Abstract
Hydrologic loads can stimulate seismicity in the Earth’s crust1. However, evidence for the triggering of large earthquakes remains elusive. The southern San Andreas Fault (SSAF) in Southern California lies next to the Salton Sea2, a remnant of ancient Lake Cahuilla that periodically filled and desiccated over the past millennium3,4,5. Here we use new geologic and palaeoseismic data to demonstrate that the past six major earthquakes on the SSAF probably occurred during highstands of Lake Cahuilla5,6. To investigate possible causal relationships, we computed time-dependent Coulomb stress changes7,8 due to variations in the lake level. Using a fully coupled model of a poroelastic crust9,10,11 overlying a viscoelastic mantle12,13, we find that hydrologic loads increased Coulomb stress on the SSAF by several hundred kilopascals and fault-stressing rates by more than a factor of 2, which is probably sufficient for earthquake triggering7,8. The destabilizing effects of lake inundation are enhanced by a nonvertical fault dip14,15,16,17, the presence of a fault damage zone18,19 and lateral pore-pressure diffusion20,21. Our model may be applicable to other regions in which hydrologic loading, either natural8,22 or anthropogenic1,23, was associated with substantial seismicity.
This is a preview of subscription content, access via your institution
Access options
Access Nature and 54 other Nature Portfolio journals
Get Nature+, our best-value online-access subscription
$29.99 / 30 days
cancel any time
Subscribe to this journal
Receive 51 print issues and online access
$199.00 per year
only $3.90 per issue
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




Data availability
The Abaqus data files, Lake Cahuilla and Salton Sea data files and processed model results are available on Zenodo (http://doi.org/10.5281/zenodo.7714217). Source data are provided with this paper.
Code availability
All relevant MATLAB post-processing codes and sample-plotting codes are available on Zenodo (http://doi.org/10.5281/zenodo.7714217).
References
Talwani, P. On the nature of reservoir-induced seismicity. Pure Appl. Geophys. 150, 473–492 (1997).
Tostrud, M. B. The Salton Sea, 1906-1996, Computed and Measured Salinities and Water Levels. Draft Report, Colorado River Board of California (1997).
Waters, M. R. Late Holocene lacustrine chronology and archaeology of ancient Lake Cahuilla, California. Quat. Res. 19, 373–387 (1983).
Philibosian, B., Fumal, T. & Weldon, R. San Andreas fault earthquake chronology and Lake Cahuilla history at Coachella, California. Bull. Seismol. Soc. Am. 101, 13–38 (2011).
Rockwell, T. K., Meltzner, A. J. & Haaker, E. C. Dates of the two most recent surface ruptures on the southernmost San Andreas fault recalculated by precise dating of Lake Cahuilla dry periods. Bull. Seismol. Soc. Am. 108, 2634–2649 (2018).
Rockwell, T. K., Meltzner, A. J., Haaker, E. C. & Madugo, D. The late Holocene history of Lake Cahuilla: two thousand years of repeated fillings within the Salton Trough, Imperial Valley, California. Quat. Sci. Rev. 282, 107456 (2022).
King, G. C. P., Stein, R. S. & Lin, J. Static stress changes and the triggering of earthquakes. Bull. Seismol. Soc. Am. 84, 935–953 (1994).
Cocco, M. Pore pressure and poroelasticity effects in Coulomb stress analysis of earthquake interactions. J. Geophys. Res. Solid Earth 107, 2030 (2002).
Rice, J. R. & Cleary, M. P. Some basic stress diffusion solutions for fluid-saturated elastic porous media with compressible constituents. Rev. Geophys. 14, 227–241 (1976).
Wang, H. Theory of Linear Poroelasticity: With Applications to Geomechanics and Hydrogeology (Princeton Univ. Press, 2000).
LaBonte, A., Brown, K. & Fialko, Y. Hydrogeologic detection and finite-element modeling of a slow slip event in the Costa Rica prism toe. J. Geophys. Res. Solid Earth 114, B00A02 (2009).
Segall, P. Earthquake and Volcano Deformation (Princeton Univ. Press, 2010).
