The geology of New Zealand has been shaped by tectonic plate interactions driven by mantle convection over the past 60 million years, but the effects of these interactions on the transition to the lower mantle are not yet well understood. We analyze 10 years of teleseismic -wave receiver functions using common conversion point stacking to investigate mantle transition zone discontinuities. The resulting topography of the 410-km and 660-km discontinuities, smoothed using kriging interpolation, reveals localized thermal anomalies. We observe more than 10 km of thinning in the mantle transition zone beneath central North Island coincident with the Hikurangi slab, accompanied by a significantly thickened and hydrous melt-rich layer above the 410-km discontinuity. In northern South Island, the 410-km discontinuity is uplifted by approximately 15 km, likely reflecting subducted materials from the oblique Hikurangi slab material reaching mantle transition zone depths. These observations are consistent with patterns of seismicity and seismic velocity anomalies from tomographic models. Additionally, we identify significant thinning of the mantle transition zone beneath both the Northland region (10 km) and the Great South Basin (15 km), indicating the influence of potential local thermal upwellings. The different slab polarities and transpressional boundary may drive lateral mantle flow, shaping the complex transition zone topography beneath New Zealand.

Divergent plate boundaries drive magmatism and plate spreading, yet their role in whole-mantle convection remains unclear, as most systems are submarine. When spreading centres subduct, slab windows open pathways for asthenospheric and potentially deeper mantle upwelling. Here we use 20 years of teleseismic P-to-S receiver functions from 429 stations to map the 410 and 660 km discontinuities beneath western Canada. Common-conversion-point stacking reveals coherent double depressions across a slab window formed by the subducted Resurrection/Farallon Ridge system, in contrast to the adjacent craton. Mineral physics and seismic tomography constraints indicate mantle transition zone temperature anomalies of 100-300 K that follow the slab-window geometry. Monte Carlo estimates of constrained whole-mantle geotherms further suggest a warmer-than-average lower mantle. Together, these observations suggest that the fossil slab window provides a conduit for lower-mantle heat into the upper mantle and lithosphere, contributing to long-term uplift of the Canadian Cordillera.

Uncertainty in key input parameters, particularly rupture initiation depth (focal depth) and average rupture propagation velocity, can significantly degrade the reliability of finite-fault earthquake source modeling results. These parameters are often estimated empirically or determined through trial-and-error, leading to solutions with large uncertainties and inefficient computation. To address this problem, we develop a parallel grid-search procedure that systematically constrains initial rupture depth and rupture velocity prior to performing a final source rupture modeling. The method evaluates a two-dimensional grid of depth–velocity models by assigning subsets of the grid in parallel to multiple CPU threads. Compared to conventional sequential grid search, the computation time is dramatically reduced. Misfit variance contours are then used to identify the optimal parameter combinations, providing reasonable initial conditions for subsequent rupture inversion. The procedure is validated using three large earthquakes with diverse tectonic environments, demonstrating its efficiency, robustness, and applicability to modern large-scale seismic modeling. This approach offers a practical and scalable solution for improving the accuracy of finite-fault earthquake source modeling.

The absence of plate tectonics and the young surface age (0.3-1 billion years) of Venus have led to diverse geodynamic models for Venus. The energetics of the Venusian interior drives these models; however, the lack of direct constraints on surface heat flow hampers their quantitative assessment. Here we present a global heat flow map for Venus, as well as estimates of the total heat loss, obtained from an inversion of geophysical data, including lithospheric effective elastic thickness, crustal thickness, and radioactive heat production. Heat flow on Venus is lower and less geographically structured than on Earth, with an average of 31 mW m-2, but with highs associated to rifts systems reaching values typical of active terrestrial areas. The obtained total heat loss is 11-17 TW, similar to estimates of the total radioactive heat production. Therefore, at present, Venus proportionally dissipates much less heat than Earth.

