to iii_a_2. During the highstand SL3, volcanic progradation partially fills the shelf (iii_a_1), which was already covered by a sedimentary body (iii). Erosion of the coastline lava delta produces a small cliff (iii_a_2). From iii to iii_a_2, the shelf width is reduced but the depositional shelf edge maintains its posi- tion (represented by the green star in iii_a_2). 3. From iii to iii_b_2. During the highstand SL3, a significant volcanic progradation occurs filling the entire shelf, overlapping a previous highstand sedimentary body (iii_b_1). This provokes instability of the shelf edge and upper slope and mass wasting occurs, moving the shelf edge landward and upward. After this event, erosion of the coastline lava delta and production of a small cliff occurs (iii_b_2). This model can be applied to other reefless volcanic island shelves where wave erosion and volcanic pro- gradation are the main development processes, with the latter prevailing over the former.

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1412 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

Figure 7. Conceptual model of the morphologic development of rejuvenated shelves (see section 4.1 for details). Each figure represents schematically a cross-shore profile of the shelf with very high vertical exaggeration. The model starts at i or iii and develops into the three possible stages i_a_3, iii_a_2, and iii_b_2. Inside the squares, dark gray areas represent the cliff (CLF or CLIFFS), the insular shelf (INSULAR SHELF), and the upper slope (SLP or SLOPE) of a shelf sector of the island, light gray represents erosion within those areas, yellow areas represent sandy sedimentary deposits, the orange areas preserved or eroded lava flows, rounded blocks are boulder fields that result from erosion of lava flows, blue lines represent sea level, black dots represent erosional (shelf) edges, red stars shelf edges formed by volcanic progradation and green stars depositional edges. Between the squares, blue curves represent sea level oscillations; blue dots position of sea level inside the curve; blue arrows the direction of sea level variation and equal signal, stable sea level. Black horizontal and vertical arrows represent time evolution.

4.2. Insular Shelf Sedimentation Most of the superficial samples collected on the shelves around Pico Island revealed coarse sands to gravels and no relationship between grain size and depth (see Figures S1_a1_i, S1_b1_i, S1_b1_iii, S1_c1_i, S1_c3_i, S1_d1_i, and S1_d3_i and text in the supporting information). The finer samples were found on more pro- tected bays and downstream of the mouths of small streams. Therefore, the lack of well-developed drain- age, the narrow shelves, and the presence of high wave energy does not allow the development of wave- graded shelves where the grain size of seafloor sediments decreases with distance to the shore as seen in other environments [Dunbar and Barrett, 2005; George and Hill, 2008]. Consequently and similarly to Faial island’s shelves [Quartau et al., 2012], there is no relationship between depth and grain size on these shelves.

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1413 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

4.3. The Shelf Morphological Record as Key for Understanding the Evolution of the Faial-Pico Ridge The shelf edge in the north part of the Faial-Pico channel (E4) has the same trend (NW-SE) as the onshore western Achada Plateau and its respective shelf edge (C2). The depth of the edge NW of Cedros volcano (G2) is extremely incised by gullies, but reaches depths of 300 m in the most preserved segments. These gullies together with well incised streams at the coastline probably reflect structural weaknesses related to the graben faults that extend to the west of the island (see dashed lines in Figure S1_f1 of the supporting information). The depth of the edges along all these sectors also shows that they have suffered subsidence, implying a relatively old age for their wave incision. The tectonic structures in the channel and the exten- sion of this graben system onto the western Achada Plateau as well as to the west Faial, the depth of the Matuyama body and the three high-density bodies (1, 2, and 4 in Figure 3b) in the channel, all trending NW-SE, point to an old structure controlling the construction of the proto Faial-Pico ridge. Hildenbrand et al. [2012b] suggested a ridge structure for the Ribeirinha volcano based on the extension of magnetic negative values west of Faial. However, the above evidence suggests that instead a subaerial ridge structure devel- oped before the construction of Ribeirinha volcano at 850 ka (a typical circular shield volcano [Madeira, 1998]) and extended below the present-day Faial-Pico ridge (see also the supporting information). Subsec- tors A1, A2, and C2 still show this ridge direction, however the depth of their edges is well above that of subsector E4. Continuous volcanic progradation has most likely erased the morphological evidence of the deep shelf edge of this old structure on subsectors A1, A2, and C2. The convex shelf edges at B1 and B2 sectors and their positions significantly deeper than 2123 m suggests that these sectors correspond to the incision of Topo volcano (see also sections S1.2 and S2.2 in the sup- porting information). In addition, the lack of convexity of the SW part of Topo suggests that also this part has collapsed (red dashed line in Figure 5b represents the likely location of the debris avalanche produced by this collapse). After the three collapses of Topo volcano (in the N, in the SW and in the SE), significant vol- canism from the fissural system of Achada Plateau filled the scars. In areas where morphological evidence of volcanic progradation exists, it is possible to roughly infer the timing of volcanic progradation of these shelves with the help of a sea level curve. We are assuming that erosional surfaces record the water level at that time (see section 3.2): 1. If the seafloor is covered by well-preserved lava flows or the basement under the sediments is irregular, volcanic progradation occurred after the LGM. The presence or not of sharp erosional features and their depth range marks the timing of the volcanic progradation. For instance, if sharp erosional features occur above 240 m, volcanic progradation occurred at 11.15 ka (using the sea level curve of Bintanja et al. [2005]). A typical example of this is i_a_2 of Figure 7 (the black dot marks the depth of the beginning of wave incision). If no sharp erosional features are present besides a small shore platform and cliffs behind, it means that volcanic progradation occurred after 6.5 ka when sea level approached the present-day level. A typical example of this is iii_a_2 of Figure 7. 2. Uncliffed or slightly cliffed subaerial lava deltas also imply that these were emplaced when sea level approached the present-day level, 6.5 ka or later. A typical example of this is iii_b_2 of Figure 7, which most likely corresponds in Pico to several undated examples, such as Ribeiras (Figure S1_b1_i) or S~ao Mateus (Figure S1_c1_i) lava deltas.

