Recent Macaronesian kinematics from GNSS ground displacement analysis

Stud. Geophys. Geod., 65 (2021), DOI: 10.1007/s11200-020-1122-x, in print i © 2021 Inst. Geophys. CAS, Prague Recent Macaronesian kinematics 1 from GNSS ground displacement analysis 2 3 I GNACIO B ARBERO 1,7 , C RISTINA T ORRECILLAS 2 , R AÚL P ÁEZ 3 , G ONÇALO P RATES 4,5,6 4 AND M ANUEL B ERROCOSO 4 5 6 1 Escuela Técnica Superior de Ingeniería, Universidad de Sevilla, Sevilla, Spain 7 (ignacio_barber[email protected]) 8 2 Departamento de Ingeniería Gráfica, Escuela Técnica Superior de Ingeniería, Universidad de 9 Sevilla, Spain ([email protected]s) 10 3 Departamento de Estadística e Investigación Operativa, Facultad de Ciencias, Universidad de 11 Cádiz, Spain (raul.p[email protected]) 12 4 Laboratorio de Astronomía, Geodesia y Cartografía. Universidad de Cádiz, Puerto Real, 13 Cádiz, Spain ([email protected]) 14 5 Centro de Estudos Geográficos, IGOT, Universidade de Lisboa, Lisboa, Portugal 15 6 Instituto Superior de Engenharia, Universidade do Algarve, Faro, Portugal ([email protected]) 16 7 Ministerio de Hacienda, Dirección General del Catastro, Spain 1 17 18 Received: March 21, 2020; Revised: August 24, 2020; Accepted: November 5, 2020 19 ABSTRACT 20 Macaronesia is a complex oceanic region spanning three tectonic plates in the north-21 east Atlantic ocean. It is composed of four archipelagos, widely distributed and limited to 22 the east by the Iberian Peninsula and north-western coast of Africa. This study aims to 23 clarify recent Macaronesian kinematics from 19 GNSS stations located on the four 24 archipelagos and the Iberian and African coastlines. The analysis is based on nearly 15 25 years of common data acquisition and aimed to detect new effects of intraplate tectonics 26 or similar local/regional events consistent with calculated ground displacements. 27 Evaluating the GNSS stations residual velocities relative to those expected from the NNR-28 MORVEL56 model, higher residuals were found at continental coastal stations (Africa) 29 than at oceanic ones (Canaries and Madeira). From the computed strain rate map, the 30 possible existence of a shear zone connecting the Gloria and Transmoroccan fault 31 systems, already mentioned by other authors, was depicted. Cluster statistical analysis of 32 the horizontal residual velocities helped to identify tectonic boundaries in Macaronesia 33 and four groups of analogous intraplate residual velocities within this region. Three of 34 four groups were identified in the Azores, highlighting the African-Nubian-Eurasian 35 diffuse plate boundary in this region. Furthermore, in the Canary Islands, two distinct 36 kinematic behaviours were detected, possibly due to the activity along a previously 37 detected tectonic fault between Tenerife and Gran Canaria, where some stations have 38 similar intraplate residuals to those at Madeira and Cape Verde stations, while others 39 have similar intraplate residuals to those of continental stations. Finally, all stations on 40 I. Barbero et al. ii Stud. Geophys. Geod., 65 (2021) oceanic crust, except Cape Verde, present recent ground subsidence which may be 1 attributed to isostatic adjustment. 2 3 Keywords: continuous GNSS time series, Canary Islands, cluster analysis, strain 4 rate analysis, isostatic adjustment 5 1. INTRODUCTION 6 Macaronesia is the collective name given to the group of islands that lie in the north-7 east Atlantic ocean between 10°N and 40°N latitude (Fig. 1). These islands form four 8 archipelagos, which are, from north to south: the Azores, Madeira, the Canary Islands and 9 Cape Verde. The Mid-Atlantic Ridge (MAR) marks its western boundary and the Gloria 10 transform fault its northern boundary (Scheidegger, 2002). However, the Azores lie on 11 a triple junction, where the North American, Eurasian and Nubian Plates interact. 12 Although these archipelagos are widely distributed, these islands and the coastlines of 13 north-west Africa and the Iberian Peninsula are the available emerged land where Global 14 Navigation Satellite System (GNSS) observations can be made. 15 Fig. 1. Macaronesian GNSS stations and African-Nubian plate boundary elements. Focal mechanisms from the ISC catalogue, within each event from the ISC Bulletin for focal mechanism solutions from the International Seismological Centre (2019); between 40°W0°W and 15°N40°N and weighted by the earthquake magnitude. A seismic hypocentre depth limit < 70 km and a magnitude M w > 3.0 were established between 2000/01/01 and 2016/12/31. Standard regime of normal faulting (NF), thrust faulting (TF), strike-slip (SS) and unknown (U) is shown; ETOPO1 basemap is obtained from NOAA National Centres for Environmental Information (NCEI). Map is represented in Mercator projection over WGS84 Earth reference system. AFZ: Atlantis Fracture Zone, EAFZ: East Azores Fracture Zone, ATJ: Azores Triple Junction, TFS: Transmoroccan Fault System. See Table 1 for GNSS station codes. 2 Recent Macaronesian kinematics from GNSS … Stud. Geophys. Geod., 65 (2021) iii Only a few studies have focused on Macaronesian kinematics, since rigidity, 1 continuity and insignificant seismicity seem to be fundamental aspects of the main plate, 2 the African-Nubian, regardless of the activity of both the MAR and Nubian-Eurasian plate 3 boundaries (Fig. 2). Additionally, its low speed relative to its neighbouring plates of 4 nearly 1 mm yr 1 (Burke and Wilson, 1972; Kogan et al., 2000; Malservisi et al., 2013) 5 and its minimum internal strain, with no significant deformation with values <1 mm yr 1 6 (Calais et al., 2003; McClusky et al., 2003; Serpelloni et al., 2007), are other suitable 7 descriptions for this plate. The global stress map database (Heidbach et al., 2018) and 8 some physical models of Macaronesia (Geyer et al., 2016; Jiménez-Munt and Negredo, 9 2003) are other sources where low values of internal stress are shown. 