UC Berkeley

UC Berkeley Previously Published Works

Title
Crustal structure and upper mantle anisotropy of the Afar triple junction

Permalink
https://escholarship.org/uc/item/6tf4n3jg

Journal
Earth, Planets and Space, 73(1)

ISSN
1343-8832

Authors
Kumar, U
Legendre, CP
Huang, BS

Publication Date
2021

DOI
10.1186/s40623-021-01495-0

Copyright Information
This work is made available under the terms of a Creative Commons Attribution License,
availalbe at https://creativecommons.org/licenses/by/4.0/

Peer reviewed

eScholarship.org Powered by the California Digital Library

University of California
Kumar et al. Earth, Planets and Space (2021) 73:166
https://doi.org/10.1186/s40623-021-01495-0

FULL PAPER Open Access

Crustal structure and upper mantle

anisotropy of the Afar triple junction
U. Kumar1, C. P. Legendre1,2* and B. S. Huang1

Abstract
The Afar region is a tectonically distinct area useful for studying continental break-up and rifting. Various conflicting
models have been suggested to explain the lateral variations of the anisotropy in this region. To address this issue,
we investigated the tectonics of the Afar region using receiver function and shear-wave splitting measurements
based on broadband seismic data from 227 stations in the region. Further, the receiver function results were inverted
to obtain the crustal thickness and Vp/Vs ratio of the region. Our results reveal a thick African crust (thicker than 40
km) with typical Vp/Vs values for the continental crust, elongated down to 21 km along the rift system with very high
Vp/Vs values near the fractured zones, suggesting crustal thinning near the fractured zones. Our shear-wave splitting
measurements indicate a general fast axis orientation of N030E. However, substantial disparities in the fast anisotropy
direction exist in the triple junction region, with some stations displaying a direction of N120E, which is perpendicular
to the fast directions measured at the surrounding stations. In addition, many stations located close to the rifts and
within the Arabian Plate provide mostly null measurements, indicating the presence of fluids or isotropic media. This
study uses several methodologies to unravel the structure and evolution of the Afar region, providing valuable insight
into the Afar, a tectonically distinct region, which will be useful for elucidating the mechanisms and characteristics of
a continental break-up and the rifting process.
Keywords: Afar, Continental break-up, Continental rifting, Crustal structure, Receiver function, Shear-wave splitting

Introduction called East African Rift System), an active continental rift

Geological settings zone in East Africa, is a divergent tectonic plate bound-
The Afar triple junction (Fig. 1) is located in East Africa ary (Kusky et al. 2010) dividing the African Plate into
connecting three branches of a complex rift system. This two tectonic units: the Nubian and Somalian plates (Gar-
area is an example of the last stage of continental rifting funkel and Beyth 2006). The Afar triple junction accom-
and the early stage of seafloor spreading, leading to the modates the divergent motions between the Arabian,
creation of oceanic crust (Mohr 1970; Tesfaye et al. 2003; Nubian, and Somalian plates along the Red Sea, Gulf of
Rychert et al. 2012). The Afar hotspot is located at the tri- Aden, and East African rifts. The kinematics of the Afar
ple junction between the Red Sea rift, the Gulf of Aden, triple junctions are generally studied using long-term
and the Ethiopian rift zone (Mohr 1970; Legendre 2013). deformation and geodetic observations (McClusky et al.
The Red Sea rift is considered as the boundary between 2010) as well as geodynamical modeling (Koptev et al.
the Arabian and African (Nubian) plates, whereas the 2018).
Gulf of Aden separates the Arabian and African (Soma- The opening of the Red Sea and the Gulf of Aden
lian) plates (Bird 2003). The Ethiopian rift zone (also started in the Late Eocene–Early Oligocene (Ghebreab
1998) and lead to the split between Africa and Arabia
in the Early Miocene (Joffe and Garfunkel 1987). The
*Correspondence: [email protected] early stage of the collision between the Mediterranean
1
Institute of Earth Sciences, Academia Sinica, Taipei, Taiwan and Biltis segments (McQuarrie et al. 2003) occurred
Full list of author information is available at the end of the article

© The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/.
Kumar et al. Earth, Planets and Space (2021) 73:166 Page 2 of 17

Fig. 1 Tectonic setting of the Afar triple junction. The red triangles are active Holocene volcanoes (Venzke 2013). The magenta line represents the
plate boundary (Bird 2003). The black circles represent the locations of the epicenters of Earthquakes (Mw > 4). Most of the Earthquakes in the
region have shallow depths. The black text framed in magenta provides the names of the different plates (Bird 2003): AFn, African (Nubian) plate;
AFs, African (Somalian) plate; Ar, Arabian plate. The green text framed in blue provides the names of the three rifts: ER, Eden rift; GA, Gulf of Aden;
R,- Red Sea rift. The blue text framed in brown (Af ) indicates the location of the Afar triple junction

synchronously with the massive eruption of flood basalt rotation with respect to the Nubian Plate (with a rotation
associated with the Afar plume at around 30 Ma (Hof- pole located in the central Red Sea). Doubre et al. (2017),
mann et al. 1997). based on a combination of GPS and Interferometric Syn-
thetic Aperture Radar (InSAR) measurements, suggested
Previous geophysical studies that the northern part of the Somalia plate evolves inde-
Several geophysical tools have been employed in this pendently from the entire African (Somalian) plate.
region to investigate its tectonics. In recent decades, Previous seismological models from body-wave tomog-
the Global Positioning System (GPS) has been widely raphy (Benoit et al. 2006; Koulakov 2007) and surface-
deployed in this region to investigate the relative motions wave tomography (Sebai et al. 2006; Guidarelli et al. 2011;
of the different tectonic units (Walpersdorf et al. 1999; Legendre 2013) corroborated the presence of negative
McClusky et al. 2010; Doubre et al. 2017). velocity anomalies beneath the triple rift systems. Also,
The main results of geodetic measurements along the they have suggested that the Afar hotspot and other vol-
Afar triple junction have suggested counterclockwise canic provinces in the region (Wignall 2001; Mège and
Kumar et al. Earth, Planets and Space (2021) 73:166 Page 3 of 17

