During the 2021 La Palma strombolian and fissure eruption, two faults were identified that controlled the spatial distribution of earthquake hypocenters and effusive eruptive vents. One of these faults has a NW-SE trend (Tazacorte Fault: TZF) and the other one shows an ENE-WSW trend (Mazo Fault: MZF). Previous works on fault structural analysis in La Palma indicated that the eruption zone was compatible with an extensional tectonic strain ellipsoid which activated normal-strike-slip directional faults at the confluence of TZF and MZF. These fractures were activated during the 2021 Tajogaite eruption, determining the NW-SE and WSW-ENE spatial distribution of vents. Both faults were mapped in real time during the volcanic eruption from fieldwork and remote sensing imagery (aerial drone images). We have collected more than 300 fracture data associated with the effusive vents and post-eruption seismic creep. Since the affected area was densely inhabited, most of these fractures affect houses and infrastructures. Some of the houses affected by the TZF were damaged 9 months after the eruption, although they were not damaged during the eruption. Surprisingly, these houses already had repairs made to the same fractures since 1980, giving information of previous fault creep movement. During the 2021 Tajogaite eruption, shallow seismicity was spatially related to both faults, suggesting a seismic behavior instead of the precedent creep movement. However, the lack of seismicity after the eruption indicates that the faults went back to creep aseismic behavior, similarly to 1980. The mapping and monitoring of these faults (TZF and MZF) is relevant bearing in mind that they have been active since 1980 and the post-eruptive phase of the 2021 volcanic eruption, which has to be included in the land use planning in areas affected by the volcanic eruption and creep movement. Furthermore, both faults could act as seismogenic sources triggering volcanic earthquakes with potential high macroseismic intensities and mass movements. The data presented here show the importance of having this type of study before the onset of the eruption, thus allowing a better interpretation of seismic data during volcanic unrest.
Keywords: Tajogaite eruption 2021; active fault; strain tensor; earthquake; creep; La Palma
The 2021 Tajogaite eruption was a subaerial fissure eruption with both effusive and explosive styles. The activity was mainly strombolian with episodic phreatomagmatic phases. The eruption opened along a NW-SE-oriented fissure, which was conditioned by previous fracturing. This fracturing has been acting aseismically (creep), affecting houses and infrastructure. Examples of fracturing in volcanic environments with seismic and aseismic movement appear in the works of [[
The previous structural geology work based on the population analysis of faults [[
La Palma Island (Figure 1), located at 28°39′59″ N–17°51′59″ W, is one of the main eight emerged volcanic islands that comprises the Canary archipelago. The Canary Islands are an intraplate archipelago that was formed from the Miocene to the Holocene in a passive margin [[
The island has a complex geology formed essentially by three volcanic units bound by unconformities: the basal complex, the older volcanic series, and the recent volcanic series [[
The Tajogaite 2021 volcanic eruption had a violent strombolian style (Volcanic Explosivity Index 3 (VEI)) with high effusive activity. During the first stages of lava emission, the La Palma volcano issued basanite–tephrite lavas, developing mostly a'a flows that eventually increased in basanite composition, increasing fluidity [[
The study of tectonic stress and strain deformation fields in active zones makes it possible to describe the role of faults within volcanotectonic processes, which are ultimately those that dominate the geodynamics of volcanic eruptions. In order to establish the orientation of the deformation fields generated by the tectonic stresses, it is necessary in such studies to establish the kinematics of the brittle deformation determined by both the seismic and aseismic movement of the faults. One of the most powerful techniques for determining the stress/strain field is the so-called "fault population analysis" (FPA) based on the geometry of the fracture and the kinematic markers observed on the plane of the microfracture, assuming that this process is self-affine [[
Generally speaking, the kinematics of the deformation are obtained from the failure plane and the measurement of its striation (Figure 2). This means that it is necessary in the FPA to take field measurements of pairs of fault plane orientation and striation angle data. In the methods used in this work and described below, the Landslide Model [[
- (a) The Right-Dihedron Method, RDM, is a geometric and dynamic method for obtaining deformation fields [[
28 ]]. - (b) The Slip Model, SLIPM, is a method for reconstructing paleostresses through stress inversion [[
25 ], [27 ]].