10 Observations from continuous GNSS stations in Macaronesia started in 2000 with 11 a limited number of permanent stations. Most previous studies were carried out by mixing 12 observations from periodic or episodic GNSS campaigns in their calculations (Barbero et 13 al., 2018; Garcia et al., 2014; Marques et al., 2013a, 2015; Mendes et al., 2013; Pérez-14 Peña et al., 2010; Prates et al., 2013; Vernant et al., 2010). Overall, these locally or 15 regionally determined velocities are similar to the velocities of geologic plate motion 16 computed from global models such as NNR-MORVEL56 (DeMets et al., 2010) and NNR-17 NUVEL1A (DeMets et al., 1994), and azimuths only disagreeing slightly by a few 18 degrees. 19 Table 1. Continuous GNSS stations used with public access data and managed by public institutions: International GNSS Service (IGS, http://www.igs.org), EUREF Permanent GNSS Network (EPN, http://www.epncb.oma.be), Rede Nacional de Estações Permanentes GNSS (ReNEP, https://renep.dgterritorio.gov.pt/estacoes). GNSS Location Region Owner Start [year, day#] End [year, day#] 3 FLRS Flores Azores IGS 2008,159 2015,090 PDEL São Miguel Azores IGS 2000,241 2015,090 TERC Terceira Azores EPN 2008,016 2015,090 PIED Pico Azores ReNEP 2008,016 2012,177 FRNS São Miguel Azores ReNEP 2008,016 2014,005 AZGR Graciosa Azores ReNEP 2009,347 2012,177 FUNC Madeira Madeira IGS 2005,314 2015,090 MAS1 Gran Canaria Canary Islands IGS 2000,211 2015,090 IZAN Tenerife Canary Islands EPN 2008,110 2015,090 LPAL La Palma Canary Islands EPN 2001,179 2015,090 TGCV Sal Cape Verde IGS 2000,004 2014,155 CASC Portugal Iberian Peninsula EPN 2000,241 2015,090 LAGO Portugal Iberian Peninsula EPN 2000,229 2015,090 SFER Spain Iberian Peninsula EPN 2000,242 2015,090 VILL Spain Iberian Peninsula IGS 2000,116 2015,090 CEU1 Spain Africa EPN 2008,314 2015,090 TETN Morocco Africa EPN 2002,083 2015,090 RABT Morocco Africa EPN 2000,166 2015,090 DAKR Morocco Africa IGS 2012,209 2015,090 I. Barbero et al. iv Stud. Geophys. Geod., 65 (2021) In 2004, the unusual increase in seismic events in Tenerife, with some earthquakes felt 1 by the population, and the underwater eruption off the coast of El Hierro augmented 2 volcanic studies in the Canary Islands (Domínguez-Cerdeña et al., 2011; Garcia et al., 3 2014; Prates et al., 2013). One of these recent studies focused on Tenerife showed distinct 4 kinematics between this island and the neighbouring island of Gran Canaria, with 5 different horizontal velocities computed over 10 years (Barbero et al., 2018). The increase 6 of permanent GNSS stations in Macaronesia and their data accessibility since 2000 offer 7 the possibility of carrying out a precise ground displacement analysis of the whole Canary 8 archipelago and the opportunity of finding similarities and differences to the kinematics of 9 other archipelagos. Therefore, the purpose of this study is to evaluate the regional 10 geodynamics of the Canary Islands and provide an overall interpretation of the recent 11 intraplate kinematics of the African-Nubian Plate. For this, velocities between May 2000 12 and March 2015 were calculated for 19 stations, aiming at accuracy of less than 13 1 mm yr 1 over such long periods (Blewitt and Lavallée, 2002; Bos et al., 2010). 14 Furthermore, a statistical cluster analysis was performed to identify and group residuals 15 with similar velocity and, finally, strain maps were generated to define intraplate 16 kinematics in Macaronesia and, particularly, in the Canary Islands. 17 2. REGIONAL SETTING 18 Although all archipelagos in Macaronesia have a minimum of five islands and 19 volcanic origin, the nature of this volcanism is not always clear (Geldmacher et al., 2006; 20 Fig. 2. Macaronesia residual kinematic map for 20002015. The arrows represent residual horizontal and vertical velocities at GNSS stations after removing NNR-MORVEL56. Biases with a 95% confidence level are shown as ellipses for the horizontal and intervals for the vertical velocities. AFZ: Atlantis Fracture Zone, EAFZ: East Azores Fracture Zone, GB: , TFS: Transmoroccan Fault System. 4 See Table 1 for station codes. 5 Recent Macaronesian kinematics from GNSS … Stud. Geophys. Geod., 65 (2021) v Métrich et al., 2014; Pacheco et al., 2013; Ramalho, 2011). The ages of their old 1 seamounts date from 142 Myr to 200000 years (von den Bogaard, 2013). From the north-2 west, near the MAR, the first of the four archipelagos is the Azores which includes nine 3 islands aligned NW-SW (see Fig. 1). The two most westerly islands, Flores and Corvo, 4 are located on the North American plate. The other seven islands lie to the east of the 5 MAR, arranged in two groups: the central group includes Graciosa, Terceira, São Jorge, 6 Faial and Pico, while the eastern group includes São Miguel and Santa Maria. Both groups 7 are in the vicinity of the Eurasian-African-Nubian plate boundary. This region shows an 8 anomalously thick crust (Silveira et al., 2010) with modern volcanism (1957 Capelinhos 9 volcano in Faial and 19992000 underwater eruption near Terceira) (Pacheco et al., 2013) 10 and recent moderate to high magnitude seismicity (1980 - M7.2, 1997 - M5.1 and 1998 - 11 M6.2 earthquakes) (Borges et al., 2007). The most important geological structures are the 12 Terceira Rift, aligned by Graciosa, Terceira and São Miguel islands, and the East Azores 13 Fracture Zone (EAFZ). 14 The archipelagos of Madeira, the Canary Islands and Cape Verde are concentrated in 15 the east Atlantic Ocean away from plate boundaries. The Madeira archipelago comprises 16 Madeira, Porto Santo, Deserta Grande and Selvagem Grande islands. The most recent 17 volcanic eruptions occurred 6500 years ago. The Canary Islands archipelago is formed by 18 seven major islands: La Palma, El Hierro, La Gomera, Tenerife, Gran Canaria, 19 Fuerteventura and Lanzarote, located less than 100 km off the north-west coast of Africa. 