The composition and structure of the mantle wedge corner (MWC) exert a primary control on seismogenic and fluid processes in subduction zone forearcs. Seismic evidence suggests widespread serpentinization of the MWC in Cascadia, but constraints on its internal structure and fabrics remain sparse. Here, we compile receiver function data for a line of seismograph stations in northern Cascadia to characterize the layered structure and seismic anisotropy in the MWC. We first invert for flat-layer shear-wave velocity (Vs) models and identify the presence of a dipping, similar to 4 km thick seismic low-velocity layer (LVL) marked by low Vs (similar to 3 km/s) and high compressional-to-shear velocity ratio (Vp/Vs > 2). Above the LVL, the continental Moho is taken as the base of the low Vp/Vs (similar to 1.7) zone, previously interpreted as the silica-rich lower continental crust. Second, we solve for anisotropic parameters in the MWC, including layer thickness, Vp/Vs, percent anisotropy, and symmetry-axis orientation. We infer an similar to 5 km thick layer at the base of the MWC with Vs of 3.6 km/s and Vp/Vs of similar to 1.9. Anisotropy is characterized by a slow axis of hexagonal symmetry with 10% anisotropy and an orientation corresponding with sub-horizontal foliation fabrics. Based on these results, we infer 40%-100% serpentinization of the MWC localized within a thin layer above the plate interface shear zone, potentially hosting high pore-fluid pressure. Shear-induced foliated fabrics in serpentinites above the plate interface may play an essential role in controlling seismogenic processes by channelling fluids up dip and aiding in generating fluid overpressures.

Receiver functions are a powerful tool to image lithospheric stratigraphy. For flat lying structures, receiver functions can be stacked azimuthally to achieve high signal-to-noise ratios and h-κ-stacks allow to estimate the depth of interfaces (h) and P-to-S wave velocity ratio of the hanging layers (κ). For dipping layers, characteristic for the slab structure in a subduction zone forearc, these methods fail, because the moveout of phases arriving from different azimuths violates the basic assumptions of these methods. We here present a simple routine to simultaneously search for the depth of the top of slab and of the oceanic Moho, for strike and dip of the downgoing slab, as well as for the S-wave velocities and the P‑to-S wave velocity ratios of multiple layers of the overriding and downgoing plates in subduction zone forearcs. Our approach is based on the recent Python port PyRaysum of Frederiksen and Bostock's classic (2000) code for modeling ray-theoretical plane body-wave propagation in dipping anisotropic media, and on SciPy's simulated annealing global parameter search. We applied the routine to hundreds of azimuthally-dependent receiver function sections from the subduction zones of Cascadia (North America) and the central Andes (South America) and retrieved laterally coherent station measurements of the depth and orientation of the top of the subducting slab and the subducting Moho, with only weakly constrained seismic velocities. In Cascadia, we interpolated a regional slab model through fitting of regularized spline surfaces. Small scale structures that are not present in previous slab models can be resolved, e.g. under Olympic Peninsula (Cascadia) and Mejillones Peninsula (northern Chile). Where the receiver functions are more complex than can be accounted for by our model, the labeling of the modeled receiver function phases and comparison to the observed receiver functions allows us to confidently interpret the additional subsurface complexities and reconcile them with our interpretations.

Earthquakes abruptly release tectonic stress that builds slowly over time through the coupled evolution of faults and the surrounding crust. Seismic wavespeeds track crustal deformation and stress changes, but typical monitoring methods are most sensitive to shallow depths. Using receiver functions, we tracked rupture-zone wavespeed and anisotropy changes throughout the crust during the 2019 Ridgecrest earthquake sequence. Shallow coseismic wavespeed reductions recovered within months, whereas a deeper postseismic wavespeed drop persisted without measurable recovery over several years. The deep, persistent wavespeed drop likely reflects accumulating damage driven by postseismic deformation, suggesting two possible scenarios: (i) a slow interseismic recovery where wavespeed and anisotropy track long-term stress evolution; or (ii) permanent deformation of an immature fault zone. Both scenarios affect the dynamics and energy budget of the seismic cycle.

The Canadian Cordillera is characterized by higher mean elevation and thinner crust compared to the adjacent craton to the east, in apparent contradiction with the Airy isostatic model. Roy Hyndman hypothesized that the high Cordillera elevations may be supported by an overall hot, low-density Cordilleran backarc mantle, but this idea hasn't been tested at the regional scale. We use a new high-resolution crustal thickness model for western Canada derived from teleseismic receiver functions and active source data, to conduct a detailed analysis of elevation-crustal thickness trends in four distinct regions: forearc, backarc, foreland, and craton. The backarc and foreland regions have higher elevations relative to the forearc and craton, requiring additional buoyancy beneath these areas. Using a grid search, we find that unrealistically low crustal densities are required to explain the high elevations and, therefore, the support must come from below the crust. To isolate the mantle component, we use a global crustal density model to correct for the effects of crustal density variations. After correction, the elevation-crustal thickness trends in each area align with the Airy hypothesis. For a given crustal thickness, the elevation of the backarc region is ~700 m higher than that of the craton. This is consistent with uplift caused by a hot backarc mantle. However, the difference is lower than that determined for the entire North American Cordillera (~1600 m). This may suggest the mantle under the Canadian Cordillera is slightly cooler than regions to the adjacent US Cordillera, perhaps due to the absence of ongoing subduction.