The main method applied in this work is the Slip Model (SLIPM), as it has several advantages over other paleostress reconstruction methods by means of inverse techniques. (a) It obtains a graphic representation that is applicable to the study of the geodynamics of fracturing by spatial distribution (deformation cluster) and (b) it is a direct method that avoids subsequent manipulations that could be mathematical artifacts. SLIPM, in addition, allows the separation of fault subpopulations that are compatible with different maximum horizontal shortening directions (D
K' = ey/ez (
In which ez is the vertical deformation axis and ey is the maximum horizontal shortening axis.
The Right-Dihedral (RDM) method is a semi-quantitative method based on the overlapping of compressional and extensional zones by means of a stereographic plot [[
The advantage of applying these methods is that they make it possible to discriminate regional deformation fields, resulting from large tectonic structures such as oceanic ridges or continental collision zones, for example, from local deformation fields resulting from specific volcanic phenomena or the emplacement of magmatic dikes. As for the study of deformation due to regional fields, [[
Once the population analysis of faults has been carried out and the strain tensor has been calculated, as defined by the K' shape factor and the D
The calculation and mapping of the strain trajectories have been carried out with the local tensor interpolation method developed by [[
A total of 344 faults with slickensides distributed along 39 stations on the island of La Palma have been measured in different field campaigns (Figure 3). The FPA has been used to obtain the D
The Atlantic deformation field on the island of La Palma is defined by both reverse fault populations (7 substations) and normal fault populations (16 substations) (Figure 3 and Figure 4A). The areas with extensive K' are located in the NE and NW areas as well as in the most modern area of the island, the CV ridge in the S. However, the local field is fundamentally extensive (21 substations) with a strike-slip component in the NE and only two compressional stations (in the NW and SW of the island) (Figure 3 and Figure 4B).
For monitoring seismicity during the volcanic eruption, hypocenter data from the Instituto Geográfico Nacional (IGN: https://
In September 2021, earthquakes of M > 3 at a depth of 9 km began to be recorded. Some of these hypocenters showed a NW-SE alignment at a depth of 9 km. This line was probably contained in a fault plane to which the rest of the hypocenters adapted. This plane could be parallel to the one inferred for the Jedey eruption (1585), since our surface microfracturing data measured at this volcano (Figure 5) show a fault plane with a 100°/80° SW orientation and dip that controlled this eruption and the emission of volcanic material [[
Ref. [[
- 12–19 September (Figure 6A): Seismicity prior to the eruption indicating the rise of magma in a dike towards the surface in an accelerated manner [[
37 ]]. The hypocentral distribution fits the Tazacorte Fault plane (TZF) with a NW-SE orientation (N100° E/80° SW). The seismicity was mainly concentrated at a depth of 9 km and began its rise to the surface on day 15 [[38 ]]. - 20 September–1 October (Figure 6B): The eruption began on 19 September 2021 at 14:12 h (UTC) in favor of the TZF (NW-SE) with a clear alignment of eruptive vents that rapidly developed a cinder cone. The TZF remained active throughout the eruption with the main vents aligned with this structure (Figure 7). The distribution of the hypocenters adjusted to a fault plane compatible with the TZF, at a depth of mainly between 15 and 5 km. During this entire interval, there was an opening of different eruptive vents aligned in accordance with the TZF [[
39 ]] and of surface fractures NW of the main vent. During this stage, the distribution of epicenters of stage 1 was maintained, differing only in that the eruption had already occurred. - 2 October–25 November (Figure 6C). This time interval marked the change in the orientation of the hypocenter distribution. It went from a NW-SE direction (N100° E) to a WSW-ENE direction (N080° E), adjusting to a 70°–80° SE dip plane that we have called Mazo Fault (MZF). These hypocenters were concentrated at a depth of between 8 and 16 km. There is another concentration of epicenters along this same direction, N080° E, and at a depth of between 25 and 35 km, but in this case with a dip towards the NW [[
35 ]]. This alignment was reflected in the eruption on 25 November, in which new distal eruptive vents began to open in the southern area of the flows [[39 ]]. On 4 December, the last distal emission center opened in this southern area. All these eruptive vents were aligned in accordance with the MZF. - 26 November–13 December (Figure 6D): This was the last stage of the eruption, with seismic activity decreasing significantly, but also with the maintenance of the two sets of epicenters defined in the previous stage (MZF). The eruption finished abruptly on 13 December 2021 at 22:21 UTC, its end being officially announced by the authorities on 25 December (https://www..gobiernodecanarias.org/noticias/wp-content/uploads/2021/12/251221-INFORME-Comit%C3%A9-Cient%C3%ADfico-PDF.pdf; accessed on 15 January 2024).