Korme 2004) are originating from the African super- rigid blocks in the region. Further geophysical modeling
plume (Ni et al. 2002; Simmons et al. 2007), upwelling (Simmons et al. 2007; Reilinger and McClusky 2011) has
from the lower mantle. enabled the kinematics and dynamics of the region to
The crustal thickness in the region is of great interest as be refined. However, conflicting models have been sug-
it is crucial to unraveling the mechanisms of the tectonic gested to explain the lateral variations of the anisotropy
evolution of the Afar triple junction. The Moho depth in the Afar triple junction region. Gao et al. (2010) sug-
has been widely investigated using the receiver function gested that the optimal source of anisotropy is located
technique (Dugda et al. 2005; Dugda and Nyblade 2006; at approximately 300 km depth (in the asthenosphere),
Hammond et al. 2011; Reed et al. 2014; Thompson et al. whereas Keir et al. (2011) found that this anisotropy orig-
2015) and the crustal thickness ranges from ~ 15 km inates mostly from deformation-related structures in the
beneath the spreading centers up to ~ 40–45 km outside crust and lithosphere.
the continental break-up area (Hammond et al. 2011). In this study, we performed a systematic investigation
Previous anisotropy observations have mainly been of the crustal structure and mantle anisotropy in the Afar
achieved by the shear-wave splitting of teleseismic region by means of receiver function and shear-wave
(Gashawbeza et al. 2004; Walker et al. 2005) or local splitting analyzes. Knowledge about the crustal thick-
Earthquakes (Keir et al. 2011). Several other shear-wave ness as well as crustal and lithospheric anisotropies (Leg-
splitting measurements performed in the region and sur- endre et al. 2017, 2020) can provide helpful insight into
roundings have been compiled into international shear- the crustal and mantle flow affecting the regional dynam-
wave splitting databases (Wüstefeld et al. 2009; IRIS ics (Legendre et al. 2016; Fan et al. 2020). The main con-
2012). Deformation-related structures generally explain tribution of this study is a total of 431 new shear-wave
the anisotropic pattern in the Afar region. The general splitting measurements in the region as well as 34 meas-
orientation of the anisotropies is associated with mantle urements of crustal thickness and ratio of seismic com-
flow; in contrast, subtle changes in the fast direction are pressional and shear-wave velocities (Vp/Vs). In the first
related to the presence of fluids, isotropic media, or verti- step, we will describe the data and methodologies used in
cal mantle flow. Dyke-induced faulting and alignment of this study (“Data and methods”). “Results” presents our
melt pockets near volcanic centers also have very distinct new findings based on crustal thickness, velocity ratio
anisotropic patterns in the region (Gashawbeza et al. between compressional waves and shear waves as well as
2004). shear-wave splitting of the SKS phase. “Discussion” dis-
V
cusses the implications of our computed crustal depth, Vps
Tectonics of the Afar triple junction ratio, and anisotropy in terms of the structures, composi-
The tectonics of the Afar triple junction is dominated by tion, and evolution of local and regional features.
the motion of three plates: Arabia, Africa (Nubia), and
Africa (Somalia). The first rift to open is thought to have Data and methods
been the Gulf of Aden around 30 Ma (Ghebreab 1998). Seismological data
Simultaneously, the intense activity of the Afar hot- We searched for all available broadband seismic stations
spot (Schilling et al. 1992; Hofmann et al. 1997) led to a deployed in the study region (Fig. 2). In total, 227 stations
massive amount of flood basalt in the region. The open- were found, but only 223 used for receiver functions and
ing of the Red Sea rift (Kusky et al. 2010) is thought to 224 used for shear-wave splitting constraints provided
have started between 24 and 21 Ma. This event was fol- sufficient data for further investigation. To compute the
lowed by high magnetic anomalies (Wignall 2001; Mège receiver functions, we selected each available Earthquake
and Korme 2004) and subsequently the opening of the with Mw > 5.5 and an epicentral distance between 30° and
Ethiopian rift between 18 and 15 Ma (Kusky et al. 2010). 90°, and to perform the shear-wave splitting measure-
The local seismicity (Abdallah et al. 1979; Illsley-Kemp ments at each station, we selected Earthquakes with Mw
et al. 2018), surface displacement (Walpersdorf et al. > 6.5 with epicentral distances ranging from 90° to 120°.
1999; Kusky et al. 2010) and geometry of the active faults We selected the accessible Earthquakes independently
(Doubre et al. 2017) were used to fine-tune the geom- for each station. The number of Earthquakes mainly
etry and complexity of the tectonic units interacting in varies with the station operating time. We selected 427
the region. Further information from body-wave tomog- events among all stations to compute the receiver func-
raphy (Benoit et al. 2006) and surface-wave tomogra- tions and 514 for the shear-wave splitting measurements.
phy (Guidarelli et al. 2011), as well as receiver function Waveform data were downloaded from global open-
results (Hammond et al. 2011), have provided additional access data centers (mostly from the GEOFOrschung-
constraints on the nature of these tectonic units, such as sNetz program (GEOFON), GeoForschungsZentrum
their lateral and vertical extensions and the presence of Potsdam, Observatories and Research Facilities for
Kumar et al. Earth, Planets and Space (2021) 73:166 Page 4 of 17

Fig. 2 Locations of the 227 seismic stations deployed in Afar region. The color code indicates the networks to which the stations belong

European Seismology Data Center and Incorporated The ZRT coordinates (vertical, radial, and transverse
Research Institutions for Seismology Data Management components) were conveniently adopted for the nearly
Center) (Megies et al. 2011; Trabant et al. 2012). vertical incidence of teleseismic events. Assuming the
location of the station–hypocenter pair is known, it
is possible to infer the wave coordinate system. The L
Receiver functions component is associated with the P-wave polarization,
Receiver functions are time series representing the inter- whereas the Q component is tied to ­SV-wave polariza-
nal boundaries of the Earth near the receiver, convert- tion, and the T component is related to the ­SH-wave
ing the incident P waves into S waves, or vice versa. The polarization.
receiver function method is well established and widely In this study, the particle motions of the incident and
used to investigate crustal and upper mantle velocity dis- converted waves were separated using the wave coordi-
continuities (Langston 1979; Ammon et al. 1990; Zhu and nates. To derive the P-receiver functions, the teleseis-
Kanamori 2000; Kind et al. 2015; Maguire et al. 2018). mic waveform of the L (or Z) component (assumed to
Herein, we illustrate the method implemented by resemble that of a P wave striking the conversion bound-
Eulenfeld (2020) using a single station, ZF-MAYE. Fig- ary) is deconvolved from the QT (or RT) components to
ure 3 summarizes the potential events selected for this separate the source effects of the converted phases. The
station and satisfying the magnitude and epicentral dis- resulting receiver function (the Q component in particu-
tance requirements. lar) reveals the delay time (relative to the P-wave onset)
Kumar et al. Earth, Planets and Space (2021) 73:166 Page 5 of 17

