Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Article
  • Published:

Central shutdown and surrounding activation of aftershocks from megathrust earthquake stress transfer

Abstract

Megathrust earthquakes release and transfer stress that has accumulated over hundreds of years, leading to large aftershocks that can be highly destructive. Understanding the spatiotemporal pattern of megathrust aftershocks is key to mitigating the seismic hazard. However, conflicting observations show aftershocks concentrated either along the rupture surface itself, along its periphery or well beyond it, and they can persist for a few years to decades. Here we present aftershock data following the four largest megathrust earthquakes since 1960, focusing on the change in seismicity rate following the best-recorded 2011 Tohoku earthquake, which shows an initially high aftershock rate on the rupture surface that quickly shuts down, while a zone up to ten times larger forms a ring of enhanced seismicity around it. We find that the aftershock pattern of Tohoku and the three other megathrusts can be explained by rate and state Coulomb stress transfer. We suggest that the shutdown in seismicity in the rupture zone may persist for centuries, leaving seismicity gaps that can be used to identify prehistoric megathrust events. In contrast, the seismicity of the surrounding area decays over 4–6 decades, increasing the seismic hazard after a megathrust earthquake.

This is a preview of subscription content, access via your institution

Access options

Rent or buy this article

Prices vary by article type

from$1.95

to$39.95

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Aftershocks of M ≥9.0 megathrust ruptures since 1960.
Fig. 2: Change in seismicity rate beginning 5 yr after the Tohoku M 9 earthquake.
Fig. 3: Cross-sections of seismicity-rate and stress change for the Tohoku M 9 earthquake.
Fig. 4: Modelled response of seismicity to a megathrust earthquake.
Fig. 5: Contemporary gaps in seismicity could represent prehistoric megathrust earthquakes.

Similar content being viewed by others

Data availability

We used the USGS ANSS catalogue (https://earthquake.usgs.gov/earthquakes/search/), the JMA catalogue (https://www.data.jma.go.jp/svd/eqev/data/daily_map/index.html and https://www.data.jma.go.jp/svd/eqev/data/bulletin/eqdoc.html) and the NIED F-net focal mechanism catalogue (https://www.fnet.bosai.go.jp/event/search). We also used a published 1960 Chile earthquake catalogue25, and a published 1960–1966 Alaska earthquake catalogue43. All seismic slip (‘finite fault’) models are published and cited; those also available from http://equake-rc.info/SRCMOD/searchmodels/allevents/ include the 2003 Tokachi-oki57 and 2011 Tohoku58 earthquakes (used for Coulomb calculations) and 1944 Tonankai53, 1946 Nankai53 and 2010 Maule59 (used for display in figures). Seismic slip models available only from publications include the 1700 M ~9.0 Cascadia52, 1762 M ~8.8 Arakan49, 1868 M ~9.0 Arica, Peru–Chile51, 1906 M ~8.8 Ecuador50, 1952 Kamchatka48, 1960 Valdivia24, 1964 Prince William Sound44 and 2004 Sumatra45 earthquakes. Source data are provided with this paper.

Code availability

The numerical methodology used in this study is described in Methods and in refs. 55,56. We used the Coulomb 3.3 software60,61,62,63,64 (software, tutorial files and user guide accessible via http://www.temblor.net/coulomb). For magnitude of completeness and aftershock decay calculations, we used ZMAP18 (http://www.seismo.ethz.ch/en/research-and-teaching/products-software/software/ZMAP/ and https://github.com/swiss-seismological-service/zmap7).

References

  1. Ryder, I. et al. Large extensional aftershocks in the continental forearc triggered by the 2010 Maule earthquake, Chile. Geophys. J. Int. 188, 879–890 (2012).

    Article  Google Scholar 

  2. Kato, A. & Igarashi, T. Regional extent of the large coseismic slip zone of the 2011 Mw 9.0 Tohoku-oki earthquake delineated by on-fault aftershocks. Geophys. Res. Lett. 39, L15301 (2012).

    Article  Google Scholar 

  3. Henry, C. & Das, S. Aftershock zones of large shallow earthquakes: fault dimensions, aftershock area expansion and scaling relations. Geophys. J. Int. 147, 272–293 (2001).