A reconstruction of the evolution of the Faial-Pico ridge is therefore proposed based on the information given by the shelf morphological analysis (see supporting information), known volcanic stratigraphy, tecton- ics, gravity data and magnetization values:

1. Development of the Faial-Pico volcanic ridge (Figure 8a) sometime before the emplacement of Ribeirinha volcano (i.e., before 850 ka).

2. Development of the proto island of Faial with the emplacement of the Ribeirinha volcano reaching maxi- mum elevation at 850 ka [Hildenbrand et al., 2012b] (Figure 8b). 3. Rejuvenation of volcanism in NE Faial around 400–350 ka [Hildenbrand et al., 2012b] (Figure 8c). 4. Development of volcanism in the intersection of the two subridges in Pico, with emersion of Topo shield volcano at 250 ka [Chovelon, 1982] (Figure 8d).

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1414 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

Figure 8. Evolutionary stages (a–j) of the Faial-Pico ridge (the positions of the shore line and shelf segments or gradient breaks are schematic) (see section 4.3 for details). Color polygons represent built up of successive volcanic edifices and have the same color legend as in Figure 2. The dashed black line represents the position of the shore line after volcanic construc- tion and/or wave erosion during the time elapsed between each stage. The color lines offshore represent the same shelf edge or gradient break segments as in Figure 5b. Circles and arrows represent, respectively, volcanic cones and direction of the flows issuing from them.

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1415 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

Figure 9. Mapping of the onshore areas that can potentially collapse because lava deltas have reached the shelf break (in red). Legend is the same as Figures 5a and 5b.

5. Rejuvenation of volcanism in W Faial with the development of the Caldeira volcano at 120 ka [Hilden- brand et al., 2012b] (Figure 8e). 6. Collapses of the northern and southern flanks of the Topo volcano. The minimum age estimation of the northern collapse is 70 ka [Costa et al., 2014] (Figure 8f). 7. New volcanic phase with the development of the Pico stratovolcano sometime before 53 ka [Costa et al., 2014]. Infilling of the Topo volcano landslide scars by reactivation of the volcanism of the western and eastern Achada Plateau ridge (Figure 8g). 8. Collapse of the south flank of Pico volcano sometime after 53 ka and eruption of the Almoxarife Forma- tion in Faial at 30 ka [Feraud et al., 1980] (Figure 8h). 9. During the last sea level rise coastlines receded and insular shelves widened around the volcanic edifices. Simultaneously, volcanic progradation occurred and filled previous erosional spaces on the shelves of Faial and Pico. Rejuvenation of volcanism in W Faial with the development of the Capelo Peninsula around 8 ka [Di Chiara et al., 2014] (Figure 8i). 10. Development of post human-settlement lava deltas in Pico (Ponta do Misterio, S~ao Jo~ao, Santa Luzia, and Silveira). In Faial Island eruptions occurred in 1672–1674 and in 1957–1958 A.D [Madeira and Brum da Silveira, 2003] (Figure 8j).

4.4. Implications for Hazard Assessment In several places where the lava deltas reached the shelf edge in Pico Island, they have apparently triggered small slope failure and a consequent retreat of the edge (as the example iii_b_2 of Figure 7). This is in accord- ance with reports by other authors of the effects of lava deltas reaching steep submarine slopes, covered by unstable volcanic aprons [Bosman et al., 2014; Chiocci et al., 2008; Poland and Orr, 2014; Sansone and Smith, 2006]. Mapping the seabed offshore of these structures may therefore provide an indication of their potential for land collapse. Figure 9 shows the areas where lava progradation was voluminous enough to reach the shelf edge, triggering slope failures and provoking its retreat. These areas are potentially unstable and this can be a simple and time-effective way to assess likely future collapses, perhaps more efficient than using time series InSAR or GPS data.