In summary, the TZF remained active throughout the eruption, generating vents parallel to this route in the main vent zone and N area of the flows (but without significant seismicity from phase 3 onwards). The MZF was activated from 1 October onwards, with the appearance of distal eruptive vents on 25 November, aligned according to the normal echelon faults contained in this fault. In this way, these two fault breaks were revealed, breaking the upper crust and spatially conditioning the distribution of eruptive vents.
Both the TZF and the MZF generated a multitude of newly formed surface fractures during the eruption, similar to those mapped by [[
The observed faults have lengths that range between tens and hundreds of meters (Figure 8, Figure 9 and Figure 10). The fractures that appear during the eruption have either an extensional or a transtensive character (Figure 8A). Transtensive fractures suggest a movement in the horizontal along the planes and are not related to lava flow emission, thus evidencing a dominant role of tectonics in their geometry and distribution. These faults affect anthropic buildings (Figure 8B), generating significant damage to many of them. During the eruptive process, many of these faults display high temperatures and gas anomalies.
In the TZF and the MZF, there was a development of normal echelon faults that was to be included within two deformation bands corresponding to each of the main faults (Figure 9). This distribution of normal echelon faults could indicate a left-lateral movement direction for TZF and dextral movement direction for MZF and explain the appearance of eruptive vents along these normal fault systems. Therefore, there would appear to have been a transtensive activity of these two faults at least in the eruption area (Figure 10).
A total of 362 measurements of surface faults (Figure 9) and 49 of striated faults (Figure 11) have been made. Figure 9 shows the spatial distribution of the fifteen measurement stations taken in the MZF with their corresponding stereograms and the two stations taken in the TZF in zone N of the lava flows. Following Anderson's fracturing model, the distribution of conjugate fracture systems appears to indicate a NW-SE shortening direction, which is compatible with the Atlantic strain field associated with the opening of the Atlantic Ridge. It has been possible, at six measuring stations, to collect data on fault striations that allow us to analyze these data to obtain the strain tensor. We have used the Right-Dihedral method, which offers solutions that are compatible with the NW-SE shortening direction related to the Atlantic deformation field (Figure 11).
Monitoring of anthropic construction elements after the eruptive process demonstrates the formation of new fractures. This became evident in March 2022, when houses that had not suffered damage from fractures during the eruption began to fracture in the N area of the lava flows. After the first studies, we also observed that some fractures had been repaired prior to the eruption, suggesting that they had moved before the eruptive process. The information given by the owners of the houses is that these fractures appeared after the 1980s, when these houses were built. These movements have been intermittent, without the owners of the houses being able to determine the intervals at which these fractures moved. These pre-eruption movements have been observed in both the TZF and the MZF (Figure 12).
We are currently monitoring these fractures with a Novatest FIPSG100 precision fissurometer, with a measurement resolution of 0.001 mm and with an error of ±0.003 mm. In order to eliminate measurement errors, displacements of less than 0.1 mm have not been taken into account. This system can be used with fixation of measurement points using steel screws on both sides of the fissure or with perforated steel buttons attached using epoxy resin. In this case, we have chosen the latter option. The measurement points have been preferably placed on concrete sheets affected by fractures on the exterior of the homes, so as to forestall the effects of expansion and contraction of the homes as much as possible. There was previous evidence of movement thanks to plaster cores placed by the homeowners when the fissure movements began in March 2022 (Figure 10A), with accumulated displacements of 11 mm (measurement carried out with a caliper before placing precision fissurometer measurement points) until March 2023. The measurement stations for the precision fissurometer were installed in March 2023, after two data collection campaigns, one in May and another in November 2023. We have 15 measurement stations in the TZF and 25 in the MZF. The results obtained offer maximum displacement speeds in both the TZF and the MZF of 2.80 mm/year and minimum speeds of 0.24 mm/year. No notable seismic activity has been detected, so these movements must be associated with aseismic creep. This study is in the preliminary phase and will require mid- and long-term monitoring to obtain results that give an idea of the temporal evolution of both faults.
The fracturing of the island of La Palma has played an important role in the Tajogaite eruptive process of 2021. The data available before the eruption revealed the areas subject to extension that could be susceptible to hosting volcanic eruptions (344 data on fault/striation pairs distributed into 39 measurement stations [[
From these results, one can infer that both fields have coexisted since the beginning of the volcanism that generated the Canary Islands archipelago, since the two data populations can be observed from materials belonging to the ancient massifs to the most modern materials. The Atlantic Field may be linked to the opening of the ridge and the drag of the oceanic crust, while the local one might be conditioned by the great lines of weakness in a NE-SW direction (abyssal hills parallel to the ridge axis), generated by the opening of the ridge in the oceanic crust.