Abstract Slow slip events (SSEs) at subduction zones relieve tectonic stress over periods ranging from days to years. A pervasive feature in SSE regions worldwide is the presence of high fluid pressures in the subducting slab, which are thought to fluctuate during SSEs through the fault-valving action of an impermeable layer at the plate interface. Yet, geophysical observations of this fault-valving behaviour remain sparse. Here we show, by applying timelapse receiver function analysis at the Manawatu deep slow slip region, that VP/VS in the top ~7.3 km of the subducting Pacific Plate increases up to 2.40 during inter-SSE periods and decreases down to 2.20 during SSE periods, indicating sustained high fluid pressures across slip cycles. In a ~2 km thick layer above, VP/VS fluctuates in the opposite sense, decreasing to ~1.4 during inter-SSE to >2.0 during SSEs. We interpret these results to indicate a reduction in fluid saturation due to decreased crack density and mineral precipitation between SSE cycles, followed by increased crack density and fluid saturation induced by slip on the subduction thrust. Our work shows that this thin layer relieves high fluid pressures in the subducting slab by acting as a valve.

AbstractSlow-slip events at global subduction zones relieve tectonic stress over days to years. Through slow-slip cycles, high fluid pressures observed at the top of subducting plates are thought to fluctuate, potentially due to the valving action of an impermeable layer near the plate interface. We model teleseismic scattering data at the Manawatu deep slow-slip patch at the Hikurangi margin in New Zealand and find high seismic P-to-S wave velocity ratios, VP/VS, in the upper ~5 km of the subducting Pacific Plate, reflecting sustained elevated fluid pressures that decrease during slow-slip and increase during inter-slow-slip periods. Within a ~ 3 km thick lower crustal layer of the overriding Australian Plate, decreasing VP/VS during inter-slow-slip periods reflects permeability reduction due to mineral precipitation. Increasing VP/VS during slow-slip reflects increasing permeability and crack density, facilitating upward fluid transfer through this layer. Our results suggest it acts as a valve to relieve high fluid pressures in the subducting slab.

SUMMARY Seismic velocity models provide important constraints on Greenland’s deep structure, which, in turn, has profound implications for our understanding of the tectonic history of this region. However, the resolution of seismic models has been limited by a sparse network, particularly in northern and central Greenland. We address these limitations by generating new high-resolution Rayleigh-wave phase velocity maps encompassing Greenland and northeastern Canada by processing over three decades of teleseismic earthquake records and incorporating recently added stations in Greenland and Arctic Canada. These phase velocity maps are sensitive to structure from the lower crust down to the sub-lithospheric mantle (25–185 s period). We find significant heterogeneity and a strong correlation between isotropic and anisotropic seismic velocities with inferred geological structure. High seismic velocities associated with cratonic lithosphere are broadly divided into two regions, with a belt of reduced velocity spanning central Greenland, which we interpret as lithospheric erosion resulting from interaction between the Greenland continental keel and the Iceland plume. Within each region, we identify tectonic subdivisions that suggest fundamental differences between the blocks that make up Precambrian Greenland. In the south, the North Atlantic craton (NAC) has a high-velocity keel exhibiting anisotropic stratification. Between the NAC and the cratonic lithosphere further north, the Proterozoic Nagssugtoqidian orogenic belt shows a distinct signature of reduced seismic velocity to $\sim$75 s period, but then appears to pinch out at depth. The northern Greenland lithosphere exhibits significant isotropic heterogeneity, with a distinct core of high velocities in the northwest ($\sim$55–75 s period) giving way to a set of distinct east-west trending high-velocity belts at longer periods. At all periods sensitive to the lithospheric mantle in this region, anisotropic fast orientations are E–W, consistent with a north–south Precambrian assembly of the Greenland shield. In contrast to the NAC, there is no evidence of anisotropic stratification in the northern part of the cratonic keel. Based on both isotropic and anisotropic phase-velocity anomalies, we suggest that the Phanerozoic Caledonian and Ellesmerian-Franklinian fold belts are relatively thin-skinned features onshore Greenland, though the Caledonian belt may have a stronger signature off the east coast. At the longest periods, a prominent low-velocity anomaly initially centred on Iceland migrates northwards and spreads beneath central-eastern Greenland. Coupled with NW–SE trending anisotropy, this feature is interpreted as the effect of mantle flow radiating outward from the Iceland plume and interacting with the eroded Greenland lithosphere.