Fig. 3 Selected events used to compute the receiver functions for a single seismic station (ZF-MAYE). The seismic station is indicated by a black
triangle, and the locations of the events are marked by circles, color-coded by epicentral depth

and relative amplitudes of the P waves converted into S of the S v phase movement whereas T is aligned in the
waves (Ps) by the significant discontinuity, as well as mul- direction of the SH phase movement).
tiples (PPPmP PpSmP, PPPmS, and PpSmS) caused by rever- The L and Q receiver functions are obtained by decon-
berations within the layer. P indicates a P-wave in the volving the vertical component seismogram of a single
mantle, p indicates a P-wave ascending to the surface, S teleseismic event from the L and Q components seismo-
indicates an S-wave in the mantle, s indicates an S-wave gram, respectively, using the time domain source equali-
ascending to the surface, whereas m indicates a top-side zation method of Langston (1979), as shown in Eq. (1):
reflection from the Moho.
L(ω)
Those phases are not easily identified on a seismo- RFL (ω) = c(ω) · ,
gram, because they generally have low amplitude and Z(ω)
Q(ω)
(1)
are hidden in the coda of teleseismic P-wave coda. A RFQ (ω) = c(ω) · .
conventional approach is to build the receiver function, Z(ω)
and to highlight those compressional motion converted
Figure 4 displays the receiver functions computed at the
into shear-polarized motion. The underlying assump-
selected station (ZF-MAYE).
tion is that the vertical component is primarily related
to the compressional motion. Thus, deconvolving it
from the horizontal components removes the signature
H–κ stacking
of the Earthquake source (present in both horizontal
We stacked all receiver functions that were reasonable
and vertical components of the seismogram), as well as
for each station to produce an average receiver function
all compressional reverberations. The resulting product
for that station (Fig. 4, top). Stacking receiver functions
of the deconvolution is the time series of shear motion
enhances the signal-to-noise ratio of the trace, thereby
within the teleseismic P-wave coda. More details about
increasing the amplitude of the converted phases in the
the deconvolution process are available at Walpole et al.
receiver function. Theoretically, all converted phases
(2014) and Kumar and Legendre (2021).
should be visible on the stacked receiver function. How-
The horizontal components of the seismogram are
ever, some converted phases have very low amplitudes
rotated into LQT components (L is aligned in the direc-
and are not easy to discriminate.
tion of P wave propagation; Q is aligned in the direction
Kumar et al. Earth, Planets and Space (2021) 73:166 Page 6 of 17

Fig. 5 (Bottom) Stack of all receiver function for station ZF-MAYE

(orange line) and automatically picked phases (blue cross and
their name in black font). (Top) H–κ grid search following Zhu and
V
Kanamori (2000). The best combination of Moho depth and Vps ratio is
indicated by a black circle

To ensure the reliability of the H–κ stack, we selected

only the receiver functions that were of high quality and
discarded those with more uncertainty. In a first step, we
discarded all receiver functions that had a signal-to-noise
ratio lower than 3. Only the receiver functions with clear
direct P, PS, and ­PPPS phases were retained. We used an
automated tool (Eulenfeld 2020; Kumar and Legendre
2021) to identify the potential phases in the stacked seis-
mogram (Fig. 5, bottom) where only the phases with strong
amplitude contrasts were easily recognized. We set the
first peak with the highest amplitude as the direct P-wave
arrival. The first peak with an amplitude more significant
than 10% of the P-wave arrival amplitude was considered
to correspond to the P-to-S converted (PS) wave. The sec-
ond peak with an amplitude greater than 5% of the P-wave
arrival amplitude was considered the ­PPPS wave.
The conventional H–κ stacking method (Zhu and Kan-
amori 2000) requires another phase: the ­PpSs or ­PsPS that
arrive 15–20 s after the direct P arrival. However, this
phase has a negative amplitude and was not always picked
by our automated picking method. Therefore, we only use
those three arrival times to constrain the Moho depth (H)
V
as well as the Vps ratio (κ), following:

Fig. 4 (Left) Q component of the receiver function for the seismic (t1 − t0 )
H= ,
station indicated in Fig. 3. (Right) All 155 receiver functions sorted by

( Vp ) − p2 − ( V1p )2 − p2
κ 2
distance (purple dots) and back-azimuth (blue dots). (Top) Stacked
receiver function (t2 − t0 )
(2)
H=  ,
( Vκp )2 − p2 − ( V1p )2 − p2
Kumar et al. Earth, Planets and Space (2021) 73:166 Page 7 of 17

where t0 is the arrival time of the P wave, t1 is the arrival Shear‑wave splitting
time of the PS wave, t2 is the arrival time of the ­PpPs wave, Shear-wave polarization anisotropy from core phases
Vp is set to 6.3 km/s (based on the IASP91 reference is a well-established concept (Ando et al. 1983; Fukao
V
model from Kennett et al. (1995), κ is the Vps ratio. H is 1984; Obrebski et al. 2010). To download and process
the Moho depth (in km). A 65% weight was given to the the data, we followed the workflow of STADIUM-py
PS arrival time in Eq. (2) and 35% was given to the ­PpPs (Walpole et al. 2014).
arrival time, following Zhu and Kanamori (2000). Note For each seismic event that matches the magnitude
that the calculated crustal thickness is strongly influ- and epicentral distance requirements (Fig. 6), ZNE raw
enced by the initial velocity model, which is also affected data were downloaded and rotated from ZNE to ZRT
by the thickness of the sedimentary layer. components. The predicted arrival time of the SKS
For each station where the stacked receiver function phase was calculated using the IASP91 reference model
displayed enough amplitude on the selected phases, both (Kennett et al. 1995). The traces were cut around the
V
Moho depth (H) and Vps ratio (κ) could be retrieved, as SKS phase predicted arrival time (from 30 s before to
displayed in Fig. 10. 45 s after the predicted time), as displayed in Fig. 7.
Some receiver functions failed to provide enough We then used an automatic selection approach to
energy at the predicted arrival times for the ­PPPS phase identify the beginning and ending times of the SKS
due to the instability of the deconvolution (Kind et al. wave on the radial component (Fig. 8). Characteris-
2015). Of the 223 seismic stations that provided data tic functions calculated using short-term-average/
for receiver function inversion, 34 also supplied meas- long-term-average (STA/LTA) algorithms are often
urements for the H–κ stacking method (Zhu and Kan- employed for automated detection of P and S waves
amori 2000). Most stations provided receiver functions and arrival time estimation (Walpole et al. 2014). This
results, but we imposed highly restrictive quality checks study used an automatic selection method based on an
to ensure that the H–κ stacking method would yield con- STA/LTA algorithm to determine the SKS arrival time
sistent results. The quality checks consist of two steps: accurately. Two threshold values in the characteristic
first, the receiver function requires a signal-to-noise ratio function were set to pick the beginning and end times
higher than 3. Then all the phases used for the H–κ stack- of the SKS wave accurately (Walpole et al. 2014): the
ing method need to be consistently determined for each first threshold was for the beginning of the SKS phase,
station specific stacked receiver function.