Barbot, S. & Fialko, Y. A unified continuum representation of post-seismic relaxation mechanisms: semi-analytic models of afterslip, poroelastic rebound and viscoelastic flow. Geophys. J. Int. 182, 1124–1140 (2010).
Fialko, Y. Interseismic strain accumulation and the earthquake potential on the southern San Andreas fault system. Nature 441, 968–971 (2006).
Lin, G., Shearer, P. M. & Hauksson, E. Applying a three-dimensional velocity model, waveform cross correlation, and cluster analysis to locate southern California seismicity from 1981 to 2005. J. Geophys. Res. Solid Earth 112, B12309 (2007).
Fuis, G. S., Scheirer, D. S., Langenheim, V. E. & Kohler, M. D. A new perspective on the geometry of the San Andreas fault in Southern California and its relationship to lithospheric structure. Bull. Seismol. Soc. Am. 102, 236–251 (2012).
Lindsey, E. O. et al. Interseismic strain localization in the San Jacinto fault zone. Pure Appl. Geophys. 171, 2937–2954 (2014).
Fialko, Y. et al. Deformation on nearby faults induced by the 1999 Hector Mine earthquake. Science 297, 1858–1862 (2002).
Cochran, E. S. et al. Seismic and geodetic evidence for extensive, long-lived fault damage zones. Geology 37, 315–318 (2009).
Caine, J. S., Evans, J. P. & Forster, C. B. Fault zone architecture and permeability structure. Geology 24, 1025–1028 (1996).
Bense, V., Gleeson, T., Loveless, S., Bour, O. & Scibek, J. Fault zone hydrogeology. Earth Sci. Rev. 127, 171–192 (2013).
Nof, R. et al. Rising of the lowest place on Earth due to Dead Sea water-level drop: evidence from SAR interferometry and GPS. J. Geophys. Res. Solid Earth 117, B05412 (2012).
Gupta, H. K. Reservoir Induced Earthquakes (Elsevier, 1992).
Weldon, R. J., Fumal, T. E., Biasi, G. P. & Scharer, K. M. Past and future earthquakes on the San Andreas fault. Science 308, 966–967 (2005).
Field, E. H. et al. Uniform California Earthquake Rupture Forecast, version 3 (UCERF3)—the time-independent model. Bull. Seismol. Soc. Am. 104, 1122–1180 (2014).
Fumal, T. E. Timing of large earthquakes since AD 800 on the Mission Creek strand of the San Andreas fault zone at Thousand Palms Oasis, near Palm Springs, California. Bull. Seismol. Soc. Am. 92, 2841–2860 (2002).
Gurrola, L. D. & Rockwell, T. K. Timing and slip for prehistoric earthquakes on the Superstition Mountain fault, Imperial Valley, southern California. J. Geophys. Res. Solid Earth 101, 5977–5985 (1996).
Thomas, A. P. & Rockwell, T. K. A 300- to 550-year history of slip on the Imperial fault near the U.S.-Mexico border: missing slip at the Imperial fault bottleneck. J. Geophys. Res. Solid Earth 101, 5987–5997 (1996).
Luttrell, K., Sandwell, D., Smith-Konter, B., Bills, B. & Bock, Y. Modulation of the earthquake cycle at the southern San Andreas fault by lake loading. J. Geophys. Res. Solid Earth 112, B08411 (2007).
Brothers, D., Kilb, D., Luttrell, K., Driscoll, N. & Kent, G. Loading of the San Andreas fault by flood-induced rupture of faults beneath the Salton Sea. Nat. Geosci. 4, 486–492 (2011).
Sieh, K. Slip rate across the San Andreas fault and prehistoric earthquakes at Indio, California. Eos Trans. AGU 67, 1200 (1986).
Mueller, K. Neotectonics, Alluvial History and Soil Chronology of the Southwestern Margin of the Sierra de Los Cucapas, Baja California Norte. Master’s thesis, San Diego State Univ. (1984).
Bronk Ramsey, C. Bayesian analysis of radiocarbon dates. Radiocarbon 51, 337–360 (2009).
Reimer, P. J. et al. The IntCal20 Northern Hemisphere radiocarbon age calibration curve (0–55 cal kBP). Radiocarbon 62, 725–757 (2020).