Geologically, the Arctic is one of the least-explored regions of Earth. Obtaining data in the high Arctic is logistically, economically, and environmentally expensive, but the township of Longyearbyen (population of 2617 as of 2024) at 78 degrees N represents a relatively easily accessible gateway to Arctic geology and is home to The University Centre in Svalbard (UNIS). These unique factors provide a foundation from which to teach and explore Arctic geology via the classroom, the laboratory, and the field. UNIS was founded in 1993 as the Norwegian "field university", offering field-based courses in Arctic geology, geophysics, biology, and technology to students from Norway and abroad.In this contribution, we present one of the educational components of the international collaboration project NOR-R-AM (a Norwegian-Russian-North American collaboration in Arctic research and collaboration, titled Changes at the Top of the World through Volcanism and Plate Tectonics) which ran from 2017 to 2024. One of the key deliverables of NOR-R-AM was a new graduate (Master's and PhD-level) course called Arctic Tectonics and Volcanism that we have established and taught annually at UNIS since 2018 and detail herein. The course's main objective is to teach the complex geological evolution of the Arctic from the Devonian period (similar to 420 million years ago, Ma) to the present day through integrating multi-scale datasets and a broad range of geoscientific disciplines. We outline the course itself before presenting student perspectives based on both an anonymous questionnaire (n=27) and in-depth perceptions of four selected students. The course, with an annual intake of up to 20 MSc and PhD students, is held over a 6-week period, typically in spring or autumn. The course comprises modules on field and polar safety, Svalbard/Barents Sea geology, wider Arctic geology, plate tectonics, mantle dynamics, geo- and thermochronology, and geochemistry of igneous systems. A field component, which in some years included an overnight expedition, provides an opportunity to appreciate Arctic geology and gather field observations and data. Digital outcrop models, photospheres, and tectonic plate reconstructions provide complementary state-of-the-art data visualization tools in the classroom and facilitate efficient fieldwork through pre-fieldwork preparation and post-fieldwork quantitative analyses. The course assessment is centred around an individual research project that is presented orally and in a short and impactful Geology journal-style article. Considering the complex subject and the diversity of students' backgrounds and level of geological knowledge before the course, the student experiences during this course demonstrate that the multi-disciplinary, multi-lecturer field-and-classroom teaching is efficient and increases their motivation to explore Arctic science.

The Canadian Cordillera marks a transition region from the current plate boundary through the Phanerozoic Cordilleran orogen to the Precambrian cratons. Knowledge of the subsurface structure of western Canada has been greatly advanced by seismological investigations during the past two decades, pioneered by the Lithoprobe project and, more recently, by regional passive seismic arrays. In this study, we construct a new model (WCANM22) of crustal thickness and P- to S-wave velocity ratio, or V-p/V-s, by compiling receiver function data from 473 stations and existing constraints from over 2,600-km long active source experiments. Our model covers a broad swath (about 1/4) of the land area of North America (105 degrees-140 degrees W, 48 degrees-72 degrees N) and shows an overall flat Moho beneath the Cordillera with an average depth of similar to 36 km and a standard deviation of 3 km across orogenic belts. This study provides a comprehensive catalog of V-p/V-s in western Canada and reveals a moderate correlation between Poisson's ratio and the age of crustal domains. The average V-p/V-s values are 1.72, 1.79, and 1.82 for the Phanerozoic Cordillera, Proterozoic Cratons, and Archean-aged Medicine Hat Block, respectively, suggesting continued modifications to crustal composition through episodic tectonothermal events. This distinct trend in western Canada sheds new light on the debated role of secular changes in the composition of continental crust. Plain Language Summary The crust of western Canada consists of complex tectonic domains that collided during a protracted geological history dating back to 4 billion years ago. To uncover crustal structures and the associated tectonic processes, we systematically determine the depth to the base of the crust (Moho depth) and compressional-to-shear seismic wave speed ratio (V-p/V-s ) beneath 473 seismic stations integrated with the results from existing active source experiments. Along a northwest-southeast profile, V-p/V-s ratios exhibit a gradually increasing trend with the age of the crust. This variation in V-p/V-s may be caused by modifications to the crustal composition during the geological evolution of western Canada.