Fig. 6 Selected events for the computation of shear-wave splitting parameters for a single seismic station (ZF-MAYE)
Kumar et al. Earth, Planets and Space (2021) 73:166 Page 8 of 17

Fig. 7 Raw waveforms of unfiltered ZNE components for an event recorded at station ZF-MAYE (8/9/2009, 10:55:56.28, epicenter (33.1474,
138.0594), epicentral depth of 302.2 km, and Mw =7.1

Fig. 8 Automatic selection of the SKS phase using the recursive STA LTA method for the trace depicted in Fig. 7. The red and blue bars indicate the
beginning and ending times of the SKS phase, based on values of the characteristic function (bottom)

and the second was for the end of the SKS phase (red search successfully, we set some thresholds on the errors
and blue lines in Fig. 8, respectively). in the phase (δφ ) and the delay time (δt ). Measure-
Once the SKS phase had been selected, we performed ments with excessive uncertainties in the fast directions
a grid search (Fig. 9) for the best-fitting splitting param- (δφ > 7◦) and the delay time (δt > 1.5s) were also been
eters following (Walpole et al. 2014). To pass the grid discarded. Those threshold values in the phase (δφ ) and
Kumar et al. Earth, Planets and Space (2021) 73:166 Page 9 of 17

Fig. 9 (Top left) North and East components of the seismogram (same traces as in Figs. 7 and 8), with corresponding particle motion (center).
(Bottom left) Radial and transverse components of the seismogram, with corresponding particle motion (center). (Right) Grid search results used to
find the best-fitting parameters for the phase and delay time

the delay time (δt ) are quite conservative and ensure 20–25 km near the active rift segments and observed a
that the final measurements do not include too many northward thinning of the crust beneath the Afar triple
uncertainties. junction.
Vp
This grid search (Fig. 9) yielded the fast shear phase Vs ratio is expected to be in the range 1.7–1.9 in the
(φ ) and a delay time (δt ). The procedure was repeated at continental crust (Zandt and Ammon 1995; Musacchio
each station and for all respective events. In the end, the et al. 1997; Kandilarov et al. 2015); most of the measure-
V
average of all the measurements was computed at each ments in the region (Fig. 11), the value of Vps is reasonable.
station. However, the measurements observed near the continen-
tal rift display relatively high values (greater than 2.0).
Note that the receiver function results are not displayed
Results for all stations, as their main primary purpose was to
We obtained successful measurements of H and κ for compute H and κ using the H–κ stacking method (Zhu
a total of 34 stations. Figures 10 and 11 display the and Kanamori 2000).
receiver function inversions for H and κ , respectively. Figure 12 summarizes the shear-wave splitting meas-
Our results suggest that regional thick continental urements, and Table 1 lists them in detail. In most of
crust (with a thickness of up to 38 km) exists beneath the regions on the Nubian and Somalian plates, the fast
the African and Arabian plates, which agrees with pre- direction is oriented N030E. This fast direction orienta-
vious measurements (Dugda et al. 2005; Dugda and tion is highly consistent with previous shear-wave split-
Nyblade 2006; Hammond et al. 2011; Reed et al. 2014; ting measurements (Wüstefeld et al. 2009; IRIS 2012).
Thompson et al. 2015). In contrast, a very thin crust Along with some additional measurements, we provide
(with a thickness as low as 21 km) is found along the detailed information by classifying our results into three
three branches of the continental rifting and beneath categories:
the Afar triple junction. This thin crust is consistent
with the very thin crustal thickness found beneath the • no data, unreliable measurements (due to insufficient
spreading centers and the area of continental break-up operation time),
(Hammond et al. 2011). Hammond et al. (2011) found • new measurements computed in this study,
that the crustal thickness ranged from 16 km down • and Null measurements (if number of Null meas-
the spreading centers in the northern Afar region to urements (Wüstefeld and Bokelmann 2007) was
Kumar et al. Earth, Planets and Space (2021) 73:166 Page 10 of 17

Fig. 10 Moho depths computed by the H–κ stacking method (Zhu and Kanamori 2000) near the Afar triple junction

more significant than the number of regular meas- measurements or no measurements (e.g., beneath the
urements). Arabian Plate), which contrasts with the high scatter-
ing present in previous measurements (Wüstefeld et al.
Null measurements can occur for several reasons, such 2009; IRIS 2012). In the null measurements, we do not
as limited data coverage and azimuthal distribution of report any direction of ( φ ) or ( δt ) in the station sum-
seismic events. They can also occur if the wave propa- mary (Table 1).
gates through an isotropic medium, if the wave encoun-
ters fluid, if mantle upwelling occurs with a vertical
flow of peridotite, or if the initial polarization direc- Discussion
tion is oriented along either the fast or slow axis. As the Lateral variations of Moho depth
incoming shear wave is not split (Savage 1999), it is not The results we obtained by inverting the Moho depth
possible to determine or accurately. However, in the receiver functions (Fig. 10) agree well with previous
central part of the rift, there are several stations with measurements (Dugda et al. 2005; Dugda and Nyblade
fast directions perpendicular to those of the others, at 2006; Hammond et al. 2011; Reed et al. 2014; Thomp-
N120E. For several stations, we could only retrieve null son et al. 2015). A thick African crust (40–45 km) is
Kumar et al. Earth, Planets and Space (2021) 73:166 Page 11 of 17

Vp
Fig. 11 Vs ratio (κ) computed by the H–κ stacking method (Zhu and Kanamori 2000) near the Afar triple junction

observed near the large igneous volcanic provinces Vp

ratio reveals the presence of fluids
Vs
(Wignall 2001; Mège and Korme 2004) beneath the Vp
stations located in the continental crust of the Nubian Vs ratio has been the topic of numerous studies in oceanic
and Somalian plates. These results contrast strongly (Clague and Straley 1977; Hyndman 1979; Bloch et al.
with the Moho depths we found along the rifts, with a 2016), continental (Zandt and Ammon 1995; Musacchio
crustal thickness of 21–25 km. This strong dichotomy et al. 1997; He et al. 2014), and passive margin domains
V
between the unaltered continental crust and the con- (Kodaira et al. 1996; Kandilarov et al. 2015). Generally, Vps
tinental crust thinned by the active continental rifting has been in the range of 1.7–1.9 for the continental crust.
(Kusky et al. 2010) confirms the theory of the African These values agree well with the results obtained for the
Plate being split into two tectonic units: the Nubian stations located far away from the rift system (Fig. 11).
and Somalian plates (Garfunkel and Beyth 2006).
Kumar et al. Earth, Planets and Space (2021) 73:166 Page 12 of 17

Fig. 12 Shear-wave splitting results for the broadband stations around the Afar triple junction. Blue: previous measurements available from
international databases (Wüstefeld et al. 2009; IRIS 2012). Yellow: due to the data quality, no convincing measurements could be obtained. Green:
null measurements. Black: each line indicates the mean fast polarization direction given by the median of all observations at one station. The circle
size is proportional to the delay time

V
However, Vps increases drastically in the regions affected fast anisotropy axis throughout the African plate. How-
by active rifting (reaching 2.1, as high as the values found ever, some scattered measurements exist, mainly beneath
for oceanic domains). This strong contrast of between the Arabian Plate. Our results (Fig. 12) display a similar
the unaltered continental crust and thinned crust regions trend for the stations located on both sides of the rift sys-
is highly consistent with the mechanical extension of tem beneath the African plates. However, beneath the
the continental thickness imaged by receiver functions Arabian Plate, the stations with reliable measurements
V
(Fig. 10). The anomalously high Vps , possibly caused by the show null measurements. This null identification sug-
reduction in shear-velocity (VS), can be plausible indirect gests that the Arabian block could be isotropic, as it has
evidence for fluids, melts, or oceanic crust within the rift not experienced substantial deformation since its forma-
system branches. tion. It is considered a substantially rigid block with lim-
ited internal deformation (Alothman et al. 2016; Tesauro
Complex anisotropy beneath the rift system et al. 2018) which is consistent with our results. In addi-
Previous shear-wave splitting measurements (Wüstefeld tion, the high level of inconsistency observed beneath the
et al. 2009; IRIS 2012) display a constant direction of the Arabian plate between the present results and previous
Kumar et al. Earth, Planets and Space (2021) 73:166 Page 13 of 17