Roeloffs, E. Fault stability changes induced beneath a reservoir with cyclic variations in water level. J. Geophys. Res. Solid Earth 93, 2107–2124 (1988).
Segall, P. Earthquakes triggered by fluid extraction. Geology 17, 942–946 (1989).
Gupta, H. K. A review of recent studies of triggered earthquakes by artificial water reservoirs with special emphasis on earthquakes in Koyna, India. Earth Sci. Rev. 58, 279–310 (2002).
Rajendran, K. & Talwani, P. The role of elastic, undrained, and drained responses in triggering earthquakes at Monticello Reservoir, South Carolina. Bull. Seismol. Soc. Am. 82, 1867–1888 (1992).
Simpson, D., Leith, W. & Scholz, C. Two types of reservoir-induced seismicity. Bull. Seismol. Soc. Am. 78, 2025–2040 (1988).
Tao, W., Masterlark, T., Shen, Z. & Ronchin, E. Impoundment of the Zipingpu reservoir and triggering of the 2008 Mw 7.9 Wenchuan earthquake, China. J. Geophys. Res. Solid Earth 120, 7033–7047 (2015).
Mitchell, T. & Faulkner, D. The nature and origin of off-fault damage surrounding strike-slip fault zones with a wide range of displacements: a field study from the Atacama fault system, northern Chile. J. Struct. Geol. 31, 802–816 (2009).
Dor, O., Ben-Zion, Y., Rockwell, T. & Brune, J. Pulverized rocks in the Mojave section of the San Andreas Fault Zone. Earth Planet. Sci. Lett. 245, 642–654 (2006).
Rockwell, T. et al. Chemical and physical characteristics of pulverized Tejon Lookout granite adjacent to the San Andreas and Garlock faults: implications for earthquake physics. Pure Appl. Geophys. 166, 1725–1746 (2009).
Morton, N., Girty, G. H. & Rockwell, T. K. Fault zone architecture of the San Jacinto fault zone in Horse Canyon, southern California: a model for focused post-seismic fluid flow and heat transfer in the shallow crust. Earth Planet. Sci. Lett. 329, 71–83 (2012).
Rempe, M. et al. Damage and seismic velocity structure of pulverized rocks near the San Andreas Fault. J. Geophys. Res. Solid Earth 118, 2813–2831 (2013).
Morrow, C., Lockner, D., Moore, D. & Hickman, S. Deep permeability of the San Andreas fault from San Andreas fault observatory at depth (SAFOD) core samples. J. Struct. Geol. 64, 99–114 (2014).
Xue, L., Brodsky, E. E., Erskine, J., Fulton, P. M. & Carter, R. A permeability and compliance contrast measured hydrogeologically on the San Andreas Fault. Geochem. Geophys. Geosyst. 17, 858–871 (2016).
USGS and California Geological Survey. Quaternary Fault and Fold Database for the United States (accessed 10 July 2019); https://www.usgs.gov/natural-hazards/earthquake-hazards/faults/.
Schulte Pelkum, V., Ross, Z. E., Mueller, K. & Ben Zion, Y. Tectonic inheritance with dipping faults and deformation fabric in the brittle and ductile southern California crust. J. Geophys. Res. Solid Earth 125, e2020JB019525 (2020).
Goebel, T., Weingarten, M., Chen, X., Haffener, J. & Brodsky, E. The 2016 Mw5.1 Fairview, Oklahoma earthquakes: Evidence for long-range poroelastic triggering at >40 km from fluid disposal wells. Earth Planet. Sci. Lett. 472, 50–61 (2017).
Verdecchia, A., Cochran, E. S. & Harrington, R. M. Fluid-earthquake and earthquake-earthquake interactions in Southern Kansas, USA. J. Geophys. Res. Solid Earth 126, e2020JB020384 (2021).
Toda, S., Stein, R. S. & Sagiya, T. Evidence from the AD 2000 Izu islands earthquake swarm that stressing rate governs seismicity. Nature 419, 58–61 (2002).
Qin, Y., Chen, T., Ma, X. & Chen, X. Forecasting induced seismicity in Oklahoma using machine learning methods. Sci. Rep. 12, 9319 (2022).