20 In the past 500 years, volcanic eruptions have been recorded and documented at El Hierro, 21 Lanzarote, Tenerife and La Palma. The Cape Verde archipelago lies 450600 km off the 22 western coast of Africa and is composed of ten islands: Santo Antão, São Vicente, Santa 23 Luzia, São Nicolau, Sal, Boa Vista, Maio, Santiago, Fogo and Brava, arrayed in a west-24 facing horseshoe shape. The volcanic activity probably started in the Oligocene/Miocene 25 and extended well into the Holocene. Historical eruptions are unknown except on Fogo, 26 a highly active volcano whose last eruptions occurred in 1995 and 2014 (Ramalho, 2011). 27 3. RECENT NUBIAN KINEMATICS 28 The African plate consists of two subplates separated by the East African rift system: 29 the Nubian to the west and Somalian to the east. Most of Macaronesia is part of the 30 African-Nubian Plate. The main Atlantic fractures present in this area are related to 31 different velocities of the MAR opening revealed by two main transform faults: the Gloria 32 Fault, linked to the MAR by the EAFZ and the Terceira Rift (Fernandes et al., 2006), and 33 the Atlantis Fracture Zone (AFZ), extending from the MAR to the Transmoroccan Fault 34 System (TFS) and limiting the Morocco subplate (Mantovani et al., 2007) (Fig. 1). The 35 Gloria transform fault, which is part of the Nubian-Eurasian plate boundary, involves 36 predominantly right-lateral compression along the eastern segment, strike-slip along the 37 central segment and oblique extension on the westernmost segment in the Azores. 38 Mantovani et al. (2007) proposed another shear zone connecting the Gloria Fault and the 39 EAFZ propagation towards the TFS, marking the Atlantic-Mediterranean transition, 40 where the Nubian-Eurasian plate boundary changes to a diffuse transpressive region, 41 comprising a wide range of active deformation. Other authors have suggested that 42 a subduction zone is forming at the south-west of the Iberian Peninsula, as a result of both 43 I. Barbero et al. vi Stud. Geophys. Geod., 65 (2021) propagations of the Gibraltar Arc compressive stresses and pressures related to the large-1 scale Nubian-Eurasian convergence (Duarte et al., 2013). 2 Mostly, GNSS calculations in the African-Nubian Plate offer values with slight 3 velocity differences relative to global models, around ±1 mm yr 1 , but nevertheless show 4 greater differences in azimuth. Specifically, from campaign and continuous GNSS data 5 from 2001 until 2013, the African-Nubian Plate was inferred to move 4.6 ± 0.3 mm yr 1 6 toward S87.9°W (167.9°) ± 3.3° (95% uncertainty) relative to Eurasia in the Azores 7 (Marques et al., 2013a), similar to the NNR-MORVEL56 model value of 8 4.5 ± 0.4 mm yr 1 toward S68.1°W (148.1°) ± 2.8° (DeMets et al., 2010), mainly 9 depicting an azimuth difference of nearly 20°. Another relevant aspect is that the NNR-10 NUVEL1A model (DeMets et al., 1994) considers Terceira (station TERC, see Fig. 1), 11 Graciosa (station AZGR) and eastern São Miguel (station FRNS) as part of the Eurasian 12 Plate, and western São Miguel (station PDEL) and Pico (station PIED) as part of the 13 African-Nubian Plate. The horizontal velocity of the African-Nubian Plate computed by 14 the UNAVCO plate motion calculator (https://www.unavco.org/software/geodetic-15 utilities/plate-motion-calculator/plate-motion-calculator.html), with Eurasia as the 16 reference plate, was 4.04.2 mm yr 1 with an azimuth of S6575°W (245255°) for 17 both stations (PDEL and PIED). Within the central group, the Pico/Faial volcanic ridge 18 was shown to move mostly with the African-Nubian Plate, Terceira island moves mostly 19 with the Eurasia Plate, and São Jorge island has an intermediate position between African-20 Nubian and Eurasian Plates (Marques et al., 2013a; Mendes et al., 2013). Similar values 21 were calculated by Borges et al. (2007) who found an average seismic strain rate of about 22 4.4 mm yr 1 with a NW direction for the period 19801998. Subsidence was detected on 23 Terceira island between 1999 and 2006 based on GPS (Miranda et al., 2012) and also 24 between 2001 and 2013 by combining InSAR and GPS techniques (Marques et al., 2015). 25 Near Gibraltar, velocities in stations TENT, SFER and CEU1 present values between 26 4.4 and 4.7 mm yr 1 and azimuths of N4050°W (310320°) from NUVEL1A, with 27 Eurasia as the reference plate, using the UNAVCO Plate Motion Calculator, while 28 southwards horizontal velocity magnitude decreases to 3.9 mm yr 1 and an azimuth of 29 N41°W (319°) in station RABT. Koulali et al. (2011) inferred a south-westward motion of 30 the Rif Mountains in northern Morocco 6 , with velocities between 3.5 and 4 mm yr 1 . 31 McClusky et al. (2003) determined an NW-SE convergence within the Mediterranean, 32 with a larger westerly component of motion than NUVEL1A, with Eurasia as the 33 reference plate, ranging from 5.4 ± 1 mm yr 1 in the Eastern Mediterranean to 34 4.5 ± 1 mm yr 1 near Gibraltar. The study of Pérez-Peña et al. (2010) found slight 35 differences in azimuth between their calculations and NUVEL1A between 1995 and 2005 36 in Southern Spain. Finally, Vernant et al. (2010) showed that the deformation associated 37 with the Nubian-Eurasian plate boundary that occurs in the Betic-Rif-Alboran domain has 38 velocities of around 4 mm yr 1 , and showed azimuths in clockwise rotation between 39 CASC and RABT stations between 1997 and 2009. 40 In the region defined by Madeira, the Canary Islands and Cape Verde archipelagos, 41 horizontal velocities follow the MAR transform fault direction of SW-NE. A recent GPS 42 study from 19982005 showed similar velocities between MAS1 and the continental 43 TETN station and slight differences in azimuth between both and the TGCV station 44 Recent Macaronesian kinematics from GNSS … Stud. Geophys. Geod., 65 (2021) vii (Serpelloni et al., 2007). Martín et al. (2014) showed similar horizontal velocities in all 1 Canary Islands, between 2002 and 2009, so no strain within the archipelago was detected 2 with root mean square (RMS) 7 values of 2 mm. However, the Canary Islands have 3 experienced an increase in seismic activity since 2003, which is considered a precursor of 4 the recent volcanic unrest and eruption on El Hierro island (López et al., 2017) and an 5 overall uplift until 2011 and subsidence from then on has also been detected (López et al., 6 2017). Finally, in Cape Verde, overall subsidence has been detected with data from 7 Sentinel-1 between 2017 and 2018 (Dias et al., 2018). 