Nature Geoscience spoke with Samantha Hansen, a geophysicist at the University of Alabama and Sebastian Rost, a global seismologist at the University of Leeds about the ultralow velocity zones in the lowermost mantle.
How would you describe the concept of ultralow velocity zones to a non-specialist and how does an ultralow velocity zone differ from a low velocity zone?
Samantha Hansen: Ultralow velocity zones (ULVZs) are unusual spots deep within the Earth that sit just above our planet’s core–mantle boundary at approximately 2900 km depth. These zones are generally tens of kilometers thick, with a potentially larger lateral extent. When seismic waves generated by earthquakes pass through ULVZs, their velocities are dramatically reduced, and this is where the name comes from. Compressional (P) waves may move up to 20% slower in ULVZs compared to their speed in the surrounding mantle, and shear (S) waves slow down even more, with velocity reductions up to 50%. Some studies suggest that ULVZs are up to 20% denser than the surrounding mantle. Various hypotheses have been proposed to explain the presence and structure of ULVZs, yet their origins remain a topic of ongoing debate.
In contrast, low velocity zones also exhibit slower seismic wave propagation, albeit with less pronounced reduction in speed (approximately 2–6%). The term ‘low velocity zone’ typically refers to the portion of the Earth where the lithosphere (that is, rigid tectonic plates) interfaces with the underlying asthenosphere (the weak, ductile part of the upper mantle). This transition occurs at depths ranging from about 40 to 250 kilometers, depending on whether the low velocity zone is situated beneath the oceanic or the continental lithosphere. Consequently, low velocity zones are considerably shallower in comparison to ULVZs.
Sebastian Rost: Ultralow velocity zones (ULVZs) are small patches in the Earth’s lowermost mantle, with maximum lateral dimensions of a few hundred kilometers and a few low-tens of kilometers thick, characterized by extreme seismic velocity drops of up to 50%. Their nature, genesis, origin, and whether they are transient or enduring mantle features are unknown. They are imaged using seismic data and studied in seismology, high-temperature and high-pressure mineral physics and geodynamics.
Large Low Velocity Provinces (LLVPs) are two continent-sized regions with lowered seismic velocities at the equator beneath the Pacific and Africa. Velocity reductions of a few percent are characteristic of LLVPs, and their location and structure are impacted by large scale mantle convection processes such as oceanic plate subduction; however, their origin is unknown, and ULVZs might be related to LLVPs. Early studies indicated that ULVZs were found at the edges of LLVPs, presumably due to mantle flow outside of LLVPs, composition, internal convection, and higher temperatures. Recent studies suggest a wider distribution of ULVZs likely due to wider geographical sampling of the core–mantle boundary, but LLVPs might still be the host and/or origin of a specific type of ULVZ.
Why is there so much interest in ultralow velocity zones? How can they help us to understand the lower mantle?
SH: The internal structure of our planet is often presented as a simple ‘layer cake’, but I think our recognition of ULVZs and other deep Earth anomalies have made us question this approach. The absolute change in physical properties (for example, viscosity, temperature, density) across the core–mantle boundary is greater than that between solid rock and air, and the complicated nature of this portion of our planet is intriguing. It is fascinating to contemplate how the deep interior of our planet interacts with materials originating from its surface. For example, some studies (including work conducted by my colleagues and I) suggest a potential association between ULVZs and subducted oceanic crust that has descended to the core–mantle boundary. Variable accumulations of subducted materials, advected throughout the lowermost mantle, could account for the distribution and diverse characteristics of ULVZs. If this hypothesis holds true, it will emphasize that plate tectonics is a global process, extending beyond the shallow portion of our planet.
SR: ULVZs are enigmatic structures in the Earth’s deep mantle, challenging to image seismically, conduct mineral physics experiments at the relevant pressure and temperature conditions, and implement in numerical geodynamic studies due to their size and unknown composition. Their possible origins include remnants of a global magma ocean from the early history of our planet, piles of core–mantle reaction products, features related to melting or graveyards of subducted oceanic crust, and exotic materials. Understanding ULVZs can provide insight into Earth’s mantle structure, dynamics, evolution, and relative stationarity of hot spots. It can also help to understand chemical core–mantle interactions and potential material flux across the core–mantle boundary. If ULVZs are passive piles within the mantle convective system, their locations can map up- and down-stream mantle material locations. If they are linked to partial melting of subducted material, they can help in understanding oceanic material recycling and temperature changes in the core–mantle boundary’s thermal boundary layer. Improved seismic characterization of ULVZs, better understanding of the lowermost mantle composition and behaviour under core–mantle boundary conditions, and better simulation of the lowermost mantle’s dynamics, are crucial for understanding the internal processing of our planet.
What is the significance of seismic anisotropy in understanding ultralow velocity zones and mantle dynamics more generally?