On the other hand, the combination of surface fracture data generated after and during the eruption and the seismicity data made it possible to identify two active faults that conditioned the eruption during the three months it lasted. The first fault to which seismicity is adapted is in the NW-SE direction and we have called it Tazacorte Fault (TZF). On 1 October, there began to be a variation in the orientation of the distribution of hypocenters, adapting to a fault with a mean ENE-WSW direction that we have called Mazo Fault (MZF) and which generated the first distal lava emission centers on 25 November. Surface fractures were mapped with field and drone work throughout the eruption. There are 326 data on faults associated with the MZF and 36 with the TZF. The emission centers of the main cone zone were then placed parallel to echelon fractures of the TZF, with the appearance of more distal centers distributed on these relay faults indicating dextral movement of the TZF. The same occurred in the MZF, in which the distal lava emission centers opened on normal-directional echelon faults associated with the dextral movement of the MZF. In both cases, this seems to be transtensive activity on the part of these two faults (Figure 13). The movement of these two faults could be compatible with the deformation field called the Atlantic Strain Field.
Aerial photographs of the 1946–1956 American flight have been used to study the fracturing prior to the 2021 Tajogaite eruption and contemporary with the San Juan eruption (1949). From these images, 191 lineations that are compatible with surface fractures have been identified (Figure 13). The average orientation of these fractures is E-W, which is fundamentally compatible with the MZF, although in the N zone of Tajogaite lava flows, the orientations of the fractures tend to be parallel to the TZF. This seems to indicate a prior conditioning of surface fracturing by both active faults. The route of the TZF in a NW-SE direction can be extended with the eruptive fractures that gave rise to the eruption of the San Juan volcano in 1949. These eruptive fractures of the San Juan volcano have the same NW-SE orientation [[
The activity of the TZF and the MZF has continued after the eruption, with creep movement without associated seismicity. Deformations in houses and human constructions have been observed in the surface route of these two faults. It has been confirmed that these fractures have been repaired before the 2021 eruption, so their activity has been continuous over time. In order to quantify these displacements, precision fissurometer measurement stations have been installed. There are 15 measurement stations in the TZF and 25 in the MZF. This study should be extended in the mid and long term, but to date, the data available provide maximum movement speeds of 2.80 mm/year and minimum speeds of 0.24 mm/year. This gives an idea of the intense activity of these faults, which has persisted over time, as well as the possibility of a flank eruption on the western flank of the area of the last eruption of 2021, with a movement of blocks towards the W or the SW, similar to that interpreted by [[
Utilizing structural analysis of tectonic strain fields in active volcanic regions enables us to forecast areas with a high likelihood of volcanic center formation and potential dike emplacement orientations.
Two primary tectonic strain fields have been identified on La Palma island: a regional Atlantic extensional tectonic field trending NW-SE, and a local tectonic field characterized by a NE-SW trend.
The Atlantic tectonic field plays a crucial role in the eruption dynamics of La Palma, as evidenced by the orientation of the main dike emplacement (NW-SE) during eruptions, constrained by an extensional tectonic regime.
The Cumbre Vieja eruption and the formation of the Tajogaite volcanic edifice were primarily influenced by two main faults: the NW-SE-trending Tazacorte Fault and the ENE-WSW-trending Mazo Fault. Both faults have the potential to trigger moderate earthquakes. Additionally, post-eruption creep movements have been mapped, underscoring the continuous volcanic activity beyond the eruptive phase. This information is critical for post-disaster reconstruction efforts and for restoring normalcy for the island's inhabitants between volcanic eruptions.
Graph: Figure 1 Geographical and geological setting: (A) geographic location of the Canary Islands archipelago; (B) location of La Palma island within the Canary Islands archipelago; (C) digital elevation model (DEM) of the island of La Palma and the location of the area covered by lava flows from the 2021 Tajogaite volcanic eruption; (D) detailed view of the extent of the lava flows and cinder cone of the 2021 Tajogaite eruption.
Graph: Figure 2 Examples of normal faults in La Palma (DEM of La Palma with the photograph location): (A) conjugate normal faults affecting pyroclastic deposits (Taburiente volcanic edifice (superior), Barlovento, La Palma); (B) slickensides on normal fault plane affecting basaltic lava flows (Bejenado volcanic edifice, Barranco de las Angustias, La Palma).