Earth's tectonic history is punctuated by several cycles of supercontinent assembly and breakup that profoundly influenced the lithospheric structure; however, the roles of the various factors controlling continental strength and deformation during the cycles remain debated. The effective elastic thickness (Te) reflects the lithosphere's long-term, depth-integrated strength and is useful for deciphering the complex evolution of continents. In this study, we estimate a new global map of continental Te projected onto a 15′×15′ grid by inverting the cross-spectral properties (admittance and coherence) between Bouguer gravity and topography data obtained from a continuous wavelet transform. Continental Te ranges from <5 to ∼140 km, with a mean and standard deviation of 50 and 33 km, respectively. Based on a gaussian mixture model-based cluster analysis, we delineate tectonically active provinces, stable Archean cratons and transitional lithosphere. We find an obvious positive correlation between Te and lithospheric thickness obtained from calibrated upper mantle surface wave tomography models. Further comparing the Te distribution with orogenic age data shows that Te exhibits a clear time dependence where the strength is governed by the time since the last orogeny. Based on plate cooling models, we indicate that continental Te corresponds approximately to the depth of the 300±150∘C isotherm. These results favour a diffusive (cooling) model that considerably influences the strength of the continental lithosphere, despite the complex relation between Te and the thermal, compositional and rheological structure.

The recent deployment of temporary broadband seismic networks, notably the EarthScope USArray-Transportable Array (TA), has drastically improved the station coverage across northwestern Canada over the last ten years, enabling application of high-resolution passive-source seismic methods (i.e., seismic tomography, receiver functions and core phase shear wave splitting). This review highlights the main discoveries pertaining to the seismic velocity structure, origin and deformation of the lithosphere in the northern Canadian Cordillera (NCC). High-resolution seismic tomography models reveal that the lower crust in the NCC is marked by low velocity anomalies extending from the Gulf of Alaska to the Cordilleran deformation front, which are interpreted to reflect elevated temperatures that buoyantly support regional high elevations and potentially represent the seismic signature of strain transfer from the Yakutat collision zone to the Mackenzie Mountains. The Moho is relatively flat and shallow across the NCC, and is underlain by a thin layer of mantle lithosphere. Seismic velocity models further unveiled large-scale mantle structures associated with the unexposed Mackenzie craton in the north, and the Liard Transfer Zone in the south, which appear to buttress the NCC and further focus deformation in the eastern NCC. Seismic anisotropy and tomography provide evidence that the Tintina and Denali faults penetrate into the lithospheric mantle and played a first order role in shaping the present-day NCC. We propose that future studies should aim to: 1) resolve the shape of the Cordillera-craton boundary at upper mantle depths; 2) accurately estimate the lithosphere thickness in the NCC; and 3) improve coverage in the Beaufort Sea to understand the controls on convergent tectonics in the northern NCC.

Deep-learning (DL) algorithms are increasingly used for routine seismic data processing tasks, including seismic event detection and phase arrival picking. Despite many examples of the remarkable performance of existing (i.e., pre-trained) deep-learning detector/picker models, there are still some cases where the direct applications of such models do not generalize well. In such cases, substantial effort is required to improve the performance by either developing a new model or fine-tuning an existing one. To address this challenge, we present Blockly Earthquake Transformer(BET), a deep-learning platform for efficient customization of deep-learning phase pickers. BET implements Earthquake Transformer as its baseline model, and offers transfer learning and fine-tuning extensions. BET provides an interactive dashboard to customize a model based on a particular dataset. Once the parameters are specified, BET executes the corresponding phase-picking task without direct user interaction with the base code. Within the transfer-learning module, BET extends the application of a deep-learning P and S phase picker to more specific phases (e.g., Pn, Pg, Sn and Sg phases). In the fine-tuning module, the model performance is enhanced by customizing the model architecture. This no-code platform is designed to quickly deploy reusable workflows, build customized models, visualize training processes, and produce publishable figures in a lightweight, interactive, and open-source Python toolbox.