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1416 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

5. Conclusions Multibeam bathymetry, seismic reflection profiles, and sediment samples from the nearshore submarine areas of Faial and Pico islands were used to map and characterize the morphology of the shelf environ- ments. These insular shelves were formed by wave erosion of the flanks of the volcanic edifices of Faial and Pico islands. Pico insular shelves have been significantly modified by later volcanic progradation. Nevertheless, most of Pico’s insular shelves still have edges at depths below 123 m (the LGM sea level) and, given that the adjacent volcanoes are older than the LGM they were probably created before and have subsided. The morphology of Pico insular shelves, particularly expressed by the depth of their edges, well-preserved lava flow morphologies and the nature of their basement (sharp or irregular) beneath the sediments sug- gests whether they were last affected by wave erosion or volcanic progradation. This allowed us to recon- struct the history of these shelves and to put forward a conceptual model of the post-LGM morphologic development of the Pico shelves. They show a dominant role of volcanic progradation in their evolution, prevailing over wave erosion and allow us to classify many of these shelves as rejuvenated. In places where volcanic progradation has not been homogeneous along the coastline, a dimensionless variable (the ratio between the maximum and minimum shelf width) is proposed to discriminate between mainly erosional or mainly rejuvenated shelves. The ratio shows that Pico shelves are mostly rejuvenated when compared with the majority of the Faial and Terceira shelves. In addition, the history of these shelves together with the dis- tribution of the gravity data, magnetization values and the onshore geomorphology were used to develop an evolutionary model for the Faial-Pico ridge. The interpretation of these data suggests that the first subae- rial volcanism along this ridge is older than has been reported, i.e., before 850 ka. Lava deltas reaching the shelf break around Pico Island have triggered small mass-wasting events and retreat of the shelf edges. At Ribeiras, this lava delta seems to be unstable and subsiding faster than the sur- Acknowledgments Seismic acquisition and sediment rounding areas (see text in supporting information). Collecting multibeam bathymetry can reveal where sampling of Pico shelves were these lava deltas have reached or not the shelf edge, hence providing an effective way, complementary to conducted in the scope of the GEMAS ground deformation monitoring, to assess hazard and produce maps of onshore areas that have a height- Project funded by Secretaria Regional do Ambiente dos Ac¸ores. The ened risk for human occupation. following agencies are gratefully acknowledged for providing funding These results improve the understanding of the processes shaping island shelves at reefless volcanic ocean for the multibeam surveys: the Royal islands. In addition, the morphological analysis of the shelves provides a significant complement to unveil- Society, the British Council, the Higher ing the geological history of the islands as well as information for risk assessment and coastal land use Education Funding Council for Wales, the Regional Directorate for Science management. and Technology of the Azores and Portuguese projects MARINOVA (INTERREG IIIb/MAC/4.2/M11), MAROV, References MAYA (AdI/POSI/2003), and MeshAtlantic (AA-10/1218525/BF). This Agencia Estatal de Meteorologıa, Madrid and Instituto de Meteorologia de Portugal, Lisboa (2012), Climate Atlas of the Archipelagos of the work was also supported by IDL , Madeira and the Azores, 78 pp. through UID/GEO/50019/2013, Andrade, C., R. M. Trigo, M. C. Freitas, M. C. Gallego, P. Borges and A. M. Ramos (2008), Comparing historic records of storm frequency and financed by Fundac¸~ao para a Ciencia^ e the North Atlantic Oscillation (NAO) chronology for the Azores region, Holocene, 18(5), 745–754. € a Tecnologia (FCT). The authors also Babonneau, N., C. Delacourt, R. Cancouet, E. Sisavath, P. Bache`lery, A. Mazuel, S. J. Jorry, A. Deschamps, J. Ammann, and N. Villeneuve acknowledge the support of Landmark (2013), Direct sediment transfer from land to deep-sea: Insights into shallow multibeam bathymetry at La Reunion Island, Mar. Geol., Graphics Corporation to IPMA via the 346, 47–57. Landmark University grant Program Bintanja, R., R. S. W. van de Wal, and J. Oerlemans (2005), Modelled atmospheric temperatures and global sea levels over the past million and CARIS, Inc., for the CARIS HIPS and years, Nature, 437(7055), 125–128. SIPS license granted under the Borges, P. (2003), Ambientes litorais nos Grupos Central e Oriental do arquipelago dos Ac¸ores, PhD thesis, 413 pp., Univ. dos Ac¸ores, Ponta academic agreement with Delgada, Portugal. IMAR-Instituto do Mar (Ref.: 2009-02- Bosman, A., D. Casalbore, C. Romagnoli, and F. Chiocci (2014), Formation of an aa lava delta: Insights from time-lapse multibeam bathyme- APP08). Rui Quartau and Fernando try and direct observations during the Stromboli 2007 eruption, Bull. Volcanol., 76(7), 1–12. Tempera acknowledge, respectively, Camacho, A. G., J. C. Nunes, E. Ortiz, Z. Franca, and R. Vieira (2007), Gravimetric determination of an intrusive complex under the Island of the Ciencia^ 2008 research contract and Faial (Azores): Some methodological improvements, Geophys. J. Int., 171(1), 478–494. the post-doc grant ref. SFRH/BPD/ Cannat, M., A. Briais, C. Deplus, J. Escartin, J. Georgen, J. Lin, S. Mercouriev, C. Meyzen, M. Muller, and G. Pouliquen (1999), Mid-Atlantic 79801/2011 funded by FCT. We finally Ridge-Azores hotspot interactions: Along-axis migration of a hotspot-derived event of enhanced magmatism 10 to 4 Ma ago, Earth wish to express our gratitude for the Planet. Sci. Lett., 173(3), 257–269. constructive reviews by Claudia Casalbore, D., C. Romagnoli, A. Bosman, and F. L. Chiocci (2011), Potential tsunamigenic landslides at Stromboli Volcano (Italy): Insight Romagnoli and Christian Huebscher from marine DEM analysis, Geomorphology, 126(1–2), 42–50. which helped to significantly improve Casalbore, D., C. Romagnoli, A. Pimentel, R. Quartau, D. Casas, G. Ercilla, A. Hipolito, A. Sposato, and F. L. Chiocci (2015), Volcanic, tectonic the manuscript. Data can be provided and mass-wasting processes of offshore (Azores) revealed by high-resolution seafloor mapping, Bull. Volcanol., 77, 24, on request to Rui Quartau doi:10.1007/s00445-015-0905-3. ~ ([email protected]). Catalao, J., G. Nico, R. Hanssen, and C. Catita (2011), Merging GPS and atmospherically corrected InSAR data to Map 3-D terrain displace- ment velocity, IEEE Trans. Geosci. Remote Sens., 49(6), 2354–2360.