Table 1 Summary of all shear-wave splitting parameters measured in this study for the Afar triple junction region
NET STA Long. Lat. φ δt nb nb Null

2H ABAE 39.7636 13.3535 14 0 1 3

2H AFME 40.8585 13.2040 – – 0 3
2H AHME 40.2784 14.0889 – – 0 3
2H ASYE 41.4422 11.5607 48 0.7 1 4
2H BTIE 40.0218 11.1949 – – 0 1
2H DALE 40.2178 14.2290 20 0.1 1 1
2H DAME 40.9625 11.6868 – – 0 1
2H ERTE 40.4969 13.4463 – – 0 5
2H GALE 40.3940 13.7251 14 1.1 2 3
2H GPSE 40.5283 13.4882 – – 0 7
2H GULE 39.5885 13.6945 – – 0 2
2H HITE 40.3169 13.1012 – – 0 1
2H IGRE 40.4612 12.2526 – – 0 1
2H KOZE 40.9849 12.4948 – – 0 2
2H LULE 40.7037 11.9892 – – 0 2
2H MAYE 39.5343 12.7832 – – 0 2
2H SAHE 40.9770 12.0402 17.95 1.5 6 2
2H SEME 41.0043 11.7926 20.45 1.1 3 4
2H TRUE 40.3148 12.4812 -40 3 1 3
4H NAB1 41.6554 13.3873 – – 0 2
4H NAB2 41.7146 13.4285 35.39 0.5 3 4
4H NAB3 41.7549 13.3759 12 0.3 1 1
4H NAB4 41.6786 13.4788 – – 0 1
4H NAB7 41.7562 13.4256 – – 0 2
5H ASSE 42.6545 13.0628 31 1 2 2
5H DOLE 39.9806 15.0968 27 0.9 2 4
5H FAME 41.5196 13.5684 – – 0 2
5H TIOE 40.8666 14.6651 – – 0 3
AF AAUS 38.7665 9.03490 43.61 0.9 5 10
AF ANKE 39.7418 9.58270 24 1.2 2 3
AF DESE 39.6350 11.1180 74 0.6 1 21
G ATD 42.8466 11.5307 44.73 1.5 20 18
GE DAMY 44.3917 14.5725 – – 0 8
IU FURI 38.6798 8.89520 30.68 1.2 7 31
XW ADEN 44.9824 12.7759 – – 0 3
XW ANID 43.2048 15.4736 – – 0 2
XW DABI 44.2690 15.1325 – – 0 2
XW HOTA 44.8757 13.0639 – – 0 1
XW RUSA 44.3539 14.7074 – – 0 1
XW SANA 44.2068 15.3926 – – 0 2
XW YSLE 44.2816 14.9387 – – 0 1
YJ ADEE 39.9068 7.79090 19 1.1 2 1
YJ ADUE 38.9019 8.54040 30 0.4 2 1
YJ ANKE 39.7339 9.59270 – – 0 3
YJ AREE 39.4188 8.93850 24 1.2 3 0
YJ ASEE 39.1317 7.97290 20 0.9 1 0
YJ BEDE 40.7710 8.90860 – – 0 1
YJ BUTE 38.3824 8.11700 – – 0 1
YJ DIKE 39.5566 8.06270 – – 0 1
Kumar et al. Earth, Planets and Space (2021) 73:166 Page 14 of 17

Table 1 (continued)
NET STA Long. Lat. φ δt nb nb Null

YJ DONE 39.5504 8.50900 – – 0 1

YJ DZEE 38.9959 8.78030 28 0.7 1 2
YJ GEWE 40.5743 10.0060 – – 0 2
YJ GTFE 39.8376 9.00340 22 1.5 1 0
YJ KARE 39.9349 10.4224 – – 0 1
YJ KOTE 39.3961 9.38750 – – 0 2
YJ LEME 38.6095 8.61150 30 0.5 1 1
YJ MECE 40.3241 8.59380 20 0.7 1 0
YJ MEKE 38.8330 8.16230 – – 0 3
YJ MELE 40.2008 9.31060 22 0.2 1 4
YJ MIEE 40.7581 9.24160 17 1.9 2 0
YJ NURE 39.7956 8.73120 – – 0 1
YJ SENE 39.0166 9.14660 33 0.5 2 2
ZE AFME 40.8585 13.2040 – – 0 7
ZE AWEE 40.0719 12.0673 – – 0 1
ZE BARE 40.3556 12.6443 28 0.1 1 1
ZE BERE 41.1886 12.1707 – – 0 1
ZE BREE 41.1883 12.1704 10 0.6 1 1
ZE DIGE 40.2734 12.3279 – – 0 1
ZE FINE 40.3160 12.0681 – – 0 3
ZE HARE 40.8840 11.6072 12.45 0.4 3 3
ZE MEGE 41.3381 11.4906 – – 0 1
ZE MILE 40.7650 11.4238 -24 1.6 1 4
ZE SEHE 40.9769 12.0401 44 1.4 1 2
ZE SEME 41.0044 11.7926 18 1.3 1 3
ZE SILE 41.1879 12.4070 112 1.4 1 3
ZE TRUE 40.3161 12.4841 – – 0 5
ZF ADTE 40.7570 11.1221 118 0 1 4
ZF ADYE 38.9811 13.6350 26 1.8 3 5
ZF AKEE 39.1680 10.8883 36 0.4 1 3
ZF ASYE 41.4422 11.5607 81 0.6 2 2
ZF AWSE 40.1700 8.98970 – – 0 1
ZF BTIE 40.0218 11.1949 40 0.9 1 3
ZF DERE 39.6353 11.1179 – – 0 5
ZF DICE 41.5737 11.9136 36 1.2 2 4
ZF ELLE 40.3784 11.2579 30 1.2 1 7
ZF ERTE 40.4969 13.4463 – – 0 2
ZF GASE 38.9211 11.6813 50.33 0.7 5 0
ZF GEWE 40.5739 10.0045 12 1.7 1 1
ZF HALE 40.0077 13.8422 – – 0 1
ZF HYNE 42.0962 9.31480 – – 0 3
ZF KOBE 39.6298 12.1506 21 1.1 2 5
ZF KORE 39.9344 10.4265 – – 0 4
ZF LALE 39.0375 12.0258 42 0.2 2 0
ZF LYDE 41.9257 12.0546 26 0 1 3
ZF MAYE 39.5343 12.7832 26 1.2 6 5
ZF MISE 40.7591 9.23690 10 3 1 2
ZF QATE 41.4691 9.37530 50 1.1 2 0
ZF SEKE 39.0334 12.6217 34.67 1.4 3 3
Kumar et al. Earth, Planets and Space (2021) 73:166 Page 15 of 17