    Article  Google Scholar 

  4. Kanamori, H. Rupture processes of subduction-zone. Annu. Rev. Earth Planet Sci. 14, 293–322 (1986).

    Article  Google Scholar 

  5. Rietbrock, A. et al. Aftershock seismicity of the 2010 Maule Mw = 8.8, Chile, earthquake: correlation between co-seismic slip models and aftershock distribution? Geophys. Res. Lett. 39, L08310 (2012).

    Google Scholar 

  6. Bilek, S. L. & Lay, T. Subduction zone megathrust earthquakes. Geosphere 14, 1468–1500 (2018).

    Article  Google Scholar 

  7. Lengliné, O., Enescu, B., Peng, Z. & Shiomi, K. Decay and expansion of the early aftershock activity following the 2011, Mw9.0 Tohoku earthquake. Geophys. Res. Lett. 39, L18309 (2012).

    Article  Google Scholar 

  8. Thatcher, W. Order and diversity in the modes of circum-Pacific earthquake recurrence. J. Geophys. Res. 95, 2609–2623 (1990).

    Article  Google Scholar 

  9. Woessner, J., Schorlemmer, D., Wiemer, S. & Mai, P. M. Spatial correlation of aftershock locations and on-fault main shock properties. J. Geophys. Res. 111, B08301 (2006).

    Google Scholar 

  10. Wetzler, N., Lay, T., Brodsky, E. E. & Kanamori, H. Systematic deficiency of aftershocks in areas of high coseismic slip for large subduction zone earthquakes. Sci. Adv. 4, eaao3225 (2018).

    Article  Google Scholar 

  11. Parsons, T. Global Omori law decay of triggered earthquakes: large aftershocks outside the classical aftershock zone. J. Geophys. Res. 107, 2199 (2002).

    Google Scholar 

  12. Hainzl, S., Christophersen, A., Rhoades, D. & Harte, D. Statistical estimation of the duration of aftershock sequences. Geophys. J. Int. 205, 1180–1189 (2016).

    Article  Google Scholar 

  13. Lay, T., Astiz, L., Kanamori, H. & Christensen, D. H. Temporal variation of large intraplate earthquakes in coupled subduction zones. Phys. Earth Planet. Int. 54, 258–312 (1989).

    Article  Google Scholar 

  14. Ogata, Y. Space–time point-process models for earthquake occurrences. Ann. Inst. Stat. Math. 50, 379–402 (1998).

    Article  Google Scholar 

  15. Stein, R. S. The role of stress transfer in earthquake occurrence. Nature 402, 605–609 (1999).

    Article  Google Scholar 

  16. Harris, R. A. & Simpson, R. W. Suppression of large earthquakes by stress shadows: a comparison of Coulomb and rate-and-state failure. J. Geophys. Res. 103, 24439–24451 (1998).

    Article  Google Scholar 

  17. Toda, S., Stein, R. S., Beroza, G. & Marsan, D. Aftershocks halted by static stress shadows. Nat. Geosci. 5, 410–413 (2012).

    Article  Google Scholar 

  18. Woessner, J. & Wiemer, S. Assessing the quality of earthquake catalogues: estimating the magnitude of completeness and its uncertainty. Bull. Seismol. Soc. Am. 95, 684–698 (2005).

    Article  Google Scholar 

  19. Iinuma, T. et al. Coseismic slip distribution of the 2011 off the Pacific Coast of Tohoku Earthquake (M9.0) refined by means of seafloor geodetic data. J. Geophys. Res. 117, B07409 (2012).

    Google Scholar 

  20. Lin, J. & Stein, R. S. Stress triggering in thrust and subduction earthquakes and stress interaction between the southern San Andreas and nearby thrust and strike–slip faults. J. Geophys. Res. 109, B02303 (2004).

    Google Scholar 

  21. Dieterich, J. H. A constitutive law for the rate of earthquake production and its application to earthquake clustering. J. Geophys. Res. 99, 2601–2618 (1994).

    Article  Google Scholar 

  22. Hardebeck, J. L., Nazareth, J. J. & Hauksson, E. The static stress change triggering model: constraints from two southern California aftershock sequences. J. Geophys. Res. 103, 24427–24437 (1998).