8 4. GNSS DATA 9 In order to study the kinematics of Macaronesia, a total of 19 continuous GNSS 10 stations were selected, managed by public institutions and with public data access; six 11 GNSS stations were selected in the Azores (AZGR, PIED, TERC, PDEL, FRNS and 12 FLRS), one in Madeira (FUNC), one in Cape Verde (TGCV), three in the Canary Islands 13 (IZAN, LPAL and MAS1), four on the Iberian Peninsula (VILL, SFER, LAGO and 14 CASC) and four on the African western coastline (TENT, CEU1, RABT and DAKR) 15 (Table 1 and Fig. 1). The observation window span was nearly 15 years, from May 2000 16 to March 2015. The most representative continuous GNSS stations were selected based on 17 their long and stable time series. 18 5. METHODOLOGY AND RESULTS 19 5.1. GNSS data processing 20 The GNSS data were processed using Bernese software v.5.0, particularly the Bernese 21 Processing Engine module (Dach et al., 2011), with satellite antenna bias corrections and 22 REPRO2 files (daily orbits, satellite clocks and Earth rotation parameters) (Rebischung et 23 al., 2016). A sampling rate of 30 s and a 10° elevation mask were applied. Hourly 24 tropospheric refraction corrections to the combined Saastamoinen (2013) and Niell (2004) 25 models were computed. The average distance between stations was 1562 km, the shortest 26 distance being 483 km and the longest 3105 km, so the Quasi Ionosphere-Free algorithm, 27 recommended for long baselines, was applied to resolve ambiguity (Mervart, 1995). Also, 28 the Ocean Tide Load bias was corrected by the GOT00.2 model (Ray, 1999; Scherneck 29 and Bos, 2002). 30 The VILL station, located on the Iberian Peninsula on the Eurasia Plate, was selected 31 as the reference in the differential positioning ionosphere-free mode, due to its very stable 32 time series in this time span (http://www.igs.org/network). The known reference station 33 position and velocity and precise satellite ephemeris data allow the daily coordinates to be 34 computed for each station in the ITRF2008 reference frame. From daily topocentric 35 positions (east, north and up), time series velocities were computed. The model fitted to 36 these time series was: 37         2 00 0 0 1 sin cos xiiii i xt x v tt a tt b tt t        , (1) 38 I. Barbero et al. viii Stud. Geophys. Geod., 65 (2021) where  xt is the value of the topocentric coordinate at time t, t 0 is the reference epoch, 1 x 0 is the initial position, v x is the velocity,  i is the angular frequency for harmonic 2 components, and a i and b i are the amplitudes of the sine and cosine, respectively. Finally, 3  t  represents the noise. 4 HECTOR software was applied to estimate velocities, as well as other parameters and 5 the noise model (Bos et al., 2013). This software uses maximum likelihood estimation 6 (MLE) to estimate these parameters and related uncertainties. White noise and power-law 7 noise close to flicker noise with a spectral index of 1 were considered as the optimum 8 noise models to describe the stochastic part of the time series (He et al., 2017; Klos et al., 9 2018, 2019; Williams et al., 2004). 10 The horizontal velocities contain the tectonic plate motion normally defined by 11 rotation around a Euler pole. NNR-MORVEL56 (Argus et al., 2011) defines the African-12 Nubian Plate Euler pole at 47.68°N and 68.44°W with an angular velocity of 13 0.292° Ma 1 . The horizontal velocities calculated by this model were subtracted from the 14 ITRF2008 horizontal velocities of each GNSS station using EPC software (Goudarzi et 15 al., 2014). Therefore, the eastward and westward 8 horizontal residual velocities of the 16 African-Nubian Plate motion (dV E_Eul and dV N_Eul , respectively) and module (dV Eul ) 9 for 17 each GNSS station were computed (see Fig. 2 and Table 2). 18 Analysing the horizontal residual velocities, without taking into account station FLRS 19 located on the North American Plate, some aspects can be highlighted: 20  All GNSS stations in the Azores archipelago have similar magnitudes and 21 azimuths with directions between E and NE. 22  There is continuous azimuth rotation of stations on the Iberian Peninsula and 23 Gibraltar, depicted by the horizontal residual velocities in CASC, LAGO, SFER, 24 CEU1 and TETN. These stations show similar magnitudes to those from Azores 25 stations, although those not on the African-Nubian Plate present, as expected, 26 higher-velocity residuals. 27  All stations on the Canary Islands have small horizontal velocity residuals with E 28 to NE azimuths, as does FUNC in the Madeira archipelago. 29  DAKR and RABT, on the African western coastline, have SE azimuths and higher 30 magnitudes than the Canary Islands. 31  TGCV in Cape Verde presents no horizontal velocity residual above the GNSS 32 accuracy, hence agreeing with NNR-MORVEL56 model velocity for the African-33 Nubian Plate there. 34  Finally, overall subsidence was detected in all Macaronesian stations, except for 35 DAKR and TGCV, with higher values north of São Jorge Island in the Azores 36 (TERC nearly 9 mm yr 1 and AZGR with 5 mm yr 1 ). 37 Recent Macaronesian kinematics from GNSS … Stud. Geophys. Geod., 65 (2021) ix 5.2. Cluster analysis 1 Statistical cluster analysis (Kaufman and Rousseeuw, 1990) was performed to evaluate 2 similarities in the horizontal residual velocities and thus to group stations based on their 3 kinematics. This method has been previously applied to GNSS data in California to 4 classify stations within the San Andres fault system (Savage and Simpson, 2013; Simpson 5 et al., 2012) and is widespread in GNSS studies (Özdemir and Karslıoğlu, 2019). 