Table 1 (continued)
NET STA Long. Lat. φ δt nb nb Null

ZF SMRE 39.2111 13.1977 32 1.3 1 4

ZF SRDE 41.3100 11.9579 – – 0 2
ZF WLDE 39.5874 11.8243 28 0.9 1 1
ZF WUCE 39.6064 11.5116 52 0.9 1 2
ZF YAYE 37.9965 11.8608 – – 0 1
The columns are for the network code, station code, longitude and latitude of the station, phase (φ ), delay time (δt ), amount of successful measurements and amount
of Null measurements. Directions of (φ ) and δt are not reported for stations considered to provide mostly null measurements

shear-wave splitting measurements (Wüstefeld et al. In addition to the lateral flow of the asthenosphere,
2009; IRIS 2012) displayed in Fig. 12, suggests that the mantle upwelling is also responsible for the presence
previous measurements in this region were not classified of fluids and melt, which is in agreement with the high
as null, despite the strong angular variations of the fast amount of null measurements, as well as the very high
direction of anisotropy. The highly scattered anisotropy values observed around the branches of the Afar rift
directions measured in previous shear-wave splitting system. Those observations are consistent with those of
studies strongly suggest the presence of isotropic media body wave (Benoit et al. 2006; Koulakov 2007) and sur-
beneath the Arabian block. face wave (Sebai et al. 2006; Guidarelli et al. 2011; Leg-
Close to the triple junction, most of the measurements endre 2013) tomographic models. The presence of very
are oriented N030E and agree with previous measure- slow velocities beneath the Afar region (for the global
ments. However, we also found a few measurements and regional models) and beneath the three branches
perpendicular to the general trend of N120E. The meas- of the rift system at all depths suggests the presence of
urements of those outliers are mainly located close to the fluids and melt, linked with potential mantle upwelling.
triple junction. Figure 12 also indicates that the stations Previous tomographic studies focusing on anisotropy on
located near the rifts tend to display null measurements. a large scale (Sebai et al. 2006) or local scale (Korostelev
Most stations with null measurements are located near et al. 2015; Sicilia et al. 2008) have revealed large regions
the active Holocene volcanoes (Fig. 1). Although null with surface wave slow anomalies that are associated
measurements are not present in the global database, with magmatism. Along the rift system, very slow veloci-
the presence of null measurements can be an important ties anomalies have been found, suggesting the presence
indicator of the presence of fluid or melt beneath the rift of either partial melt or fluids released by cooling mag-
branches. matic systems. The local seismicity (Abdallah et al. 1979;
Illsley-Kemp et al. 2018) is also very shallow, suggesting
Implications for local tectonics the presence of shallow reservoirs in the crust along the
In this study, 431 new shear-wave splitting measurements rift system. In those regions, our results indicate both
V
were obtained in the investigated region. The average fast thinner crust, higher values of Vps , and the presence of
direction measured is consistent with the results of Gao null measurements.
et al. (2010). However, the average delay time we found
is smaller than that obtained by Gao et al. (2010), which Conclusion
is close to the global average for continents (1.0 s calcu- In this study, we obtained receiver function and shear-
lated from Kennett et al. (1995) and Silver (1996). The wave splitting measurements for all 227 available broad-
fact that the average delay time is slightly smaller than band seismic stations deployed in the Afar region. The
the global average for continental crust can be explained receiver function results provide additional constraints
V
by the presence of thinned continental crust in our tar- on the crustal thickness and Vps ratio. The shear-wave
get region. The presence of several stations displaying splitting measurements provide some constraints on the
null measurements also agrees with the results of Gao average anisotropy over a column encompassing the crust
et al. (2010). The relatively homogeneous pattern of the and upper mantle beneath the seismic stations. Here, we
splitting observed in the region suggests that the main found that the thick African crust (with a thickness of
contributor to the anisotropy has a deep origin, with 40 km) has been strongly elongated and thinned (down to
local variations associated with crustal heterogeneities. a thickness of 20 km) across the rift system. In addition,
V
The relatively homogeneous pattern of the splitting is the exceptionally high Vps value within the crust along
also consistent with the model proposed by Barruol and the rifts and numerous null measurements of shear-
Ismail (2001), involving a NE flow in the asthenosphere. wave splitting for stations located near the rifts suggests
Kumar et al. Earth, Planets and Space (2021) 73:166 Page 16 of 17