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1417 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

Chiocci, F. L., C. Romagnoli, P. Tommasi, and A. Bosman (2008), The Stromboli 2002 tsunamigenic submarine slide: Characteristics and pos- sible failure mechanisms, J. Geophys. Res., 113, B10102, doi:10.1029/2007JB005172. Chiocci, F. L., et al. (2013), Bathy-morphological setting of Terceira Island (Azores) after the FAIVI cruise, J. Maps, 9(4), 590–595. Chovelon, P. (1982), Evolution volcanotectonique des iles de Faial et de Pico, Archipel des Ac¸ores—Atlantique Nord, The`se de Docteur 3e`me Cycle thesis, 193 pp., Univ. de Paris-Sud, Paris. Costa, A. C. G., F. O. Marques, A. Hildenbrand, A. L. R. Sibrant, and C. M. S. Catita (2014), Large-scale flank collapses in a steep volcanic ridge: Pico-Faial Ridge, Azores Triple Junction, J. Volcanol. Geotherm. Res., 272, 111–125. Cruz, J. V., and M. O. Silva (2001), Hydrogeologic framework of Pico island, Azores, Portugal, Hydrogeol. J., 9, 177–189. Di Chiara, A., F. Speranza, M. Porreca, A. Pimentel, F. D’Ajello Caracciolo, and J. Pacheco (2014), Constraining chronology and time-space evolution of Holocene volcanic activity on the Capelo Peninsula (Faial Island, Azores): The paleomagnetic contribution, Geol. Soc. Am. Bull., 126(9–10), 1164–1180. Dias, N. A., L. Matias, N. Lourenco, J. Madeira, F. Carrilho and J. L. Gaspar (2007), Crustal seismic velocity structure near Faial and Pico Islands (AZORES), from local earthquake tomography, Tectonophysics, 445 (3-4), 301–317. Dickson, M. E. (2004), The development of talus slopes around Lord Howe island and implications for the history of island planation, Aust. Geogr., 35(2), 223–238. Dunbar, G. B., and P. J. Barrett (2005), Estimating palaeobathymetry of wave-graded continental shelves from sediment texture, Sedimentol- ogy, 52, 253–269. Feraud, G. (1980), Contribution a la datation du volcanisme de l’archipel des Ac¸ores par la methode Potassium-Argon, Consequences geodynamiques, PhD thesis, 186 pp., Univ. de Paris-Sud, Paris. Feraud, G., I. Kaneoka, and C. J. Alle`gre (1980), K-Ar ages and stress pattern in the Azores: Geodynamic implications, Earth Planet. Sci. Lett., 46(2), 275–286. Fletcher, C. H., et al. (2008), Geology of Hawaii reefs, in Coral Reefs of the USA. Coral Reefs of the World 1, edited by B. M. Riegl and R. E. Dodge, chap. 11, pp. 435–488, Springer, N. Y. Garcia, M. O., E. H. Haskins, E. M. Stolper, and M. Baker (2007), Stratigraphy of the Hawai’i Scientific Drilling Project core (HSDP2): Anatomy of a Hawaiian shield volcano, Geochem. Geophys. Geosyst., 8, Q02G20, doi:10.1029/2006GC001379. Gente, P., J. Dyment, M. Maia, and J. Goslin (2003), Interaction between the Mid-Atlantic Ridge and the Azores hot spot during the last 85 Myr: Emplacement and rifting of the hot spot-derived plateaus, Geochem. Geophys. Geosyst., 4(10), 8514, doi:10.1029/2003GC000527. George, D. A., and P. S. Hill (2008), Wave climate, sediment supply and the depth of the sand-mud transition: A global survey, Mar. Geol., 254(3–4), 121–128. Glass, J. B., D. J. Fornari, H. F. Hall, A. A. Cougan, H. A. Berkenbosch, M. L. Holmes, S. M. White, and G. De La Torre (2007), Submarine volcanic morphology of the western Galapagos based on EM300 bathymetry and MR1 side-scan sonar, Geochem. Geophys. Geosyst., 8, Q03010, doi:10.1029/2006GC001464. Hernandez-Molina, F. J., L. Somoza, J. Rey, and L. Pomar (1994), Late Pleistocene-Holocene sediments on the Spanish continental shelves: Model for very high resolution sequence stratigraphy, Mar. Geol., 120(3–4), 129–174. Hildenbrand, A., F. O. Marques, J. Catal~ao, C. M. S. Catita, and A. C. G. Costa (2012a), Large-scale active slump of the southeastern flank of Pico Island, Azores, Geology, 40(10), 939–942. Hildenbrand, A., F. O. Marques, A. C. G. Costa, A. L. R. Sibrant, P. F. Silva, B. Henry, J. M. Miranda, and P. Madureira (2012b), Reconstructing the architectural evolution of volcanic