SH: Seismic anisotropy describes how the velocity of seismic waves depends on their propagation direction. That is, seismic waves travelling in different directions can have different speeds. One possible cause of seismic anisotropy is the alignment of mineral grains due to mantle flow, which can help us determine the direction of mantle material movement. Observations of both seismic anisotropy and ULVZs are common in certain parts of the lowermost mantle, particularly near the edges of the Large Low Velocity Provinces (LLVPs; two large-scale features that have been routinely recognized by seismic studies and that are situated beneath southern Africa and the Pacific Ocean). A recent study by Wolf and Long (J. Wolf and M. D. Long, Geochem. Geophys. Geosyst. 24, e2022GC010853; 2023) argued that by combining observations of ULVZs and anisotropy, we can gain insights into how ULVZs might migrate along the core–mantle boundary, influenced by mantle convection. Some studies suggest that mantle flow converges at the edges of LLVPs, leading to the accumulation of ULVZs along LLVP boundaries.
SR: Seismic anisotropy, the dependence of seismic velocity on direction, is a useful tool for analysing mantle deformation. It is typically sampled using S-wave splitting; the polarization of S-waves into fast and slow directions with a related difference in travel time. Understanding anisotropy within ULVZs could help to determine deformation and flow within these features. However, due to their small size, ULVZs are difficult to study with available seismic probes for anisotropy. Seismically resolving ULVZs’ 3D shapes and internal structure, including anisotropy, could provide better constraints on their dynamics and stability. However, current seismic resolution is insufficient to resolve these. Denser sampling of the seismic wavefield using recent installations of large numbers of seismic stations, and advances in high-frequency seismic wavefield simulations have made better characterization of the 3D and internal structure of ULVZs possible, but our understanding of the 3D structure of ULVZs is still in its early stages.
What are the big unknowns about ultralow velocity zones? Are there regions of particular interest that remain underexplored?
SH: There are still many questions about ULVZs. How do they originate? What constitutes their internal structure? Are they hotter and/or compositionally different than the surrounding mantle? How extensive are they across the core–mantle boundary? Clearly, these questions are related to one another, and while we have learned a lot about ULVZs, we still have much to discover.
SR: Current studies face trade-offs between ULVZ parameters (thickness, velocity) and uncertainties in their properties. Resolving these uncertainties and better constraining P-wave and S-wave velocities and density is crucial for understanding ULVZ composition and distinguishing between different ULVZ families. With the wide range of ULVZ parameters that is currently reported, there is a possibility that what we describe as ULVZs is a range of different phenomena. The global details of ULVZ locations at the core–mantle boundary are crucial for linking ULVZs to larger scale mantle structures and dynamics. The detailed 3D structure of ULVZs and how we see it in seismic data with their limited resolution is also important. High-resolution seismic wave propagation simulations at high frequencies are needed for detailed waveform studies. Understanding the validity of current non-3D models in characterizing 3D ULVZ structure and identifying information that may be available in the wavefield can help to resolve the 3D structure with the existing seismic data.
What technical advances or methodological developments are needed to test hypotheses about the origin of the ultralow velocity zones?
SH: I believe that a multidisciplinary approach is crucial. From a seismic viewpoint, we need to be able to test a wide range of 3D models, with different ULVZ parameters and geometries. Although progress is being made on this front, it is worth noting that such modelling remains computationally expensive and time consuming. A recent paper by Jenkins et al. (J. Jenkins, S. Mousavi, Z. Li and S. Cottaar, Earth Planet. Sci. Lett. 563, 116885; 2021) summarized this nicely by saying “the non-linearity, trade-offs, and high computational costs of exploring high-frequency complex 3D ULVZ models means a full understanding of 3D effects is not currently possible.” Additionally, seismic observations with new geometries are important. Different seismic phases are used to investigate ULVZs and other deep Earth anomalies, and our ability to explore different parts of the core–mantle boundary depend on source-to-station ray paths.
SR: Understanding ULVZs requires progress in several disciplines. For seismic imaging to determine where ULVZs exist, better coverage of the lowermost mantle with ULVZ sensitive probes is necessary along with improved global coverage of seismic stations and international collaboration in data exchange. Improved vertical resolution is needed to identify ULVZs with lower thicknesses in previously sampled areas. Combined analysis of multiple seismic probes can explore trade-offs and provide better constraints on ULVZ properties. Imaging the 3D structure of ULVZs is crucial for understanding their origin, composition, and dynamics. To better resolve that, better data coverage with high resolution seismic probes together with improvements in simulating the seismic wavefield in 3D velocity structures at high frequencies are needed. Advances in wave propagation techniques, machine learning, and artificial intelligence can help to predict seismic waveforms for different ULVZ parameters. As ULVZ study is a multi-disciplinary endeavour, improvements in understanding the properties of mantle materials at high temperature and pressure are also needed. Geodynamic models can help to predict how material and melts are transported through the lowermost mantle and how material transport between mantle and core works on a larger scale.
How might future research on ultralow velocity zones improve our understanding of the Earth as a dynamic and evolving planet?