Graph: Figure 3 Stereonet of fault measuring stations (Wulff net, lower hemisphere) and representation of Dey obtained from the Slip Model (red arrows).
Graph: Figure 4 Data points and results for the fault population analysis in La Palma Island (the dashed lines represent Dey's mean trajectories.): (A) Atlantic Strain Field and (B) Local Strain Field.
Graph: Figure 5 Normal fault planes of the Roques of Jedey, Jedey eruption (1585), and FPA analysis: (A) fault planes and slickensides; (B) Right-Dihedron Method solution; (C) Slip Model solution; (D) average direction of maximum horizontal shortening (Dey) obtained from the Slip Model.
Graph: Figure 6 Spatial evolution of the hypocenter distribution during the 2021 eruption: (A) 12–19 September; (B) 19 September–1 October; (C) 1–10 October; (D) 10 October–13 December. See text for further explanation.
Graph: Figure 7 Photograph of the cinder cone with the main eruptive vents aligned according to the TZF orientation (NW-SE), from RPA aerial view (23 October 2021).
Graph: Figure 8 Surface reflection of fracturing development: (A) strike-slip fault with pull-apart basins to the south of the lava field; (B) strike-slip fault affecting an agricultural irrigation pond; (C) normal fault on Cogote Mountain.
Graph: Figure 9 Fault mapping of TZF and MZF and spatial distribution of stations for measuring surface fractures of endogenous origin generated during the eruption. The red color represents the measurement stations operated in lava emission centers.
Graph: Figure 10 Transtensional strike-slip fault (MZF) with development of normal faults in echelon, in the southern sector of the lava flows: (A) orthoimage by drone (the red arrows indicate the sense of movement of the fault) and (B) field view of (A), normal faults in echelon affecting ash fallout deposits from the eruption itself (the red arrows indicate the surface fault trace).
Graph: Figure 11 Fault mapping of TZF and MZF and spatial distribution of Right-Dihedron Method solution (pair of data fault plane/slickenside) stations for measuring surface fractures of endogenous origin generated during the eruption.
Graph: Figure 12 Faults suggesting creep movement of the TZF and MZF, before and after the volcanic eruption. Northern sector of the lava flows: (A) faults in the houses repaired since 1980 with the same orientation as the TZF; (B) close view of the plaster cores on fractures generated in houses; at present it continues its creep movement after the volcanic eruption. Southern sector of the lava flows: (C) fracture repaired before the eruption (
MAP: Figure 13 Structural map of the main deformation bands associated with the TZF and MZF. The rose diagram represents the average orientation of previous fractures obtained from aerial photography of 1946–1956.
Field campaign for data record: M.Á.R.-P., M.Á.P., R.P.-L., N.S., J.L.-G., I.G. and M.B.; Structural analysis (Fault Population Analysis): M.Á.R.-P. and R.P.-L.; Photointerpretation of historical aerial photography: J.F.M., M.Á.R.-P. and G.L.; Contextualization and bibliographic compilation: M.Á.P. and D.S.-M.; Drone work: J.C.G.-D. and C.L.C.; Mapping and GIS (Geographic Information System), G.L., M.Á.R.-P. and J.F.M.; Model proposed: M.Á.R.-P., R.P.-L., M.Á.P., N.S. and I.G. The final model was approved by all the authors, and the manuscript and figures as well. Manuscript redaction: M.Á.R.-P., R.P.-L., M.Á.P. and N.S.; Figures: M.Á.R.-P., M.Á.P. and G.L. All authors have read and agreed to the published version of the manuscript.
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to the project is currently under implementation.
The authors declare no conflicts of interest.
We greatly thank the Cabildo de La Palma (La Palma Regional Government), the Dirección General de Emergencias (Directorate General for Emergencies of the Government of the Canary Islands), and the Ayuntamiento de El Paso (El Paso City Council) for their collaboration during the field work collecting data. Thanks to Mayte Pedrosa for her help in the elaboration of the figures.
By Miguel Ángel Rodríguez-Pascua; Raúl Perez-Lopez; María Ángeles Perucha; Nieves Sánchez; Julio López-Gutierrez; José F. Mediato; David Sanz-Mangas; Gonzalo Lozano; Inés Galindo; Juan Carlos García-Davalillo; Carlos Lorenzo Carnicero and Marta Béjar
Reported by Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author; Author