islands from combined K/Ar, morphologic, tectonic, and magnetic data: The Faial Island example (Azores), J. Volcanol. Geotherm. Res., 241–242, 39–48. Hildenbrand, A., F. O. Marques, J. Catal~ao, C. M. S. Catita, and A. C. G. Costa (2013), Large-scale active slump of the southeastern flank of Pico Island, Azores: Reply, Geology, 41(12), e302. Hildenbrand, A., D. Weis, P. Madureira, and F. O. Marques (2014), Recent plate re-organization at the Azores Triple Junction: Evidence from combined geochemical and geochronological data on Faial, S. Jorge and Terceira volcanic islands, Lithos, 210–211, 27–39. Hipolito, A., J. Madeira, R. Carmo, and J. L. Gaspar (2013), Neotectonics of Graciosa island (Azores): A contribution to seismic hazard assess- ment of a volcanic area in a complex geodynamic setting, Ann. Geophys., 56(6), 1–18. Holcomb, R. T., and R. C. Searle (1991), Large landslides from oceanic volcanoes, Mar. Geotechnol., 10, 19–32. Instituto Geografico do Exercito (2001a), Portuguese military map, Horta (Faial-Ac¸ores), sheet 7, scale 1:25,0000, Lisboa. Instituto Geografico do Exercito (2001b), Portuguese military map, Piedade (Pico-Ac¸ores), sheet 13, scale 1:25,0000, Lisboa. Instituto Geografico do Exercito (2001c), Portuguese military map, (Pico-Ac¸ores), sheet 12, scale 1:25,0000, Lisboa. Instituto Geografico do Exercito (2001d), Portuguese military map, Prainha de Cima (Pico-Ac¸ores), sheet 9, scale 1:25,0000, Lisboa. Instituto Geografico do Exercito (2001e), Portuguese military map, S. Roque do Pico (Pico-Ac¸ores), sheet 8, scale 1:25,0000, Lisboa. Instituto Geografico do Exercito (2001f), Portuguese military map, S. Mateus (Pico-Ac¸ores), sheet 11, scale 1:25,0000, Lisboa. Instituto Geografico do Exercito (2001g), Portuguese military map, Candelaria (Pico-Ac¸ores), sheet 10, scale 1:25,0000, Lisboa. Instituto Geografico do Exercito (2001h), Portuguese military map, Pedro Miguel (Faial-Ac¸ores), sheet 5, scale 1:25,0000, Lisboa. Instituto Geografico do Exercito (2001i), Portuguese military map, Praia do Norte (Faial-Ac¸ores), sheet 4, scale 1:25,0000, Lisboa. Instituto Geografico do Exercito (2001j), Portuguese military map, Feteira (Faial-Ac¸ores), sheet 6, scale 1:25,0000, Lisboa. Instituto Hidrografico (Ed.) (1999), Hydrographic chart, Ilha do Faial e Canal do Faial, 1st ed., Lisboa. Instituto Hidrografico (2005), Tratamentos de dados de agitac¸~ao marıtima, Ac¸ores/Terceira—Fevereiro 2005, report, PJ OC 22EO05, Lisboa. Kelfoun, K., T. Giachetti, and P. Labazuy (2010), Landslide-generated tsunamis at Reunion Island, J. Geophys. Res., 115, F04012, doi:10.1029/ 2009JF001381. Kennedy, D. M., C. D. Woodroffe, B. G. Jones, M. E. Dickson, and C. V. G. Phipps (2002), Carbonate sedimentation on subtropical shelves around Lord Howe Island and Balls Pyramid, Southwest Pacific, Mar. Geol., 188(3–4), 333–349. Krastel, S., H.-U. Schmincke, C. L. Jacobs, R. Rihm, T. P. Le Bas, and B. Alibes (2001), Submarine landslides around the Canary Islands, J. Geo- phys. Res., 106(B3), 3977–3997. Laughton, A. S., and R. B. Whitmarsh (1974), The Azores-Gibraltar plate boundary, in Geodynamics of Iceland and the North Atlantic Area, edited by L. Kristjansson, pp. 63–81, Reidel, Dordretch, Netherlands. Le Friant, A., C. L. Harford, C. Deplus, G. Boudon, R. J. S. Sparks, R. A. Herd, and J.-C. Komorowski (2004), Geomorphological evolution of Monserrat (West Indies): Importance of flank collapse and erosional processes, J. Geol. Soc. London, 161, 147–160. Le Friant, A., E. Lebas, V. Clement, G. Boudon, C. Deplus, B. de Voogd, and P. Bachelery (2011), A new model for the evolution of La Reunion volcanic complex from complete marine geophysical surveys, Geophys. Res. Lett., 38, L09312, doi:10.1029/2011GL047489. Llanes, P., R. Herrera, M. Gomez, A. Munoz,~ J. Acosta, E. Uchupi, and D. Smith (2009), Geological evolution of the volcanic island La Gomera, Canary Islands, from analysis of its geomorphology, Mar. Geol., 264(3–4), 123–139.