Lindsey, E. & Fialko, Y. Geodetic slip rates in the southern San Andreas Fault system: effects of elastic heterogeneity and fault geometry. J. Geophys. Res. Solid Earth 118, 689–697 (2013).
Blanton, C. M., Rockwell, T. K., Gontz, A. & Kelly, J. T. Refining the spatial and temporal signatures of creep and co-seismic slip along the southern San Andreas Fault using very high resolution UAS imagery and SfM-derived topography, Coachella Valley, California. Geomorphology 357, 107064 (2020).
Jin, Z. & Fialko, Y. Finite slip models of the 2019 Ridgecrest earthquake sequence constrained by space geodetic data and aftershock locations. Bull. Seismol. Soc. Am. 110, 1660–1679 (2020).
Eissa, E. & Kazi, A. Relation between static and dynamic Young’s moduli of rocks. Int. J. Rock Mech. Min. Geomech. Abstr. 25, 479–482 (1988).
Salditch, L. et al. Earthquake supercycles and long-term fault memory. Tectonophysics 774, 228289 (2020).
Meltzner, A. J. & Wald, D. J. Aftershocks and triggered events of the great 1906 California earthquake. Bull. Seismol. Soc. Am. 93, 2160–2186 (2003).
Byerlee, J. Friction of rock. Pure Appl. Geophys. 116, 615–626 (1978).
Sibson, R. H. An assessment of field evidence for ‘Byerlee’ friction. Pure Appl. Geophys. 142, 645–662 (1994).
Fialko, Y. & Jin, Z. Simple shear origin of the cross-faults ruptured in the 2019 Ridgecrest earthquake sequence. Nat. Geosci. 14, 513–518 (2021).
Fialko, Y. Estimation of absolute stress in the hypocentral region of the 2019 Ridgecrest, California, earthquakes. J. Geophys. Res. Solid Earth 126, e2021JB022000 (2021).
Lockner, D. A., Morrow, C., Moore, D. & Hickman, S. Low strength of deep San Andreas fault gouge from SAFOD core. Nature 472, 82–85 (2011).
Mitchell, E., Fialko, Y. & Brown, K. M. Temperature dependence of frictional healing of Westerly granite: experimental observations and numerical simulations. Geochem. Geophys. Geosyst. 14, 567–582 (2013).
Mitchell, E., Fialko, Y. & Brown, K. Frictional properties of gabbro at conditions corresponding to slow slip events in subduction zones. Geochem. Geophys. Geosyst. 16, 4006–4020 (2015).
Papazafeiropoulos, G., Muñiz-Calvente, M. & Martínez-Pañeda, E. Abaqus2Matlab: a suitable tool for finite element post-processing. Adv. Eng. Softw. 105, 9–16 (2017).
Durham, W. B. Laboratory observations of the hydraulic behavior of a permeable fracture from 3800 m depth in the KTB pilot hole. J. Geophys. Res. Solid Earth 102, 18405–18416 (1997).
Miller, S. A. The role of fluids in tectonic and earthquake processes. Adv. Geophys. 54, 1–46 (2013).
Chang, K. W. & Segall, P. Injection-induced seismicity on basement faults including poroelastic stressing. J. Geophys. Res. Solid Earth 121, 2708–2726 (2016).
Ge, S. Comment on “Evidence that the 2008 Mw 7.9 Wenchuan earthquake could not have been induced by the Zipingpu Reservoir” by Kai Deng, Shiyong Zhou, Rui Wang, Russell Robinson, Cuiping Zhao, and Wanzheng Cheng. Bull. Seismol. Soc. Am. 101, 3117–3118 (2011).
Biot, M. A. General theory of three dimensional consolidation. J. Appl. Phys. 12, 155–164 (1941).
Dassault Systèmes. Abaqus 2019 (Dassault Systèmes, 2019, 2020).
Tompson, A. et al. Groundwater Availability Within the Salton Sea Basin Final Report (Lawrence Livermore National Laboratory, 2008).
Allam, A. A. & Ben-Zion, Y. Seismic velocity structures in the southern California plate-boundary environment from double-difference tomography. Geophys. J. Int. 190, 1181–1196 (2012).