6 The k-means clustering method (Larose and Larose, 2014) was applied to the 7 horizontal residual velocity space. This approach groups the data set into a defined 8 number of clusters, k. The algorithm consists of randomly assigned k records (residual 9 velocities) as the initial cluster centre locations. For each record, the nearest cluster centre 10 using the Euclidean distance was found. A subset of the records was assigned to each 11 centroid, representing partition of the data set or cluster. For each of the k clusters, the 12 mean value of the records was found, named the centroid, and the location of each cluster 13 centre updated to the new value of the centroid. This process was repeated until the 14 centroids no longer changed or there was no significant shrinkage in the mean squared 15 Table 2. Geographical coordinates, first and final date of data 10 , easting (V E ), northing (V N ) and vertical (V U ) velocities, all with respect to the ITRF2008 reference frame with uncertainties and the corresponding residuals after correction from the NNR-MORVEL56 Nubian Euler pole (dV E_Eul , dV N_Eul and module dV Eul ) for each GNSS station. Except for coordinates, all the values in mm yr 1 . 11 GNSS Lat. N [°] Lon. E [°] V E V N V U  E  N  U 12 dV N_Eul dV E_Eul dV Eul AZGR 39.08786  28.02295 14.52 17.87  5.20 1.21 0.74 1.85 3.68 6.42 7.40 CASC 38.69342 9.41852 18.25 18.14  4.74 0.37 0.38 0.39  0.56 6.55 6.57 CEU1 35.89197 5.30639 15.96 17.41  5.32 0.29 0.53 0.81  2.09 2.27 3.09 DAKR 14.72090  17.43950 21.15 14.68 3.87 2.56 0.76 4.07  2.32 1.45 2.74 FLRS 39.45383  31.12639 10.59 21.43  2.84 0.64 0.45 1.32 8.13 −18.10 19.84 FRNS 37.76933  25.30821 15.11 16.02  5.12 0.63 0.49 2.15 1.02 5.92 6.01 FUNC 32.64795  16.90762 13.93 18.23  4.41 0.43 0.46 1.18 1.13 1.04 1.54 IZAN 28.30806  16.49968 18.16 17.06  2.40 0.30 0.31 0.39 1.12 1.36 1.76 LAGO 37.09894 8.66838 16.58 17.06  1.05 0.32 0.40 1.36  1.83 5.67 5.96 LPAL 28.76387  17.89383 16.94 17.21  1.25 0.26 0.34 1.32 0.16 2.19 2.20 MAS1 27.76374  15.63328 12.87 15.95  1.68 0.33 0.27 0.81  0.18 1.85 1.86 PDEL 37.74775  25.66277 11.89 14.77  2.59 1.13 0.67 1.36 1.06 3.68 3.83 PIED 38.41382  28.03246 16.73 17.85  3.54 0.19 0.26 0.46 0.57 3.40 3.45 RABT 33.99810 6.85429 15.93 16.94  1.87 0.33 0.37 0.52  1.35 2.64 2.97 SFER 36.46435 6.20564 12.21 17.50  0.24 0.41 0.39 1.04  2.45 2.63 3.59 TERC 38.71899  27.15299 15.53 16.31  8.81 0.36 0.39 0.80 3.11 3.72 4.85 TETN 35.56160 5.36300 18.82 16.00  1.18 1.10 0.53 3.00  3.18 1.73 3.62 TGCV 16.75477  22.98276 19.23 16.32 0.47 0.21 0.23 0.05 0.40 0.23 0.46 VILL 40.44359 3.95198 14.52 17.87  1.75 1.21 0.74 1.85  3.38 7.04 7.81 Average (without FLRS) 16.22 16.87 I. Barbero et al. xvi Stud. Geophys. Geod., 65 (2021) velocity residuals in regions with low to non-existent seismic activity, specifically: greater 1 than 1 mm yr 1 in Madeira, between 1 and 2 mm yr 1 in the Canary Islands and between 2 2 and 3 mm yr 1 on the continental margin of Africa (DAKR and RABT stations). In 3 addition, stations in the Azores and Gibraltar arc regions present horizontal velocity 4 residuals greater than 3 mm yr 1 , influenced mainly by their proximity to plate boundaries 5 and with directions according to previous studies. 6 The similar kinematic behaviour between islands is not followed by stations on the 7 continental margin of Africa. The difference in magnitude, but particularly in azimuth, 8 causes the inferred strain rate extension NW-SE in the Canary Islands to shift to W-E to 9 the east of Madeira and NE-SW at the Gloria Fault. From seismicity, kinematic and strain 10 data analysis, a shear zone was located in harmony with other studies (Mantovani et al., 11 2007). Seismic activity and strain rates suggest some dextral strike-slip deformation 12 occuring within the Morocco subplate along a NNW-SSE direction from the Gorringe 13 Bank to Agadir, and/or a remnant mantel plume upwelling (Saki et al., 2015). 14 In the Canary archipelago, horizontal residual velocities show a slight azimuth shift 15 that affects the strain rate between Tenerife and Gran Canaria. The strain rate in the 16 Canary Islands, inferred as positive (extension) in the NW-SE direction, is perpendicular 17 to the NE-SW seismically active oblique reverse fault identified in previous studies 18 (Barbero et al., 2018). It could be also due to a temporary magma accumulation zone that 19 contributes to deep endogenous growth and uplift of the volcanoes. 20 All GNSS stations in the Macaronesian islands show subsidence, possibly linked to 21 isostatic adjustment of volcanic islands (Moore, 1970). Finally, although this study 22 presents large areas without GNSS stations, limiting conclusions there, in regions covered 23 by several GNSS stations our GNSS data analysis presents consistency with and support 24 to previous local studies, providing an overall interpretation of the recent intraplate 25 kinematics of the African-Nubian Plate, particularly between emerged land. 26 27 28 References 21 29 Argus D.F., Gordon R.G. and Demets C., 2011. Geologically current motion of 56 plates relative to 30 the no-net-rotation reference frame. Geochem. Geophys. Geosyst., 12, Art.No. Q11001, 31 DOI: 10.1029/2011GC003751 32 Barbero I., Torrecillas C., Prates G., Páez R., Gárate J., García A. and Berrocoso M., 2018. 33 Assessment of ground deformation following Tenerife’s 2004 volcanic unrest (Canary 34 Islands). J. Geodyn., 121, 18, DOI: 10.1016/j.jog.2018.06.002 35 Blewitt G. and Lavallée D., 2002. Effect of annual signals on geodetic velocity. J. Geophys. Res.