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1418 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

Lourenc¸o, N., J. M. Miranda, J. F. Luıs, A. Ribeiro, L. A. Mendes Victor, J. Madeira, and H. D. Needham (1998), Morpho-tectonic analysis of the Azores Volcanic Plateau from a new bathymetric compilation of the area, Mar. Geophys. Res., 20(3), 141–156. Louvat, P., and C. J. Alle`gre (1998), Riverine erosion rates on S~ao Miguel volcanic island, Azores archipelago, Chem. Geol., 148, 177–200. Luis, J. F. (2007), Mirone: A multi-purpose tool for exploring grid data, Comput. Geosci., 33(1), 31–41. Luis, J. F., J. M. Miranda, A. Galdeano, P. Patriat, J. C. Rossignol, and L. A. Mendes Victor (1994), The Azores triple junction evolution since 10 Ma from an aeromagnetic survey of the Mid-Atlantic Ridge, Earth Planet. Sci. Lett., 125(1–4), 439–459. Madeira, J. (2005), The volcanoes of Azores island: A world-class heritage (examples from Terceira, Pico and Faial Islands), in IV International Symposium ProGEO on the Conservation of the Geological Heritage: Field Trip Book, edited by L. Lattex, 104 pp., Eur. Assoc. for the Con- serv. of the Geol. Heritage and Cent. de Geocienc.^ da Univ. do Minho, Braga, Portugal. Madeira, J., and A. Brum da Silveira (2003), Active tectonics and first paleosismological results in Faial, Pico e S. Jorge islands (Azores, Portu- gal), Ann. Geophys., 46(5), 733–761. Madeira, J. (1998), Estudos de neotectonica nas ilhas do Faial, Pico e S. Jorge: Uma contribuic¸~ao para o conhecimento geodin^amico da junc¸~ao tripla dos Ac¸ores, PhD thesis, Dep. de Geologia, Univ. de Lisboa, Lisboa. Madeira, J., A. M. M. Soares, A. B. D. Silveira, and A. Serralheiro (1995), Radiocarbon dating recent volcanic activity on Faial Island (Azores), Radiocarbon, 37(2), 139–147. Marques, F. O., J. C. Catal~ao, C. DeMets, A. C. G. Costa, and A. Hildenbrand (2013), GPS and tectonic evidence for a diffuse plate boundary at the Azores Triple Junction, Earth Planet. Sci. Lett., 381, 177–187. Meireles, R., R. Quartau, R. S. Ramalho, A. C. Rebelo, J. Madeira, V. Zanon, and S. P. Avila (2013), Depositional processes on oceanic island shelves—Evidence from storm-generated Neogene deposits from the mid-North Atlantic, Sedimentology, 60, 1769–1785. Menard, H. W. (1983), Insular erosion, isostasy, and subsidence, Science, 220, 913–918. Menard, H. W. (1986), Islands, 230 pp., Sci. Am. Books, N. Y. Miranda, J. M., J. F. Luıs, I. Abreu, L. A. M. Victor, A. Galdeano, J. C. Rossignol (1991), Tectonic framework of the Azores triple junction, Geo- phys. Res. Lett., 18, 1421–1424. Mitchell, N. C. (2003), Susceptibility of mid-ocean ridge volcanic islands and to large-scale landsliding, J. Geophys. Res., 108(B8), 2397, doi:10.1029/2002JB001997. Mitchell, N. C., D. G. Masson, A. B. Watts, M. J. R. Gee, and R. Urgeles (2002), The morphology of the flanks of volcanic ocean islands: A com- parative study of the Canary and Hawaiian hotspot islands, J. Volcanol. Geotherm. Res., 115, 83–107. Mitchell, N. C., C. Beier, P. Rosin, R. Quartau, and F. Tempera (2008), Lava penetrating water: Submarine lava flows around the coasts of Pico Island, Azores, Geochem. Geophys. Geosyst., 9, Q03024, doi:10.1029/2007GC001725. Mitchell, N. C., R. Quartau, and J. Madeira (2012a), Assessing landslide movements in volcanic islands using near-shore marine geophysical data: South Pico Island, Azores, Bull. Volcanol., 74(2), 483–496. Mitchell, N. C., R. Stretch, C. Oppenheimer, D. Kay, and C. Beier (2012b), Cone morphologies associated with shallow marine eruptions: East Pico Island, Azores, Bull. Volcanol., 74(10), 2289–2301. Mitchell, N. C., R. Quartau, and J. Madeira (2013), Large-scale active slump of the southeastern flank of Pico Island, Azores: Comment, Geol- ogy, 