SH: A better understanding of ULVZs, including their origins, internal structure, and variability, could improve our knowledge about the internal processes driving our planet. For example, ULVZs could buffer heat flux across the core–mantle boundary, which is important because the thermal conditions in this part of our planet have been shown to strongly impact Earth’s magnetic field. Moreover, a paper from Li and Zhong in 2017 (M. Li and S. Zhong, Earth Planet. Sci. Lett. 478, 47–57; 2017) argued that mantle plumes are also largely controlled by the temperature conditions near the core–mantle boundary, and ULVZs may contribute to determining the locations where these plumes originate.
SR: Using insight from ULVZs, we can learn a lot about mantle flow, the recycling of subducted material, core–mantle interactions (both material and heat flows) and the evolution of our planet. As we understand so little about ULVZs, it is difficult to tell what we will be able to do with ULVZ studies. It all depends on which (or which range of) models of the ULVZ that have been proposed will be the one to explain all (or some) ULVZ observations. Potentially ULVZs can help us to learn more about Earth’s evolution through resolving the dynamics of a basal magma ocean, plume dynamics and generation where ULVZs might represent plume roots or core leaks, core dynamics (for example, magnetic field reversal paths), and geochemical reservoirs. There might be other options if none of our current models of ULVZ origin is correct.
How do you use or apply geophysical data in your research on ultralow velocity zones? What kinds of collaborations are needed to advance the state of the art?
SH: Much of my work involves modelling seismic phases that have reflected off the core–mantle boundary in order to map ULVZs and gain insights into their structure and origins. I have largely been focused on seismic data recorded by stations in Antarctica, which stems back to other polar investigations I have worked on. The Antarctic dataset provides coverage of the lowermost mantle beneath a large portion of the southern hemisphere; an area that had not been extensively investigated for ULVZs before. This part of the deep Earth is also interesting because it is not associated with present-day (or recent) subduction activities and does not coincide with the LLVPs. Consequently, it is likely that large-scale mantle downwellings and upwellings do not impact this specific part of the mantle, making it a unique location for investigating anomalous lowermost mantle structures. I have also been collaborating with colleagues at Arizona State University to combine seismic observations with those from geodynamic models and mineral physics experiments. Given the complexity of ULVZs, I believe that a multi-disciplinary approach is essential for a comprehensive understanding, and we are in the process of proposing an expansion of our collaborative research efforts.
SR: I am mainly a seismologist using high-frequency seismic data, often from seismic arrays, to resolve localized ULVZ structure using core reflected and diffracted phases. I work with other seismologists using other seismic probes to reduce trade-offs for ULVZ parameters. I also work on novel seismic probes to resolve ULVZ structure to increase core–mantle boundary sampling, improve resolution and to reduce parameter uncertainty. I collaborate with colleagues from computational and experimental mineral physics, computational seismology, and geodynamics to interpret the seismic images. These multi-disciplinary collaborations are essential to come to robust interpretations of the seismic data. We are currently making progress on several fronts due to the development of new experimental and numerical techniques, new data, and the use of artificial intelligence to interpret the seismic data and characterize ULVZs in different regions of the Earth.
Studying the deep Earth is difficult due to limited direct observations and rock samples. How much can we rely on the interpretation of seismic data to unlock the compositional and geophysical structures of ultralow velocity zones?
SH: This is an excellent question. Like all methods, seismic investigations of ULVZs have limitations in terms of resolution. If they are too thin, they become indistinguishable. Additionally, there are trade-offs between the thickness of ULVZs and their seismic velocity, which can be challenging to reconcile. For instance, a thinner ULVZ with slower velocity may appear the same as a thicker ULVZ with less slow velocity. In my opinion, seismic studies still hold great potential in unravelling the structure of the deep Earth structure. However, we cannot work in a bubble. To really understand the structure and composition of ULVZs, we need insights from other fields such as high temperature or pressure mineral physics and geodynamic models. Collectively, such observations will further our ability to tease apart the details of deep Earth anomalies.
SR: Obviously, the seismic observations sampling ULVZs require interpretation. This is typically done through modelling or inversion of the seismic data together with information from other geodisciplines such as mineral physics and geodynamics. The biggest drawback here is that many of the modelling approaches are not fully using 3D structures both due to limitations in our ability to model seismic wave propagation at the required frequencies and the increase of the parameter space that needs sampling. We have made great progress in reducing the location uncertainty of ULVZs through combined analysis of ULVZ datasets as well as in better understanding of the seismic wavefield. Similar advances to resolve the internal structure are underway. The general framework of what we describe as ULVZs and where we can find them is there and is quite robust. Exploring the details will allow us to better understand ULVZ composition, origins, and dynamics.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Bahadori, A. Ultralow velocity zones in the deep Earth. Nat. Geosci. 17, 275–277 (2024). https://doi.org/10.1038/s41561-024-01415-3
Published:
Issue Date:
DOI: https://doi.org/10.1038/s41561-024-01415-3