- 36 Solid Earth, 107, Art.No. 2145, DOI: 10.1029/2001JB000570 37 Borges J.F.F., Bezzeghoud M., Buforn E., Pro C. and Fitas A., 2007. The 1980, 1997 and 1998 38 Azores earthquakes and some seismo-tectonic implications. Tectonophysics, 435, 3754, 39 DOI: 10.1016/j.tecto.2007.01.008 40 Bos M.S., Bastos L. and Fernandes R.M.S., 2010. The influence of seasonal signals on the 41 estimation of the tectonic motion in short continuous GPS time-series. J. Geodyn., 49, 42 205209, DOI: 10.1016/j.jog.2009.10.005 43 Recent Macaronesian kinematics from GNSS … Stud. Geophys. Geod., 65 (2021) xvii Bos M.S., Fernandes R.M.S., Williams S.D.P. and Bastos L., 2013. Fast error analysis of continuous 1 GNSS observations with missing data. J. Geodesy, 87, 351360, DOI: 10.1007/s00190-0122 0605-0 3 Burke K. and Wilson J.T., 1972. Is the African Plate stationary? Nature, 239, 387390, 4 DOI: 10.1038/239387b0 5 Calais E., DeMets C. and Nocquet J.M., 2003. Evidence for a post-3.16-Ma change in Nubian6 Eurasia-North America plate motions? Earth Planet. Sci. Lett., 216, 8192, DOI: 10.1016 7 /S0012-821X(03)00482-5 8 Cardozo N. and Allmendinger R.W., 2009. SSPX: A program to compute strain from 9 displacement/velocity data. Comput. Geosci., 35, 13431357, DOI: 10.1016 10 /j.cageo.2008.05.008 11 Catalán M., Martos Y.M. and Martín-Davila J., 2019. Eurasia-Africa plate boundary affected by 12 a South Atlantic asthenospheric channel in the Gulf of Cadiz region? Pure Appl. Geophys., 13 177, 17251738, DOI: 10.1007/s00024-019-02380-4 14 Dach R., Hugentobler U. and Walser P., 2011. Tutorial Processing Example Introductory Course 15 Terminal Session. http://www.bernese.unibe.ch/docs50/TERMINAL.pdf 16 de Lis Mancilla F., Stich D., Berrocoso M., Martín R., Morales J., Fernandez-Ros A., Páez R., and 17 Perez-Pena A., 2013. Delamination in the Betic Range: Deep structure, seismicity, and GPS 18 motion. Geology, 41, 307310, DOI: 10.1130/G33733.1 19 DeMets C., Gordon R.G. and Argus D.F., 2010. Geologically current plate motions. Geophys. J. 20 Int., 181, 180, DOI: 10.1111/j.1365-246X.2009.04491.x 21 DeMets C., Gordon R.G., Argus D.F. and Stein S., 1994. Effect of recent revisions to the 22 geomagnetic resersal time-scale on estimates of current plate motions. Geophys. Res. Lett., 21, 23 21912194, DOI: 10.1029/94GL02118 24 Dias P., Catalao J. and Marques F.O., 2018. Sentinel-1 InSAR data applied to surface deformation 25 in Macaronesia (Canaries and Cape Verde). Procedia Computer Science, 138, 382387, 26 DOI: 10.1016/j.procs.2018.10.054 27 Domínguez-Cerdeña I., del Fresno C. and Rivera L., 2011. New insight on the increasing seismicity 28 during Tenerife’s 2004 volcanic reactivation. J. Volcanol. Geotherm. Res., 206, 1529, 29 DOI: 10.1016/j.jvolgeores.2011.06.005 30 Duarte J.C., Rosas F.M., Terrinha P., Schellart W.P., Boutelier D., Gutscher M.A. and Ribeiro A., 31 2013. Are subduction zones invading the atlantic? Evidence from the southwest iberia margin. 32 Geology, 41, 839842, DOI: 10.1130/G34100.1 33 Fernandes R.M.S., Bastos L., Miranda J.M., Lourenço N., Ambrosius B.A.C., Noomen R. and 34 Simons W., 2006. Defining the plate boundaries in the Azores region. J. Volcanol. Geotherm. 35 Res., 156, 19, DOI: 10.1016/j.jvolgeores.2006.03.019 36 Garcia A., Fernandez-Ros A., Berrocoso M., Marrero J.M., Prates G., De la Cruz-Reyna S. and 37 Ortiz R., 2014. Magma displacements under insular volcanic fields, applications to eruption 38 forecasting: El Hierro, Canary Islands, 20112013. Geophys. J. Int., 197, 322334, 39 DOI: 10.1093/gji/ggt505 40 Geldmacher J., Hoernle K., Klügel A., von den Bogaard P., Wombacher F. and Berning B., 2006. 41 Origin and geochemical evolution of the Madeira-Tore Rise (eastern North Atlantic). 42 J. Geophys. Res.-Solid Earth, 111, Art.No. B09206, DOI: 10.1029/2005JB003931 43 I. Barbero et al. xviii Stud. Geophys. Geod., 65 (2021) Geldmacher J., Hoernle K., Klügel A., von den Bogaard P. and Duggen S., 2006. A geochemical 1 transect across a heterogeneous mantle upwelling: Implications for the evolution of the 2 Madeira hotspot in space and time. Lithos, 90, 131144, DOI: 10.1016/j.lithos.2006.02.004 3 Geyer A., Martí J. and Villaseñor A., 2016. First-order estimate of the Canary Islands plate-scale 4 stress field: Implications for volcanic hazard assessment. Tectonophysics, 679, 125139, 5 DOI: 10.1016/j.tecto.2016.04.010 6 Goudarzi M.A., Cocard M. and Santerre R., 2014. EPC: Matlab software to estimate Euler pole 7 parameters. GPS Solut., 18, 153162, DOI: 10.1007/s10291-013-0354-4 8 He X., Montillet J.P., Hua X., Yu K., Jiang W. and Zhou F., 2017. Noise analysis for environmental 9 loading effect on GPS position time series. Acta Geodyn. Geomater., 14, 131142, 10 DOI: 10.13168/AGG.2016.0034 11 Heidbach O., Rajabi M., Cui X., Fuchs K., Müller B., Reinecker J., Reiter K., Tingay M., Wenzel 12 F., Xie F.R., Ziegler M.O. and Zoback M.L., 2018. The World Stress Map database release 13 2016: Crustal stress pattern across scales. Tectonophysics, 744, 484498, DOI: 10.1016 14 /j.tecto.2018.07.007 15 Hipólito A., Hipólito A., Madeira J., Carmo R. and Gaspar J.L., 2014. Neotectonics of Graciosa 16 island (Azores): a contribution to seismic hazard assessment of a volcanic area in a complex 17 geodynamic setting. Ann. Geophys., 56, Art.No. S0677, DOI: 10.4401/ag-6222 18 International Seismological Centre, 2019. On-line Bulletin, DOI: 10.31905/D808B830 19 (http://www.isc.ac.uk/iscbulletin/) 20 Jiménez-Munt I. and Negredo A.M., 2003. Neotectonic modelling of the western part of the Africa21 Eurasia plate boundary: from the Mid-Atlantic ridge to Algeria. Earth Planet. Sci. Lett., 205, 22 257271, DOI: 10.1016/S0012-821X(02)01045-2 23 Kaufman L. and Rousseeuw P.J., 1990. Finding Groups in Data. Johm Wiley and Sons, Hoboken, 24 NJ, DOI: 10.1002/9780470316801 25 Klos A., Bos M.S. and Bogusz J., 2018. Detecting time-varying seasonal signal in GPS position 26 time series with different noise levels. GPS Solut., 22, Art.No. UNSP21, 27 DOI: 10.1007/s10291-017-0686-6 28 Klos A., Bos M.S., Fernandes R.M.S. and Bogusz J., 2019. Noise-dependent adaption of the Wiener 29 filter for the GPS position time series. Math. Geosci., 51, 5373, DOI: 10.1007/s11004-01830 9760-z 31 Klügel A., Longpré M.A., García-Cañada L. and Stix J., 2015. Deep intrusions, lateral magma 32 transport and related uplift at ocean island volcanoes. Earth