41(12), e301. Moore, J. G., W. R. Normark, and R. T. Holcomb (1994), Giant Hawaiian Landslides, Annu. Rev. Earth Planet. Sci., 22(1), 119–144. Nunes, J. C. (1999), A actividade vulc^anica na ilha do Pico do Plistocenico Superior ao Holocenico: Mecanismo eruptivo e hazard vulc^anico, PhD thesis, 357 pp., Univ. dos Ac¸ores, , Portugal. Nunes, J. C., A. Camacho, Z. Franc¸a, F. G. Montesinos, M. Alves, R. Vieira, E. Velez, and E. Ortiz (2006), Gravity anomalies and crustal signature of volcano-tectonic structures of Pico Island (Azores), J. Volcanol. Geotherm. Res., 156(1–2), 55–70. Parker, R. L., and S. P. Huestis (1974), The inversion of magnetic anomalies in the presence of topography, J. Geophys. Res., 79(11), 1587– 1593. Poland, M. P., and T. R. Orr (2014), Identifying hazards associated with lava deltas, Bull. Volcanol., 76(12), 1–10. Quartau, R., F. Curado, T. Cunha, L. Pinheiro, and J. H. Monteiro (2002a), Projecto Gemas-Localizac¸~ao e distribuic¸~ao de areias em redor da ilha do Faial, Tech. Rep. INGMARDEP 5/2002, Dep. Geol. Mar. IGM, Lisboa. Quartau, R., and F. Curado (2002b), Relatorio da Campanha FAPI2 realizada ao largo do Pico, Tech. Rep. INGMARDEP 15/2002, 17 pp., Dep. of Geol. Mar. IGM, Alfragide, Portugal. Quartau, R., and N. C. Mitchell (2013), Comment on ‘‘Reconstructing the architectural evolution of volcanic islands from combined K/Ar, morphologic, tectonic, and magnetic data: The Faial Island example (Azores)’’ by Hildenbrand et al. (2012) [J. Volcanol. Geotherm. Res., 241–242 (2012), 39–48], J. Volcanol. Geotherm. Res., 255, 124–126. Quartau, R., H. Duarte, and P. Brito (2005), Projecto Gemas–Relatorio da campanha de amostragem de sedimentos (FAPI-3) realizada na plataforma e na orla costeira das ilhas do Faial e do Pico, Tech. Rep. INGMARDEP 2/2005, Dep. of Geol. Mar. INETI, Alfragide, Portugal. Quartau, R., A. S. Trenhaile, N. C. Mitchell, and F. Tempera (2010), Development of volcanic insular shelves: Insights from observations and modelling of Faial Island in the Azores Archipelago, Mar. Geol., 275(1–4), 66–83. Quartau, R., F. Tempera, N. C. Mitchell, L. M. Pinheiro, H. Duarte, P. O. Brito, R. Bates, and J. H. Monteiro (2012), Morphology of the Faial Island shelf (Azores): The interplay between volcanic, erosional, depositional, tectonic and mass-wasting processes, Geochem. Geophys. Geosyst., 13, Q04012, doi:10.1029/2011GC003987. Quartau, R., A. Hipolito, C. Romagnoli, D. Casalbore, J. Madeira, F. Tempera, C. Roque, and F. L. Chiocci (2014), The morphology of insular shelves as a key for understanding the geological evolution of volcanic islands: Insights from Terceira Island (Azores), Geochem. Geo- phys. Geosyst., 15, 1801–1826, doi:10.1002/2014GC005248. Quidelleur, X., J. Carlut, V. Soler, J. P. Valet, and P. Y. Gillot (2003), The age and duration of the Matuyama–Brunhes transition from new K– Ar data from La Palma (Canary Islands) and revisited 40Ar/39Ar ages, Earth Planet. Sci. Lett., 208(3–4), 149–163. Ramalho, R. S., R. Quartau, A. S. Trenhaile, N. C. Mitchell, C. D. Woodroffe, and S. P. Avila (2013), Coastal evolution on volcanic oceanic islands: A complex interplay between volcanism, erosion, sedimentation, sea-level change and biogenic production, Earth Sci. Rev., 127, 140–170. Romagnoli, C. (2013), Chapter 3 Characteristics and morphological evolution of the Aeolian volcanoes from the study of submarine por- tions, Geol. Soc. Mem., 37, 13–26. Rusu, L., and C. Guedes Soares (2012), Wave energy assessments in the Azores islands, Renewable Energy, 45, 183–196. Sansone, F. J., and J. R. Smith (2006), Rapid mass wasting following nearshore submarine volcanism on Kilauea volcano, Hawaii, J. Volcanol. Geotherm. Res., 151(1–3), 133–139. Schilling, J.-G. (1975), Azores mantle blob rare earth evidence, Earth Planet. Sci. Lett., 25, 103–115.