    Article  Google Scholar 

  23. Helmstetter, A. & Shaw, B. E. Relation between stress heterogeneity and aftershock rate in the rate-and-state model. J. Geophys. Res. 111, B07304 (2006).

    Google Scholar 

  24. Barrientos, S. E. & Ward, S. N. The 1960 Chile earthquake: inversion for slip distribution from surface deformation. Geophys. J. Int. 103, 589–598 (1990).

    Article  Google Scholar 

  25. Cifuentes, I. L. The 1960 Chilean earthquakes. J. Geophys. Res. 94, 665–680 (1989).

    Article  Google Scholar 

  26. Scholz, C. H. Mechanisms of seismic quiescences. Pure Appl. Geophys. 126, 701–718 (1988).

    Article  Google Scholar 

  27. Mogi, K. Some features of recent seismic activity in and near Japan (2) Activity before and after great earthquakes. Bull. Earthq. Res. Inst. 47, 395–417 (1969).

    Google Scholar 

  28. Schurr, B. et al. Forming a Mogi doughnut in the years prior to and immediately before the 2014 M8.1 Iquique, northern Chile earthquake. Geophys. Res. Lett. 47, e2020GL088351 (2020).

    Article  Google Scholar 

  29. Pollitz, F. F., Stein, R. S., Sevilgen, V. & Bürgmann, R. The 11 April 2012 east Indian Ocean earthquake triggered large aftershocks worldwide. Nature 490, 250–253 (2012).

    Article  Google Scholar 

  30. Freed, A. M. & Lin, J. Accelerated stress buildup on the southern San Andreas fault and surrounding regions caused by Mojave Desert earthquakes. Geology 30, 571–574 (2002).

    Article  Google Scholar 

  31. Bürgmann, R., Uchida, N., Hu, Y. & Matsuzawa, T. Tohoku rupture reloaded? Nat. Geosci. 9, 183 (2016).

    Article  Google Scholar 

  32. Iinuma, T. et al. Seafloor observations indicate spatial separation of coseismic and postseismic slips in the 2011 Tohoku earthquake. Nat. Commun. 7, 13506 (2016).

    Article  Google Scholar 

  33. Hu, Y. & Wang, T. Spherical-earth finite element model of short-term postseismic deformation following the 2004 Sumatra earthquake. J. Geophys. Res. 117, B0504 (2012).

    Google Scholar 

  34. Baba, T., Hirata, K., Hori & Sakaguchi, H. Offshore geodetic data conducive to the estimation of the afterslip distribution following the 2003 Tokachi-oki earthquake. Earth Planet. Sci. Lett. 241, 281–292 (2006).

    Article  Google Scholar 

  35. Neo, J. C., Huang, Y., Yao, D. & Wei, S. Is the aftershock zone area a good proxy for the mainshock rupture area? Bull. Seismol. Soc. Am. 111, 424–438 (2020).

    Article  Google Scholar 

  36. McCann W. R., Nishenko S. P., Sykes L. R. & Krause, J. in Earthquake Prediction and Seismicity Patterns. Contributions to Current Research in Geophysics (ed. Wyss, M.) 1082-1147 (Birkhäuser, 1979).

  37. Toda, S. & Stein, R. S. Recent large Japan quakes are aftershocks of the 2011 Tohoku earthquake. Temblor https://doi.org/10.32858/temblor.175 (2021).

  38. Wang, K. & Bilek, S. L. Invited review paper: Fault creep caused by subduction of rough seafloor relief. Tectonophysics 610, 1–24 (2014).

    Article  Google Scholar 

  39. Obana, K. et al. Earthquake activity in northern Cascadia subduction zone off Vancouver Island revealed by ocean‐bottom seismograph observations. Bull. Seismol. Soc. Am. 105, 489–495 (2014).

    Article  Google Scholar 

  40. Gomberg, J. & Bodin, P. The productivity of Cascadia aftershock sequences. Bull. Seismol. Soc. Am. 111, 1494–1507 (2021).