the presence of melt and the early stage of oceanization. Benoit MH, Nyblade AA, VanDecar JC (2006) Upper mantle P-wave speed
variations beneath Ethiopia and the origin of the Afar hotspot. Geology
The results of this study provide essential constraints on 34(5):329–332
the structure of the crust and average anisotropy in the Bird P (2003) An updated digital model of plate boundaries. Geochem Geo-
Afar region. However, those measurements are limited phys Geosyst 4(3):1027
Bloch W, John T, Kummerow J, Wigger P, Salazar P, Shapiro S (2016) Watching
by the availability of the seismic station deployed in the dehydration: transient vein-shaped porosity in the oceanic mantle of the
region. Besides, those measurements provide a 1D snap- subducting Nazca slab. EGUGA, pp EPSC2016–6438
shot beneath each seismic station, and further modeling Clague DA, Straley PF (1977) Petrologic nature of the oceanic Moho. Geology
5(3):133–136
such as tomographic inversion could provide a 2D or 3D Doubre C, Deprez A, Masson F, Socquet A, Lewi E, Grandin R, Nercessian A,
view of the region. The strong lateral variations of crustal Ulrich P, De Chabalier JB, Saad I et al (2017) Current deformation in central
V
thickness, Vps and seismic anisotropy suggest a more com- Afar and triple junction kinematics deduced from GPS and InSAR meas-
urements. Geophys J Int 208(2):936–953
plex pattern with vertical variations that are not visible Dugda MT, Nyblade AA (2006) New constraints on crustal structure in eastern
from our results. However, this study provides accurate Afar from the analysis of receiver functions and surface wave dispersion
measurements of several geophysical parameters that are in Djibouti. Geol Soc Lond Spec Publ 259(1):239–251
Dugda MT, Nyblade AA, Julia J, Langston CA, Ammon CJ, Simiyu S (2005)
directly interpreted in terms of the structure and dynam- Crustal structure in Ethiopia and Kenya from receiver function analysis:
ics of the region. implications for rift development in eastern Africa. J Geophys Res Solid
Earth 110(B1):B01303
Acknowledgements Eulenfeld T (2020) rf: Receiver function calculation in seismology. J Open Sour
Source code, installation instructions and test input files for STADIUM-Py are Softw 5(48):1808
freely available for download (Kumar and Legendre 2021). Fan X, Chen QF, Legendre C, Guo Z (2020) Intraplate volcanism and regional
geodynamics in NE Asia revealed by anisotropic Rayleigh-wave tomogra-
Authors’ contributions phy. Geophys Res Lett 47(1):e2019GL085,623
UK, CL and BSH contributed to the design and implementation of the Fukao Y (1984) Evidence from core-reflected shear waves for anisotropy in the
research, to the analysis of the results and to the writing of the manuscript. All Earth’s mantle. Nature 309(5970):695
authors read and approved the final manuscript. Gao SS, Liu KH, Abdelsalam MG (2010) Seismic anisotropy beneath the Afar
depression and adjacent areas: implications for mantle flow. J Geophys
Funding Res Solid Earth 115(B12):B12330
This work is funded by the Ministry of Science and Technology under Grants Garfunkel Z, Beyth M (2006) Constraints on the structural development of Afar
MOST 107-2119-M-001-048; MOST 109-2811-M-001 -608- and MOST 109- imposed by the kinematics of the major surrounding plates. Geol Soc
2116-M-001 -021 -MY3. Lond Spec Publ 259(1):23–42
Gashawbeza EM, Klemperer SL, Nyblade AA, Walker KT, Keranen KM (2004)
Availability of data and materials Shear-wave splitting in Ethiopia: precambrian Mantle anisotropy locally
All data used in this paper came from published sources listed in the modified by Neogene rifting. Geophys Res Lett 31(18):L18602
references. Ghebreab W (1998) Tectonics of the Red Sea region reassessed. Earth-Sci Rev
45(1–2):1–44
Guidarelli M, Stuart G, Hammond JO, Kendall J, Ayele A, Belachew M (2011)
Declarations Surface wave tomography across Afar, Ethiopia: crustal structure at a rift
triple-junction zone. Geophys Res Lett 38(24):L24313
Competing interests
Hammond JO, Kendall JM, Stuart G, Keir D, Ebinger C, Ayele A, Belachew M
The authors declare no competing interests.
(2011) The nature of the crust beneath the Afar triple junction: evidence
from receiver functions. Geochem Geophys Geosyst 12(12):Q12004
Author details
1 He R, Shang X, Yu C, Zhang H, Van der Hilst RD (2014) A unified map of Moho
Institute of Earth Sciences, Academia Sinica, Taipei, Taiwan. 2 Institute of Geo-
depth and Vp/Vs ratio of continental China by receiver function analysis.
physics, Czech Academy of Science, Prague, Czech Republic.
Geophys J Int 199(3):1910–1918. https://​doi.​org/​10.​1093/​gji/​ggu365
Hofmann C, Courtillot V, Feraud G, Rochette P, Yirgu G, Ketefo E, Pik R (1997)
Received: 17 May 2021 Accepted: 11 August 2021
Timing of the Ethiopian flood basalt event and implications for plume
birth and global change. Nature 389(6653):838–841
Hyndman R (1979) Poisson’s ratio in the oceanic crust—a review. Tectonophys-
ics 59(1–4):321–333
Illsley-Kemp F, Keir D, Bull JM, Gernon TM, Ebinger C, Ayele A, Hammond JO,
References Kendall JM, Goitom B, Belachew M (2018) Seismicity during continental
Abdallah A, Courtillot V, Kasser M, Le Dain AY, Lépine JC, Robineau B, Ruegg breakup in the Red Sea Rift of northern Afar. J Geophys Res Solid Earth
JC, Tapponnier P, Tarantola A (1979) Relevance of Afar seismicity and 123(3):2345–2362
volcanism to the mechanics of accreting Plate boundaries. Nature IRIS D (2012) Data services products: SWS-DBS shear-wave splitting databases
282(5734):17–23 Joffe S, Garfunkel Z (1987) Plate kinematics of the circum Red Sea-a re-
Alothman A, Fernandes R, Bos M, Schillak S, Elsaka B (2016) Angular velocity evaluation. Tectonophysics 141(1–3):5–22
of Arabian Plate from multi-year analysis of GNSS data. Arab J Geosci Kandilarov A, Mjelde R, Flueh E, Pedersen RB (2015) Vp/Vs-ratios and anisot-
9(8):529 ropy on the northern Jan Mayen ridge, north Atlantic, determined from
Ammon CJ, Randall GE, Zandt G (1990) On the nonuniqueness of Receiver Ocean Bottom Seismic data. Polar Sci 9(3):293–310
Function inversions. J Geophys Res Solid Earth 95(B10):15,303-15,318 Keir D, Belachew M, Ebinger C, Kendall JM, Hammond JO, Stuart G, Ayele A,
Ando M, Ishikawa Y, Yamazaki F (1983) Shear wave polarization anisotropy in Rowland J (2011) Mapping the evolving strain field during continental
the upper mantle beneath Honshu, Japan. J Geophys Res Solid Earth breakup from crustal anisotropy in the Afar depression. Nat Commun
88(B7):5850–5864 2(1):1–7
Barruol G, Ismail WB (2001) Upper mantle anisotropy beneath the African iris Kennett BL, Engdahl E, Buland R (1995) Constraints on seismic velocities in the
and geoscope stations. Geophys J Int 146(2):549–561 Earth from traveltimes. Geophys J Int 122(1):108–124
Kumar et al. Earth, Planets and Space (2021) 73:166 Page 17 of 17