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1419 Geochemistry, Geophysics, Geosystems 10.1002/2015GC005733

Stretch, R. C., N. C. Mitchell, and R. A. Portaro (2006), A morphometric analysis of the submarine volcanic ridge south-east of Pico Island, Azores, J. Volcanol. Geotherm. Res., 156(1–2), 35–54. Tempera, F. (2008), Benthic habitats of the extended Faial Island shelf and their relationship to geologic, oceanographic and infralittoral biologic features, PhD thesis, 311 pp., Univ. of St. Andrews, St. Andrews, U. K. http://hdl.handle.net/10023/726. Trenhaile, A. S. (2001), Modelling the Quaternary evolution of shore platforms and erosional continental shelves, Earth Surf. Processes Land- forms, 26, 1103–1128. Trenhaile, A. S. (2000), Modeling the development of wave-cut platforms, Mar. Geol., 166, 163–178. Trippanera, D., M. Porreca, J. Ruch, A. Pimentel, V. Acocella, J. Pacheco, and M. Salvatore (2014), Relationships between tectonics and mag- matism in a transtensive/transform setting: An example from Faial Island (Azores, Portugal), Geol. Soc. Am. Bull., 126(1–2), 164–181. Wentworth, C. K. (1922), A scale of grade and class terms for clastic systems, J. Geol., 30, 377–392. Weston, F. S. (1964), List of recorded volcanic eruptions in the Azores with brief reports, Bol. Mus. Lab. Min. Geol. Fac. Ci^enc. Lisboa, 10(1), 3– 18. Youssef, W. B. H. (2005), Caracterisation in situ de l’environnement physique des habitats benthiques occupes par Codium elisabethae au sein de sites d’etudes particuliers (Faial, Ac¸ores), MS thesis, 94 pp., Univ. de Lie`ge, Lie`ge, Belgium. Zbyszewski, G. (1963), Les Phenome`nes Volcaniques Modernes Dans L’archipel des Ac¸ores, Com. Serv. Geol. Portugal. 47, 227 pp.

Erratum

In the originally published version of this article, there were some typographical errors that have now been corrected. In the first sentence in paragraph 2 of Section 2.1, the location has been clarified. The following references have been updated: Instituto Geografico do Exercito (2001i), Portuguese military map, Praia do Norte (Faial-Ac¸ores), sheet 4, scale 1:25,0000, Lisboa. Instituto Geografico do Exercito (2001j), Portuguese military map, Feteira (Faial-Ac¸ores), sheet 6, scale 1:25,0000, Lisboa. Quartau, R., F. Curado, T. Cunha, L. Pinheiro, and J. H. Monteiro (2002a), Projecto Gemas-Localizac¸~ao e distribuicc¸~ao de areias em redor da ilha do Faial, Tech. Rep. INGMARDEP 5/2002, Dep. Geol. Mar. IGM, Lisboa.

QUARTAU ET AL. THE INSULAR SHELVES OF FAIAL-PICO RIDGE 1420