    Article  Google Scholar 

  41. Lutikov, A. I., Rogozhin, E. A., Dontsova, G., Yu & Zhukovets, V. N. The Mw 7.8 earthquake of July 17, 2017 off the Commander Islands and other large earthquakes at the western segment of the Aleutian Island Arc. J. Volcanol. Seismol. 13, 112–123 (2019).

    Article  Google Scholar 

  42. Okal, E. A. & Synolakis, C. E. Far-field tsunami hazard from mega-thrust earthquakes in the Indian Ocean. Geophys. J. Int. 172, 995–1015 (2008).

    Article  Google Scholar 

  43. Page, R. Aftershocks and microaftershocks of the great Alaska earthquake of 1964. Bull. Seismol. Soc. Am. 58, 1131–1168 (1968).

    Google Scholar 

  44. Ichinose, G., Somerville, P., Thio, H. K., Graves, R. & O’Connell, D. Rupture process of the 1964 Prince William Sound, Alaska, earthquake from the combined inversion of seismic, tsunami, and geodetic data. J. Geophys. Res. 112, B07306 (2007).

    Google Scholar 

  45. Banerjee, P., Pollitz, F., Nagarajan, B. & Bürgmann, R. Coseismic slip distributions of the 26 December 2004 Sumatra–Andaman and 28 March 2005 Nias earthquakes from GPS static offsets. Bull. Seismol. Soc. Am. 97, S86–S102 (2007).

    Article  Google Scholar 

  46. Tajima, F., Mori, J. & Kennett, B. L. N. A review of the 2011 Tohoku-oki earthquake (Mw 9.0): large-scale rupture across heterogeneous plate coupling. Tectonophysics 586, 15–34 (2013).

    Article  Google Scholar 

  47. Obana, K. et al. Seismic velocity structure and its implications for oceanic mantle hydration in the trench—outer rise of the Japan Trench. Geophys. J. Int. 217, 1629–1642 (2019).

    Article  Google Scholar 

  48. MacInnes, B. T., Weiss, R., Bourgeois, J. & Pinegina, T. K. Slip distribution of the 1952 Kamchatka great earthquake based on near-field tsunami deposits and historical records. Bull. Seismol. Soc. Am. 100, 1695–1709 (2010).

    Article  Google Scholar 

  49. Cummins, P. R. The potential for giant tsunamigenic earthquakes in the northern Bay of Bengal. Nature 449, 75–78 (2007).

    Article  Google Scholar 

  50. Yoshimoto, M. et al. Depth-dependent rupture mode along the Ecuador–Colombia subduction zone. Geophys. Res. Lett. 44, 2203–2210 (2017).

    Article  Google Scholar 

  51. Okal, E. A., Borrero, J. C. & Synolakis, C. E. Evaluation of tsunami risk from regional earthquakes at Pisco, Peru. Bull. Seismol. Soc. Am. 96, 1634–1648 (2006).

    Article  Google Scholar 

  52. Wang, K. & Trehu, A. M. Invited review paper: Some outstanding issues in the study of great megathrust earthquakes—the Cascadia example. J. Geodyn. 98, 1–18 (2016).

    Article  Google Scholar 

  53. Sagiya, T. & Thatcher, W. Coseismic slip resolution along a plate boundary megathrust: the Nankai Trough, southwest Japan. J. Geophys. Res. 104, 1111–1129 (1999).

    Article  Google Scholar 

  54. Wells, D. & Coppersmith, K. J. New empirical relationships among magnitude, rupture length, rupture width, rupture area, and surface displacement. Bull. Seismol. Soc. Am. 84, 974–1002 (1994).

    Google Scholar 

  55. Toda, S., Stein, R. S., Richards-Dinger, K. & Bozkurt, S. B. Forecasting the evolution of seismicity in southern California: animations built on earthquake stress transfer. J. Geophys. Res. 110, B05S16 (2005).

    Google Scholar 

  56. Toda, S. & Stein, R. S. Long- and short-term stress interaction of the 2019 Ridgecrest sequence and Coulomb-based earthquake forecasts. Bull. Seismol. Soc. Am. 110, 1765–1780 (2020).