Planet. Sci. Lett., 431, 140149, 33 DOI: 10.1016/j.epsl.2015.09.031 34 Kogan M.G., Steblov G.M., King R.W., Herring T.A., Frolov D.I., Egorov S.G., Levin V.Y., 35 Lerner-Lam A. and Jones A., 2000. Geodetic constraints on the rigidity and relative motion of 36 Eurasia and North America. Geophys. Res. Lett., 27, 20412044, DOI: 10.1029 37 /2000GL011422 38 Koulali A., Ouazar D., Tahayt A., King R.W., Vernant P., Reilinger R.E., McClusky S., 39 Mourabit T., Davila J.M. and Amraoui N., 2011. New GPS constraints on active deformation 40 along the Africa-Iberia plate boundary. Earth Planet. Sci. Lett., 308, 211217, DOI: 10.1016 41 /j.epsl.2011.05.048 42 Larose D.T. and Larose C.D., 2014. Hierarchical and k-means clustering. Discovering Knowledge in 43 Data, 209227, John Wiley and Sons, Hoboken, NJ, DOI: 10.1002/9781118874059.ch10 44 Recent Macaronesian kinematics from GNSS … Stud. Geophys. Geod., 65 (2021) xix López C., García-Cañada L., Martí J. and Domínguez Cerdeña I., 2017. Early signs of geodynamic 1 activity before the 20112012 El Hierro eruption. J. Geodyn., 104, 114, DOI: 10.1016 2 /j.jog.2016.12.005 3 Madeira J., Brum da Silveira A., Hipólito A. and Carmo R., 2015. Chapter 3 Active tectonics in the 4 central and eastern Azores islands along the Eurasia-Nubian boundary: a review. Geol. Soc. 5 London Mem., 44, 1532, DOI: 10.1144/m44.3 6 Malservisi R., Hugentobler U., Wonnacott R. and Hackl M., 2013. How rigid is a rigid plate? 7 Geodetic constraint from the TrigNet CGPS network, South Africa. Geophys. J. Int., 192, 8 918928, DOI: 10.1093/gji/ggs081 9 Mantovani E., Viti M., Babbucci D. and Albarello D., 2007. Nubian-Eurasia kinematics: An 10 alternative interpretation from Mediterranean and North Atlantic evidence. Ann. Geophys., 50, 11 341366, DOI: 10.4401/ag-3073 12 Marques F.O., Catalão J., Hildenbrand A. and Madureira P., 2015. Ground motion and tectonics in 13 the Terceira Island: Tectonomagmatic interactions in an oceanic rift (Terceira Rift, Azores 14 Triple Junction). Tectonophysics, 651, 1934, DOI: 10.1016/j.tecto.2015.02.026 15 Marques F.O., Catalão J.C., DeMets C., Costa A.C.G. and Hildenbrand A., 2013. GPS and tectonic 16 evidence for a diffuse plate boundary at the Azores Triple Junction. Earth Planet. Sci. Lett., 17 381, 177187, Elsevier B.V, DOI: 10.1016/j.epsl.2013.08.051 18 Martín A., Sevilla M. and Zurutuza J., 2014. Crustal deformation study in the Canary Archipelago 19 by the analysis of GPS observations. J. Appl. Geodesy, 8, 129140, DOI: 10.1515/jag-201420 0002 21 McClusky S., Reilinger R., Mahmoud S., Ben Sari D. and Tealeb A., 2003. GPS constraints on 22 Africa (Nubian) and Arabia plate motions. Geophys. J. Int., 155, 126138, DOI: 10.1046 23 /j.1365-246X.2003.02023.x 24 Mendes V.B., Madeira J., Brum da Silveira A., Trota A., Elosegui P. and Pagarete J., 2013. Present25 day deformation in São Jorge Island, Azores, from episodic GPS measurements (20012011). 26 Adv. Space Res., 51, 15811592, DOI: 10.1016/j.asr.2012.10.019 27 Mervart L., 1995. Ambiguity Resolution Techniques in Geodetic and Geodynamic Applications of 28 the Global Positioning System. PhD Thesis. University of Bern, Bern, Switzerland. 29 Métrich N., Zanon V., Créon L., Hildenbrand A., Moreira M. and Marques F.O., 2014. Is the 30 ‘Azores Hotspot’ a wetspot? Insights from the geochemistry of fluid and melt inclusions in 31 olivine of Pico basalts. J. Petrol., 55, 377393, DOI: 10.1093/petrology/egt071 32 Miller M.S., O’Driscoll L.J., Butcher A.J. and Thomas C., 2015. Imaging Canary Island hotspot 33 material beneath the lithosphere of Morocco and southern Spain. Earth Planet. Sci. Lett., 431, 34 186194, DOI: 10.1016/j.epsl.2015.09.026 35 Miranda J.M., Navarro A., Catalão J. and Fernandes R.M.S., 2012. Surface displacement field at 36 Terceira island deduced from repeated GPS measurements. J. Volcanol. Geotherm. Res., 37 217218, 17, DOI: 10.1016/j.jvolgeores.2011.10.009 38 Moore J.G., 1970. Relationship between subsidence and volcanic load, Hawaii. Bull. Volcanol., 34, 39 562576, DOI: 10.1007/BF02596771 40 Negredo A.M., Bird P., Sanz de Galdeano C. and Buforn E., 2002. Neotectonic modeling of the 41 Ibero-Maghrebian region. J. Geophys. Res.-Solid Earth, 107, Art.No. 2292, 42 DOI: 10.1029/2001JB000743 43 I. Barbero et al. xx Stud. Geophys. Geod., 65 (2021) Niell A.E., 2004. Global mapping functions for the atmosphere delay at radio wavelengths. 1 J. Geophys. Res.-Solid Earth, 101, 32273246, DOI: 10.1029/95jb03048 2 Özdemir S. and Karslıoğlu M.O., 2019. Soft clustering of GPS velocities from a homogeneous 3 permanent network in Turkey. J. Geodesy, 93, 11711195, DOI: 10.1007/s00190-019 4 -01235-z 5 Pacheco J.M., Ferreira T., Queiroz G., Wallenstein N., Coutinho R., Cruz J.V., Pimentel A., 6 Silva R., Gaspar J.L. and Goulart C., 2013. Notas sobre a geologia do arquipélago dos Açores. 7 In: Dias R., Araújo A., Terrinha P. and Kullberg J.C. (Eds), Geologia de Portugal. Vol. 2, 8 595690, Escolar Editora, Lisbon, Portugal (in Portuguese). 9 Pérez-Peña A., Martín-Davila J., Gárate J., Berrocoso M. and Buforn E., 2010. Velocity field and 10 tectonic strain in Southern Spain and surrounding areas derived from GPS episodic 11 measurements. J. Geodyn., 49, 232240, DOI: 10.1016/j.jog.2010.01.015 12 Prates G., García A., Fernández-Ros A., Marrero J.M., Ortiz R. and Berrocoso M., 2013. 13 Enhancement of sub-daily positioning solutions for surface deformation surveillance at El 14 Hierro volcano (Canary Islands, Spain). Bull. Volcanol., 75, Art.No. 724, DOI: 10.1007 15 /s00445-013-0724-3 16 Ramalho R.A.S., 2011. The Cape Verde Archipelago. Building the Cape Verde Islands, Springer17 Verlag, Berlin, Germany, 1326, DOI: 10.1007/978-3-642-19103-9_2 18 Ray R.D., 1999. A Global Ocean Tide Model From TOPEX/POSEIDON Altimetry: GOT99.2. 19 Technical Memorandum 1999-209478. NASA Goddard Space Flight Center, Greenbelt, MD 20 Rebischung P., Altamimi Z., Ray J. and Garayt B., 2016. The IGS contribution to ITRF2014. 