&lt;p&gt;Global-scale seismic velocity models of the Northern Canadian Cordillera show high velocities to the east of the Cordilleran deformation front and low velocities to the west.&amp;#160;&amp;#160;&amp;#160; This velocity contrast is consistent with other geophysical observables, such as regional seismological studies, that indicate a weak and thin lithosphere to the west that transitions quickly to a strong and thick craton-like lithosphere at the deformation front.&amp;#160;&amp;#160;&amp;#160; We present new results using data collected by the Mackenzie Mountains EarthScope Project, which included an ~875 km-long line of 40 broadband seismographs across the Cordillera and into the craton extending from roughly Skagway, Alaska to Great Bear Lake, Northwest Territories.&amp;#160;&amp;#160;&amp;#160; The 3-year overlap of this deployment with other broadband seismic stations in the region, most notably the EarthScope Transportable Array and the Yukon Northwest Seismic Network, allows for detailed 3-D Rayleigh wave ambient noise imaging of the upper lithosphere.&amp;#160;&amp;#160;&amp;#160; Results show large velocity variations west of the deformation front.&amp;#160; &amp;#160;Notably, we image a 5% Vs low that extends from the upper crust to the asthenospheric mantle.&amp;#160;&amp;#160; This plume-like structure, and associated weakening, may be a primary cause for the ongoing uplift of the Mackenzie Mountains at their unusually eastward location.&amp;#160;&amp;#160; We also image a low velocity feature in the lower crust extending to the west of the deformation front, which may facilitate eastward crustal translation along a large-scale (~800 km) decollement system driven by the Yakutat indentor consistent with the orogenic float hypothesis of Mazzotti and Hyndman (2002).&amp;#160;&amp;#160;&amp;#160; We also note strong lithosphere-scale lateral heterogeneity suggesting that 3-D effects are important in focusing deformation in the Mackenzie Mountain area.&lt;/p&gt;

Our understanding of the present‐day state and evolution of the Canadian and Alaskan mantle is hindered by a lack of absolute P‐wavespeed constraints that provide complementary sensitivity to composition in conjunction with existing S‐wavespeed models. Consequently, cratonic modification, orogenic history of western North America and complexities within the Alaskan Proto‐Pacific subduction system remain enigmatic. One challenge concerns the difficulties in extracting absolute arrival‐time measurements from often‐noisy data recorded by temporary seismograph networks required to fill gaps in continental and global databases. Using the Absolute Arrival‐time Recovery Method (AARM), we extract >180,000 new absolute arrival‐time residuals from seismograph stations across Canada and Alaska and combine these data with USArray and global arrival‐time data from the contiguous US and Alaska. We develop a new absolute P‐wavespeed tomographic model, CAP22, spanning North America that significantly improves resolution in Canada and Alaska over previous models. Slow wavespeeds below the Canadian Cordillera sharply abut fast wavespeeds of the continental interior at the Rocky Mountain Trench in southwest Canada. Slow wavespeeds below the Mackenzie Mountains continue farther inland in northwest Canada, indicating Proterozoic‐Archean metasomatism of the Slave craton. Inherited tectonic lineaments colocated with this north‐south wavespeed boundary suggest that both the crust and mantle may control Cordilleran orogenic processes. In Alaska, fast upper mantle wavespeeds below the Wrangell Volcanic Field favor a conventional subduction related mechanism for volcanism. Finally, seismic evidence for the subducted Kula and Yukon slabs indicate tectonic reconstructions of western North America may require revision.

Deep-learning (DL) algorithms are increasingly used for routine seismic data processing tasks, including seismic event detection and phase arrival picking. Despite many examples of the remarkable performance of existing (i.e., pre-trained) deep-learning detector/picker models, there are still some cases where the direct applications of such models do not generalize well. In such cases, substantial effort is required to improve the performance by either developing a new model or fine-tuning an existing one. To address this challenge, we present Blockly Earthquake Transformer(BET), a deep-learning platform for efficient customization of deep-learning phase pickers. BET implements Earthquake Transformer as its baseline model, and offers transfer learning and fine-tuning extensions. BET provides an interactive dashboard to customize a model based on a particular dataset. Once the parameters are specified, BET executes the corresponding phase-picking task without direct user interaction with the base code. Within the transfer-learning module, BET extends the application of a deep-learning P and S phase picker to more specific phases (e.g., Pn, Pg, Sn and Sg phases). In the fine-tuning module, the model performance is enhanced by customizing the model architecture. This no-code platform is designed to quickly deploy reusable workflows, build customized models, visualize training processes, and produce publishable figures in a lightweight, interactive, and open-source Python toolbox.