    Article  Google Scholar 

  57. Yagi, Y. Source rupture process of the 2003 Tokachi-oki earthquake determined by joint inversion of teleseismic body wave and strong ground motion data. Earth Planets Space 56, 311–316 (2004).

    Article  Google Scholar 

  58. Ide, S., Baltay, A. & Beroza, G. C. Shallow dynamic overshoot and energetic deep rupture in the 2011 Mw 9.0 Tohoku-oki earthquake. Science 332, 1426–1429 (2011).

    Article  Google Scholar 

  59. Lorito, S. et al. Limited overlap between the seismic gap and coseismic slip of the great 2010 Chile earthquake. Nat. Geosci. 4, 173–177 (2011).

    Article  Google Scholar 

  60. Toda, S., Stein, R. S., Sevilgen, V. & Lin, J. Coulomb 3.3, Graphic-Rich Deformation and Stress Change Software for Earthquake, Tectonic, and Volcano Research and Teaching—User Guide Open-File Report 2011-1060 (USGS, 2011).

  61. Lin, Y.-N. N. et al. Coseismic and postseismic slip associated with the 2010 Maule earthquake, Chile: characterizing the Arauco Peninsula barrier effect. J. Geophys. Res. 118, 3142–3159 (2013).

    Article  Google Scholar 

  62. Lobkovsky, L. I. et al. The Komandor seismic gap: earthquake prediction and tsunami computation. Oceanology 54, 519–531 (2014).

    Article  Google Scholar 

  63. Byrne, D. E., Sykes, L. R. & Davis, D. M. Great thrust earthquakes and aseismic slip along the plate boundary of the Makran Subduction Zone. J. Geophys. Res. 97, 449–478 (1992).

    Article  Google Scholar 

  64. Yeats, R. S. Active Faults of the World 332–334 (Cambridge Univ. Press, 2012).

Download references

Acknowledgements

We thank C. Scholz, T. Parsons and W. Thatcher for insightful comments on the manuscript. We gratefully acknowledge support from the SBIR programme of the US National Science Foundation (R.S.S.) and the WTW Research Network (R.S.S.). The funders were provided with the manuscript upon submission, but had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.

Author information

Authors and Affiliations

Authors

Contributions

S.T. and R.S.S. contributed equally to the ideas, methods, text and figures in this study.

Corresponding authors

Correspondence to Shinji Toda or Ross S. Stein.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Geoscience thanks Lingling Ye, Olaf Zielke and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editor: Louise Hawkins, in collaboration with the Nature Geoscience team.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 Seismicity time series in the simplified corona and core areas.

a, Map of the seismicity rate change. The 13 Feb M 7.1, 20 Mar M 7.1, and 1 May M 6.9 earthquakes in 2021 are shown as stars. b, Time series of earthquakes in the corona. Notice that the post-M 9 rate in the core has remained low for twice the duration of the several low-rate preseismic periods. c, Time series of earthquakes in the core. We use quartiles to evaluate uncertainty because the rates are not normally distributed. d-e, Omori decay parameters fitted using ref. 18. Notice the smaller earthquakes and steeper decay (p exponent) in the core than in the corona.

Extended Data Fig. 2 Relationship between seismicity rate change and postseismic slip.

a, Seismicity rate change for 2003 Tokachi-Oki, with coseismic57 and postseismic slip during the first year34. Here we compare the period 5-10 yr after the quake (2008/09/26 - 2011/03) to the 5.7-yr background period (1998/01/01-2003/09/25). b. Seismicity associated with 2010 Maule (ANSS M ≥ 4.5 catalog), with coseismic59 and postseismic slip61. c, Postseismic slip during the first 8 months after 2011 Tohoku32 superimposed on Figs. 1a and 2a. Panel a (right) adapted with permission from ref. 34, Springer Nature Limited. Panel b (right) reproduced with permission from ref. 59, John Wiley and Sons.

Extended Data Fig. 3 Corona growth with time, and Coulomb stress imparted to focal mechanisms in the core and corona.

a–b, Two-day and two-month corona extent. c, Simplified core area with beachballs colored by maximum Coulomb stress change. Because for each mechanism, we take the nodal plane with the most positive (maximum) stress change, these results could be biased toward stress increases. Mechanisms from 1998/01/01 - 2011/03/10, M ≥ 3, depth≤150 km, from F-Net catalog.