Kind R, Yuan X, Mechie J, Sodoudi F (2015) Structure of the upper mantle in Reilinger R, McClusky S (2011) Nubia-Arabia-Eurasia plate motions and the
the north-western and central united states from us arrays-receiver func- dynamics of mediterranean and middle east tectonics. Geophys J Int
tions. Solid Earth 6(3):957–970. https://​doi.​org/​10.​5194/​se-6-​957-​2015 186(3):971–979
Kodaira S, Bellenberg M, Iwasaki T, Kanazawa T, Hirschleber HB, Shimamura H Rychert CA, Hammond JO, Harmon N, Kendall JM, Keir D, Ebinger C, Bastow ID,
(1996) Vp/ Vs ratio structure of the Lofoten continental margin, northern Ayele A, Belachew M, Stuart G (2012) Volcanism in the Afar rift sustained
Norway, and its geological implications. Geophys J Int 124(3):724–740. by decompression melting with minimal plume influence. Nat Geosci
https://​doi.​org/​10.​1111/j.​1365-​246X.​1996.​tb056​34.x 5(6):406–409
Koptev A, Gerya T, Calais E, Leroy S, Burov E (2018) Afar triple junction triggered Savage MK (1999) Seismic anisotropy and mantle deformation: what have we
by plume-assisted bi-directional continental break-up. Sci Rep 8(1):1–7 learned from shear wave splitting? Rev Geophys 37(1):65–106. https://​
Korostelev F, Weemstra C, Leroy S, Boschi L, Keir D, Ren Y, Molinari I, Ahmed A, doi.​org/​10.​1029/​98RG0​2075
Stuart GW, Rolandone F et al (2015) Magmatism on rift flanks: insights Schilling JG, Kingsley RH, Hanan BB, McCully BL (1992) Nd-Sr-Pb isotopic varia-
from ambient noise phase velocity in Afar region. Geophys Res Lett tions along the gulf of aden: Evidence for Afar mantle plume-continental
42(7):2179–2188 lithosphere interaction. J Geophys Res Solid Earth 97(B7):10,927-10,966
Koulakov IY (2007) Structure of the Afar and Tanzania plumes based on the Sebai A, Stutzmann E, Montagner JP, Sicilia D, Beucler E (2006) Anisotropic
regional tomography using ISC data. In: Doklady earth sciences, Citeseer, structure of the African upper mantle from Rayleigh and love wave
vol 417, p 1287 tomography. Phys Earth Planet Inter 155(1–2):48–62
Kumar U, Legendre CP (2021) STADIUM-Py: Python command-line interface for Sicilia D, Montagner JP, Cara M, Stutzmann E, Debayle E, Lépine JC, Lévêque
automated receiver functions and shear-wave splitting measurements. JJ, Beucler E, Sebai A, Roult G et al (2008) Upper mantle structure of
Zenodo. https://​doi.​org/​10.​5281/​zenodo.​46861​03 shear-waves velocities and stratification of anisotropy in the Afar hotspot
Kusky TM, Toraman E, Raharimahefa T, Rasoazanamparany C (2010) Active region. Tectonophysics 462(1–4):164–177
tectonics of the Alaotra-Ankay graben system, Madagascar: possible Silver PG (1996) Seismic anisotropy beneath the continents: probing the
extension of Somalian-African diffusive plate boundary? Gondwana Res depths of geology. Annu Rev Earth Planet Sci 24(1):385–432
18(2–3):274–294 Simmons NA, Forte AM, Grand SP (2007) Thermochemical structure and
Langston CA (1979) Structure under Mount Rainier, Washington, inferred from dynamics of the African Superplume. Geophys Res Lett 34(2):L02301
teleseismic body waves. J Geophys Res Solid Earth 84(B9):4749–4762 Tesauro M, Kaban MK, Petrunin AG, El Khrepy S, Al-Arifi N (2018) Strength and
Legendre C (2013) Shear wave velocity model of the upper mantle beneath elastic thickness variations in the Arabian plate: a combination of tem-
Europe and surroundings. Doctoral Thesis, Ruhr-Universität Bochum, perature, composition and strain rates of the lithosphere. Tectonophysics
Universitätsbibliothek 746:398–411
Legendre C, Zhao L, Deschamps F, Chen QF (2016) Layered anisotropy within Tesfaye S, Harding DJ, Kusky TM (2003) Early continental breakup bound-
the crust and lithospheric mantle beneath the Sea of Japan. J Asian Earth ary and migration of the Afar triple junction, Ethiopia. GSA Bull
Sci 128:181–195 115(9):1053–1067
Legendre C, Tseng T, Zhao L (2020) Surface-wave phase-velocity maps of the Thompson D, Hammond JO, Kendall JM, Stuart G, Helffrich G, Keir D, Ayele
Anatolia region (Turkey) from ambient noise tomography. J Asian Earth A, Goitom B (2015) Hydrous upwelling across the Mantle Transition
Sci 193:104322 Zone beneath the Afar triple junction. Geochem Geophys Geosyst
Legendre CP, Tseng TL, Chen YN, Huang TY, Gung YC, Karakhanyan A, Huang BS 16(3):834–846
(2017) Complex deformation in the Caucasus region revealed by ambient Trabant C, Hutko AR, Bahavar M, Karstens R, Ahern T, Aster R (2012) Data prod-
noise seismic tomography. Tectonophysics 712:208–220 ucts at the IRIS DMC: stepping stones for research and other applications.
Maguire R, Ritsema J, Goes S (2018) Evidence of subduction-related thermal Seismol Res Lett 83(5):846–854
and compositional heterogeneity below the United States from transi- Venzke E (2013) Global volcanism program, 2013. Volcanoes of the world, v
tion zone receiver functions. Geophys Res Lett 45(17):8913–8922 4.5. 2
McClusky S, Reilinger R, Ogubazghi G, Amleson A, Healeb B, Vernant P, Sholan Walker KT, Bokelmann GH, Klemperer SL, Nyblade A, Walker K, Bokelmann G,
J, Fisseha S, Asfaw L, Bendick R et al (2010) Kinematics of the southern Klemperer S, Nyblade A (2005) Shear wave splitting around hotspots: evi-
Red Sea-Afar triple junction and implications for Plate dynamics. Geophys dence for upwelling-related mantle flow? Spec Pap Geol Soc Am 388:171
Res Lett 37(5):L05301 Walpersdorf A, Vigny C, Ruegg JC, Huchon P, Asfaw LM, Kirbash SA (1999)
McQuarrie N, Stock J, Verdel C, Wernicke B (2003) Cenozoic evolution of 5 years of GPS observations of the Afar triple junction area. J Geodyn
Neotethys and implications for the causes of plate motions. Geophys Res 28(2):225–236
Lett 30(20):2036 Walpole J, Wookey J, Masters G, Kendall J (2014) A uniformly processed data
Mège D, Korme T (2004) Dyke swarm emplacement in the Ethiopian Large set of SKS shear wave splitting measurements: a global investigation of
Igneous Province: not only a matter of stress. J Volcanol Geotherm Res upper mantle anisotropy beneath seismic stations. Geochem Geophys
132(4):283–310 Geosyst 15(5):1991–2010
Megies T, Beyreuther M, Barsch R, Krischer L, Wassermann J (2011) Obspy-what Wignall PB (2001) Large igneous provinces and mass extinctions. Earth-Sci Rev
can it do for data centers and observatories? Ann Geophys 54(1):47–58 53(1–2):1–33
Mohr P (1970) The Afar triple junction and sea-floor spreading. J Geophys Res Wüstefeld A, Bokelmann G (2007) Null detection in shear-wave splitting meas-
75(35):7340–7352 urements. Bull Seismol Soc Am 97(4):1204–1211
Musacchio G, Mooney WD, Luetgert JH, Christensen NI (1997) Composi- Wüstefeld A, Bokelmann G, Barruol G, Montagner JP (2009) Identifying global
tion of the crust in the Grenville and Appalachian provinces of seismic anisotropy patterns by correlating shear-wave splitting and
North America inferred from Vp/Vs ratios. J Geophys Res Solid Earth surface-wave data. Phys Earth Planet Inter 176(3–4):198–212
102(B7):15225–15241 Zandt G, Ammon CJ (1995) Continental crust composition constrained by
Ni S, Tan E, Gurnis M, Helmberger D (2002) Sharp sides to the African Super- measurements of crustal Poisson’s ratio. Nature 374(6518):152–154
plume. Science 296(5574):1850–1852 Zhu L, Kanamori H (2000) Moho depth variation in southern Califor-
Obrebski M, Kiselev S, Vinnik L, Montagner JP (2010) Anisotropic stratification nia from teleseismic receiver functions. J Geophys Res Solid Earth
beneath Africa from joint inversion of SKS and P receiver functions. J 105(B2):2969–2980
Geophys Res Solid Earth 115(B9):B09313
Reed CA, Almadani S, Gao SS, Elsheikh AA, Cherie S, Abdelsalam MG, Thur-
mond AK, Liu KH (2014) Receiver function constraints on crustal seismic Publisher’s Note
velocities and partial melting beneath the Red Sea rift and adjacent Springer Nature remains neutral with regard to jurisdictional claims in pub-
regions, Afar depression. J Geophys Res Solid Earth 119(3):2138–2152 lished maps and institutional affiliations.