21 J. Geodesy, 90, 611630, DOI: 10.1007/s00190-016-0897-6 22 Rousseeuw P.J., 1987. Silhouettes: A graphical aid to the interpretation and validation of cluster 23 analysis. J. Comput. Appl. Math., 20, 5365, DOI: 10.1016/0377-0427(87)90125-7 24 Ruiz C.R., García-Cacho L., Araña V., Luque A.Y. and Felpeto A., 2000. Submarine volcanism 25 surrounding Tenerife, Canary Islands: Implications for tectonic controls, and oceanic shield 26 formingprocesses. J. Volcanol. Geotherm. Res., 103, 105119, DOI: 10.1016/S0377-0273 27 (00)00218-3 28 Saastamoinen J., 2013. Atmospheric correction for the troposphere and stratosphere in radio ranging 29 satellites. In: Henriksen S.W., Mancini A. and Chovitz B.H. (Eds), The Use of Artificial 30 Satellites for Geodesy, Volume 15. American Geophysical Union, Washington, D.C., 31 247251, DOI: 10.1029/GM015p0247 32 Saki M., Thomas C., Nippress S.E.J. and Lessing S., 2015. Topography of upper mantle seismic 33 discontinuities beneath the North Atlantic: The Azores, Canary and Cape Verde plumes. Earth 34 Planet. Sci. Lett., 409, 193202, DOI: 10.1016/j.epsl.2014.10.052 35 Savage J.C. and Simpson R.W., 2013. Clustering of GPS velocities in the Mojave Block, 36 southeastern California. J. Geophys. Res.-Solid Earth, 118, 17471759, DOI: 10.1029 37 /2012JB009699 38 Scheidegger A.E., 2002. Morphometric analysis and its relation to tectonics in Macaronesia. 39 Geomorphology, 46, 95115, DOI: 10.1016/S0169-555X(02)00056-9 40 Scherneck H.-G. and Bos M.S., 2002. Ocean tide and atmospheric loading. In: Vandenberg N.R. 41 and Baver K.D. (Eds), IVS 2002 General Meeting Proceedings. NASA/CP–2012–217504, 42 NASA Goddard Space Flight Center, Greenbelt, MD, 205214. 43 http://publications.lib.chalmers.se/records/fulltext/local_158760.pdf 44 Recent Macaronesian kinematics from GNSS … Stud. Geophys. Geod., 65 (2021) xxi Serpelloni E., Vannucci G., Pondrelli S., Argnani A., Casula G., Anzidei M., Baldi P. and Gasperini 1 P., 2007. Kinematics of the Western Africa-Eurasia plate boundary from focal mechanisms 2 and GPS data. Geophys. J. Int., 169, 11801200, DOI: 10.1111/j.1365-246X.2007.03367.x 3 Silva P.F., Henry B., Marques F.O., Hildenbrand A., Lopes A., Madureira P., Madeira J., Nunes J.C. 4 and Roxerova Z., 2018. Volcano-tectonic evolution of a linear volcanic ridge (Pico-Faial 5 Ridge, Azores Triple Junction) assessed by paleomagnetic studies. J. Volcanol. Geotherm. 6 Res., 352, 7891, Elsevier B.V, DOI: 10.1016/j.jvolgeores.2018.01.005 7 Silveira G., Vinnik L., Stutzmann E., Farra V., Kiselev S. and Morais I., 2010. Stratification of the 8 Earth beneath the Azores from P and S receiver functions. Earth Planet. Sci. Lett., 299, 9 91103, DOI: 10.1016/j.epsl.2010.08.021 10 Simpson R.W., Thatcher W. and Savage J.C., 2012. Using cluster analysis to organize and explore 11 regional GPS velocities. Geophys. Res. Lett., 39, Art.No. L18307, DOI: 10.1029 12 /2012GL052755 13 Teza G., Pesci A. and Casula G., 2012. Strain rate computation in Northern Victoria Land 14 (Antarctica) from episodic GPS surveys. Geophys. J. Int., 189, 851862, DOI: 10.1111 15 /j.1365-246X.2012.05403.x 16 Teza G., Pesci A. and Galgaro A., 2008. Grid_strain and grid_strain3: Software packages for strain 17 field computation in 2D and 3D environments. Comput. Geosci., 34, 11421153, 18 DOI: 10.1016/j.cageo.2007.07.006 19 von den Bogaard P., 2013. The origin of the Canary Island Seamount Province - New ages of old 20 seamounts. Sci. Rep., 3, Art.No. 2107, DOI: 10.1038/srep02107 21 Vernant P., Fadil A., Mourabit T., Ouazar D., Koulali A., Davila J.M., Garate J., McClusky S. and 22 Reilinger R., 2010. Geodetic constraints on active tectonics of the Western Mediterranean: 23 Implications for the kinematics and dynamics of the Nubian-Eurasia plate boundary zone. 24 J. Geodyn., 49, 123129, DOI: 10.1016/j.jog.2009.10.007 25 Weiss B.J., Hübscher C. and Lüdmann T., 2015. The tectonic evolution of the southeastern Terceira 26 Rift/São Miguel region (Azores). Tectonophysics, 654, 7595, DOI: 10.1016 27 /j.tecto.2015.04.018 28 Williams S.D.P., Bock Y., Fang P., Jamason P., Nikolaidis R.M., Prawirodirdjo L., Miller M. and 29 Johnson D.J., 2004. Error analysis of continuous GPS position time series. J. Geophys. Res.- 30 Solid Earth, 109, Art.No. B03412, DOI: 10.1029/2003jb002741 31 Zitellini N., Gràcia E., Matias L., Terrinha P., Abreu M.A., DeAlteriis G., Henriet J.P., 32 Danobeitia J.J., Masson D.G., Mulder T., Ramella R., Somoza L. and Diez S., 2009. The quest 33 for the Africa-Eurasia plate boundary west of the Strait of Gibraltar. Earth Planet. Sci. Lett., 34 280, 1350, DOI: 10.1016/j.epsl.2008.12.005 35 36 37 38 I. Barbero et al. xxii Stud. Geophys. Geod., 65 (2021) 1 1 Please, specify also the city. 2 Please, check and approve the edit. 3 Please, check and approve the edit. 4 Please, check and approve the edit and define GB. 5 Please, check and approve the edit. 6 Do you mean country, or subplate? 7 Please, check and approve the edit. 8 Please, check and approve the edit. 9 Please, approve formatting of the parameters. 10 This seems to be missing in the table. 11 Please, check and approve the edit. 12 Please, define these parameters (standard deviations?) 13 Please, check and approve the edit. 14 Please, check and approve the edit. 15 In the bw version of this figure, please, modify the part a). The bw scale has the same greys for different values. Change it from light do dark. 16 These are not visible. 17 Please, check and approve the edit. 18 The colored scale in a) is unclear, the same values are used for different colors. Please, clarify. 19 The same as in the bw case of Fig. 4a. Moreover, in bw Fig. 5b the scale does not correspond to the plot, there are darker greys in the plot. 20 Please, check and approve the edit. 21 References were formatted according to our style. Please, check them all carefully. Make also sure that all the references are cited in the text (and vice versa, all the references cited in the text are listed in the list of references).