Shmonov, V., Vitiovtova, V., Zharikov, A. & Grafchikov, A. Permeability of the continental crust: implications of experimental data. J. Geochem. Explor. 78-79, 697–699 (2003).
Richards-Dinger, K. B. & Shearer, P. M. Estimating crustal thickness in southern California by stacking PmP arrivals. J. Geophys. Res. Solid Earth 102, 15211–15224 (1997).
Lundgren, P. E., Hetland, A., Liu, Z. & Fielding, E. J. Southern San Andreas-San Jacinto fault system slip rates estimated from earthquake cycle models constrained by GPS and interferometric synthetic aperture radar observations. J. Geophys. Res. Solid Earth 114, B02403 (2009).
Pearse, J. & Fialko, Y. Mechanics of active magmatic intraplating in the Rio Grande Rift near Socorro, New Mexico. J. Geophys. Res. Solid Earth 115, B07413 (2010).
Johnson, K. Slip rates and off-fault deformation in Southern California inferred from GPS data and models. J. Geophys. Res. Solid Earth 118, 5643–5664 (2013).
Hampel, A., Lüke, J., Krause, T. & Hetzel, R. Finite-element modelling of glacial isostatic adjustment (GIA): use of elastic foundations at material boundaries versus the geometrically non-linear formulation. Comput. Geosci. 122, 1–14 (2019).
Brace, W. F. Permeability of crystalline and argillaceous rocks. Int. J. Rock Mech. Min. Sci. Geomech. Abstr. 17, 241–251 (1980).
Ross, Z. E., Cochran, E. S., Trugman, D. T. & Smith, J. D. 3D fault architecture controls the dynamism of earthquake swarms. Science 368, 1357–1361 (2020).
Jeppson, T. N., Bradbury, K. K. & Evans, J. P. Geophysical properties within the San Andreas Fault Zone at the San Andreas Fault Observatory at Depth and their relationships to rock properties and fault zone structure. J. Geophys. Res. Solid Earth 115, B12423 (2010).
Farr, T. & Kobrick, M. Shuttle Radar Topography Mission produces a wealth of data. Eos 81, 583–585 (2000).
Pollitz, F. F. & Sacks, I. S. The 1995 Kobe, Japan, earthquake: a long-delayed aftershock of the offshore 1944 Tonankai and 1946 Nankaido earthquakes. Bull. Seismol. Soc. Am. 87, 1–10 (1997).
Savage, J. & Burford, R. Geodetic determination of relative plate motion in central California. J. Geophys. Res. 78, 832–845 (1973).
Lin, G., Shearer, P. M. & Hauksson, E. Applying a three-dimensional velocity model, waveform cross correlation, and cluster analysis to locate southern California seismicity from 1981 to 2005. J. Geophys. Res. Solid Earth 112, B12309 (2007).
Lindsey, E. O. & Fialko, Y. Geodetic constraints on frictional properties and earthquake hazard in the Imperial Valley, Southern California. J. Geophys. Res. Solid Earth 121, 1097–1113 (2016).
Takeuchi, C. & Fialko, Y. Dynamic models of interseismic deformation and stress transfer from plate motion to continental transform faults. J. Geophys. Res. Solid Earth 117, B05403 (2012).
Turcotte, D. L. & Schubert, G. Geodynamics 2nd edn (Cambridge Univ. Press, 2002).
Acknowledgements
This work was supported by the Southern California Earthquake Center (grant 21091) to M.W. and NSF (EAR-1841273), NASA (80NSSC22K0506) and USGS (G20AP00051) to Y.F. This research benefited from correspondence with R. Guyer. This project used Quaternary fault data from the USGS. We acknowledge use of the CSRC high-performance computing cluster at San Diego State University.
Author information
Authors and Affiliations
Contributions
R.G.H. constructed the finite-element model, performed analysis of the model results and wrote the manuscript. M.W. managed the study, assisted with building the model, provided access to the modelling software, acquired funding, helped conceive the experiment and commented on the manuscript. T.K.R. carried out the palaeoseismic analysis, conceived the experiment and contributed to the manuscript. Y.F. provided advice on modelling and interpretation of the model results and contributed to the manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature thanks the anonymous reviewers for their contribution to the peer review of this work. Peer reviewer reports are available.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data figures and tables
Extended Data Fig. 1 Monte Carlo statistical test of the lake and earthquake timings.
Results of Monte Carlo statistical testing (10,000 samples) based on sampling earthquake PDF distributions and lake timings. After sampling the earthquake PDFs, we determine how many fall inside the lake timings when the lake was greater than 70% full. We compare these timings to a uniform random distribution of seven times across the same lake-loading-time range. We find that the mean timings that occur within lakes is >97% of the earthquake timings of a uniform random distribution that occur within lakes.
Extended Data Fig. 2 Numerical domain.
3D finite-element method model domain. The model mesh contains about 2 million tetrahedron elements. The light blue colour represents the extent of ancient Lake Cahuilla. The prescribed vertical load is hydrostatic, to the lake maximum water head (97.2 m). The solid red line is the SSAF fault trace. The fault zone is modelled as a slab dipping to the northeast at 60° (ref. 54), with the assumed thickness of 200 m (refs. 19,46,84).
Extended Data Fig. 3 Pore-pressure effects of lake loading.
Pore pressure (MPa) on the SSAF as a function of time (year CE) at 7 km depth for location 21, a point on the fault near the centre of the lake (see Supplementary Fig. 5). Each model is based on the variable fault permeability, with model 1 as the most permeable and model 5 as the no damage zone (Extended Data Table 2).
Extended Data Fig. 4 Analytic solution for periodic loading of a poroelastic half-space.
1D analytical model of pore pressure for a variety of different depths (blue) with surface lake-level pore pressure (black). The smaller surface profile from 1905 to the present is the Salton Sea2. Finite-element method model 2 at 7.2 km depth (green line) shows the effect of 3D diffusion with a high-permeability fault damage zone embedded in a lower-permeability host rock. The finite-element method model at 7.2 km resembles pore pressure in the 1D analytical case at 1 km, demonstrating how a fault damage zone can transmit pore pressure to depth effectively. γ = 0.1685; kfault = 1e−15 (m2); khost/1Dmodel = 1e−18 (m2).
Extended Data Fig. 6 Conceptual model of the undrained and drained effects from a lake load.
The instantaneous and transient effects of the undrained and drained effects. At t = 0, the undrained effect is felt almost instantaneously throughout the poroelastic medium beneath the lake. As time progresses, this effect attempts to equilibrate at depth. At t = 0, the drained effect is not felt except for the surface poroelastic medium and the bottom of the lake. As time progresses, this effect increases pore pressure as diffusion drives fluid from the surface down. Furthermore, as the lake load is applied, areas of compression form immediately beneath the lake, whereas areas of extension are formed near the edges.
Supplementary information
Supplementary Figures
This file contains Supplementary Figs. 1–11.
OxCal Model
Calculation of probable age ranges based on radiocarbon measurements from Rockwell et al.6.
Supplementary Video 1
Spatiotemporal pore-pressure evolution. This video describes the change in relative pore pressure across the SSAF for each time step in our model of the six lake-loading cycles of ancient Lake Cahuilla that also includes the load of the Salton Sea. The * point on the fault is the location of maximum pore pressure for average seismogenic depth of 7 km. The pore pressure at each step associated with * is plotted on the right. Two other curves are plotted that represent a ratio of what portion of the fault has positive pore pressure. The black line is for pore pressure at 7 km depth and the grey line is for pore pressure across the entire fault.
Supplementary Video 2
Spatiotemporal CFS evolution. This description is the same as for Supplementary Video 1 except for CFS instead of pore pressure.
Supplementary Video 3
Positive/negative CFS evolution. This video describes the binary positive CFS (red) versus negative CFS (blue) across the SSAF for each time step in our model. A black * plotted on the fault plane represents the location at each time step with the maximum CFS for the entire fault. Two lines are plotted on the right, which represent the positive (red) versus negative (blue) ratio for CFS on the SSAF.
Source data
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Hill, R.G., Weingarten, M., Rockwell, T.K. et al. Major southern San Andreas earthquakes modulated by lake-filling events. Nature 618, 761–766 (2023). https://doi.org/10.1038/s41586-023-06058-9
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1038/s41586-023-06058-9
Comments
By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.