Extended Data Fig. 4 Model of seismicity evolution in a heterogeneous faulting environment.

a, Simulated time histories given a standard deviation 3 times larger than a mean stress decrease. ta is the aftershock duration in rate/state friction. Each curve is a mean of 10,000 Monte Carlo simulations. b, Time history given a standard deviation equal to a mean stress increase. c–d, Time histories under different assumptions for the mean and standard deviation of the stress changes. e, Figure from Scholz (1988)26. The concentration of longer-lasting aftershocks at the periphery resembles our corona, while the briefer aftershocks (A) that fade into quiescence (Q1) at a rate lower than the background (B) resemble our core. Panel e adapted with permission from ref. 26, Springer Nature Limited.

Extended Data Fig. 5 Comparison of our model with results of Parsons (2002).

a, This study. b, Fig. 9 of Parsons (2002)11. For simplicity, we have colored the curves and removed the uncertainty bounds. Parsons reported that aftershocks with shear stress increases (red curve, b) tend to locate 25 km father from the moment centroids than aftershocks with the shear stress decreases (blue curve, b). Thus, the decreases could occur in or near the core, and the increases in or near the corona.

Extended Data Fig. 6 Seismicity holes evident today along major subduction zones for seismicity beginning after the megathrust in Fig. 1 struck.

a, Japan Trench, b, Sunda Trench, c, Alaska-Aleutian Trench, d, Peru–Chile Trench. For c–d, we begin when the seismic catalog detection markedly improves in about 1976. All maps are at the same scale.

Extended Data Fig. 7 Seismicity holes associated with candidate historic or prehistoric megathrust earthquakes.

a, The Commander (Komandor) Seismic Gap extends for 700 km along the northwest Aleutian Trench41, 60, where oblique slip is partitioned between subduction convergence and a parallel back-arc transform fault. b, The hole is most evident for subduction mechanisms6. c, The Makran Deformation Front (Makran Trench) appears to have two holes, the eastern hole at the site of a 1765 earthquake, and the western hole perhaps associated with the debated 1483 earthquake42, 61, 62.

Extended Data Fig. 8 Masking swarms and secondary aftershocks in the seismicity rate change map.

a, All data. b, Map of coefficient of variation of seismicity inter-event times with site 1 for a seismic swarm and site 2 aftershocks of a secondary mainshock during the pre-M 9 period, and site 3, steady background seismicity. c, Same as a but with sites of COV ≥ 3 masked.

Source data

Extended Data Fig. 9 Schematic illustration of how seismicity rate changes are derived from stress imparted to focal mechanisms.

a, Each focal mechanism is a proxy for a small-to-moderate fault on which that earthquake stuck (top panel in a). These earthquakes then receive coseismic stress from a nearby mainshock (second panel in a), some promoting failure (red) and some inhibiting failure (blue). The applied stress amplifies or diminishes the background seismicity rate (bottom panel in a), according the the seismicity rate equation21. Finally, to make a map of forecast seismicity as in Fig. 2b, the updated numbers on the focal mechanism plots in the bottom panel in a are spatially smoothed by a moving kernel on the grid nodes. This illustration is from ref. 56. b, Map of the Learning Period earthquakes (M ≥ 6.5 during 3-11-2011 to 3-10-2016) that are used in the model. Panel a reproduced with permission from ref. 45, Seismological Society of America.

Source data

Source Data Fig. 2a

Fig. 2a (and Extended Data Fig. 8a) numerical file.

Source Data Fig. 2b

Fig. 2b (without removal of COV areas) numerical file.

Source Data Fig. 3a

Fig. 3a numerical file.

Source Data Extended Data Fig. 8a

Extended Data Fig. 8a (and Fig. 2a) numerical file.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Toda, S., Stein, R.S. Central shutdown and surrounding activation of aftershocks from megathrust earthquake stress transfer. Nat. Geosci. 15, 494–500 (2022). https://doi.org/10.1038/s41561-022-00954-x

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s41561-022-00954-x

This article is cited by

Search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing