Influence of Farallon Slab Loading on Intraplate Stress and Seismicity in Eastern North America in the Presence of Pre‐Existing Weak Zones

The seismic velocity model (TX2019, Lu et al., 2019) used to constrain the temperature field is freely available from IRIS at http://ds.iris.edu/ds/products/emc-tx2019slab/. The BurnMan mineral physics software (Cottaar et al., 2014; Myhill et al., 2021) used in the conversion of seismic velocities to temperature is a Python package freely available from https://github.com/geodynamics/burnman/. The scripts and files used in the velocity to temperature conversion are included within the V2T_Conversion directory in the Supporting Information data available at https://data.caltech.edu/records/wh6q5-80b36 and are described in Supporting Information S1 document. The seafloor age grid used to constrain the oceanic lithospheric temperatures is from Seton et al. (2020) and is freely available from https://earthbyte.org/webdav/ftp/earthbyte/agegrid/2020/. The thermal model used to constrain the continental lithospheric temperature field is from Artemieva (2006) and is freely available at http://www.lithosphere.info/downloads.html. Stress data is available from the World Stress Map Project at https://www.world-stress-map.org/download (Heidbach et al., 2018). Data on rift geometries (polygons) and geology used to construct the weak zone input is available from the Central Eastern United States Seismic Source Characterization for Nuclear Facilities database (http://www.ceus-ssc.com/index.htm) (CEUS-SSC-Project, 2011).

CitcomS version 3.3.1 (McNamara & Zhong, 2004; Moresi et al., 2014; Tan et al., 2006; Zhong et al., 2000) is an open-source publicly available geodynamics code maintained by Computational Infrastructure for Geodynamics and available from https://geodynamics.org/resources/citcoms or https://doi.org/10.5281/zenodo.7271920. The code has excellent strong scaling on parallel computers with up to 1,000 s of cores. Our calculations were performed on the NSF ACCESS HPC clusters Stampede2 at the Texas Advanced Computing Center (TACC) and Anvil at Purdue University using 16 nodes and 768 processors per run. Typical compute times per simulation are about 8–12 hr and are highly dependent on the viscosity structure, which can span up to six orders of magnitude. The Supporting Information S1 also includes a copy of the version of CitcomS used, as well as all pre- and post-processing scripts.

Data sets containing model inputs, outputs, and procedures, as well as scripts for making the figures presented in the paper, are available from the CaltechDATA repository (https://data.caltech.edu/records/wh6q5-80b36). These include the final 3D temperature field input to CitcomS; the 3D viscosity field; the modeled 3D velocity field; the modeled S_Hmax orientation, principal stresses, stress magnitudes, and strain rates; and fault segment data taken from sources cited within the text. See Supporting Information S1 for an explanation of each of the included datasets. Figures were constructed with either Python or GMT 6 (or PyGMT) (Wessel et al., 2019).