Auxiliary material for Paper 2007JB005460 Stresses that drive the plates from below: Definitions, computational path, model optimization, and error analysis Peter Bird, Zhen Liu, and William Kurt Rucker Department of Earth and Space Sciences, University of California, Los Angeles, California, USA Bird, P., Z. Liu, and W. K. Rucker (2008), Stresses that drive the plates from below: Definitions, computational path, model optimization, and error analysis, J. Geophys. Res., 113, B11406, doi:10.1029/2007JB005460. Introduction This auxiliary material provides the 9 essential files, and detailed instructions, to permit reproduction of the preferred global-lithosphere neotectonic dynamic model Earth5-049, which was presented in the text, tables, and figures of the paper. The necessary steps would include: (1) Strip the extra filename extension ".txt" from all 9 files provided. (2) Compile SHELLS.f90 with a Fortran90 compiler. -Note that source code is in (old) fixed-form, rather than (new) free-form, and set compiler switches accordingly. -Note that standard IMPLICIT variable-typing is assumed, and set compiler switches accordingly. -Note that the source statement: USE Numerical_Libraries is used to link to subprogram DLSLRB (which solves linear systems of equations). DLSLRB is part of the I.M.S.L. (International Mathematics Subroutine Library); my INTEGER FUNCTION INDEXK is customized to work with IMSL's storage convention for banded nonsymmetric real matrices. If a different solver is substituted then INDEXK will need to be modified. Also, subprogram KSIZE, which checks whether the size of PARAMETER MAXSIZ is adequate, would also change. [N.B. If you wish to run Shells under 32-bit Windows, you may not need to compile it yourself. Obtain SHELLS.exe from: http://peterbird.name/oldFTP/neotec/SHELLS/ ] (3) At this point, decide whether you will reproduce the entire computation, (consisting of several runs of SHELLS, while waiting for the basal-strength torques to converge), or just the final run. To reproduce only the final run, jump to instruction (7) below. (4) Run SHELLS (on a computer with 1.5 GB RAM) to produce model Earth5-046: -When prompted for name for the logfile, enter "Earth5-046.log" (or any other name that you prefer). -When prompted for the name of the finite-element grid, enter "Earth5R.feg". -When prompted for the name of the input parameter file, enter "iEarth5-046.in". -When prompted for outlines of the PLATES, enter "PB2002_plates.dig" (which was published as part of the PB2002 plate model of Bird, 2003, Geosphere). (The leads to the message "Assigning each node to a plate..." which is a slow process.) -When prompted for the name of an old velocity solution, enter "X" as you don't have one. -When prompted for the name of the boundary conditions,enter "Earth5R-type4AplusA.bcs". -When prompted for the BOUNDARIES of the plates, enter "PB2002_boundaries.dig" (which was published as part of the PB2002 plate model of Bird, 2003, Geosphere). -When prompted to name the new nodal velocity file, enter "vEarth5-046.out". -When prompted to name the new nodal forces file, enter "fEarth5-046.out". -When prompted to name the new torque balances file, enter "qEarth5-046.out". (At this point, the run may be left unattended.) (5) Run SHELLS (on a computer with 1.5 GB RAM) to produce model Earth5-047: -When prompted for name for the logfile, enter "Earth5-047.log" (or any other name that you prefer). -When prompted for the name of the finite-element grid, enter "Earth5R.feg". -When prompted for the name of the input parameter file, enter "iEarth5-047.in". -When prompted for outlines of the PLATES, enter "PB2002_plates.dig" (which was published as part of the PB2002 plate model of Bird, 2003, Geosphere). (The leads to the message "Assigning each node to a plate..." which is a slow process.) -When prompted for the name of an old torque-report file, enter "qEarth5-046.out". -When prompted for the name of an old velocity solution, enter "vEarth5-046.out", or "X". -When prompted for the name of the boundary conditions,enter "Earth5R-type4AplusA.bcs". -When prompted for the BOUNDARIES of the plates, enter "PB2002_boundaries.dig" (which was published as part of the PB2002 plate model of Bird, 2003, Geosphere). -When prompted to name the new nodal velocity file, enter "vEarth5-047.out". -When prompted to name the new nodal forces file, enter "fEarth5-047.out". -When prompted to name the new torque balances file, enter "qEarth5-047.out". (At this point, the run may be left unattended.) (6) Run SHELLS (on a computer with 1.5 GB RAM) to produce model Earth5-048: -When prompted for name for the logfile, enter "Earth5-048.log" (or any other name that you prefer). -When prompted for the name of the finite-element grid, enter "Earth5R.feg". -When prompted for the name of the input parameter file, enter "iEarth5-048.in". -When prompted for outlines of the PLATES, enter "PB2002_plates.dig" (which was published as part of the PB2002 plate model of Bird, 2003, Geosphere). (The leads to the message "Assigning each node to a plate..." which is a slow process.) -When prompted for the name of an old torque-report file, enter "qEarth5-047.out". -When prompted for the name of an old velocity solution, enter "vEarth5-047.out", or "X". -When prompted for the name of the boundary conditions,enter "Earth5R-type4AplusA.bcs". -When prompted for the BOUNDARIES of the plates, enter "PB2002_boundaries.dig" (which was published as part of the PB2002 plate model of Bird, 2003, Geosphere). -When prompted to name the new nodal velocity file, enter "vEarth5-048.out". -When prompted to name the new nodal forces file, enter "fEarth5-048.out". -When prompted to name the new torque balances file, enter "qEarth5-048.out". (At this point, the run may be left unattended.) (7) Run SHELLS (on a computer with 1.5 GB RAM) to produce model Earth5-049: -When prompted for name for the logfile, enter "Earth5-049.log" (or any other name that you prefer). -When prompted for the name of the finite-element grid, enter "Earth5R.feg". -When prompted for the name of the input parameter file, enter "iEarth5-049.in". -When prompted for outlines of the PLATES, enter "PB2002_plates.dig" (which was published as part of the PB2002 plate model of Bird, 2003, Geosphere). (The leads to the message "Assigning each node to a plate..." which is a slow process.) -When prompted for the name of an old torque-report file, enter "qEarth5-048.out". -When prompted for the name of an old velocity solution, enter "vEarth5-048.out", or "X". -When prompted for the name of the boundary conditions,enter "Earth5R-type4A.bcs". <***CHANGE*** -When prompted for the BOUNDARIES of the plates, enter "PB2002_boundaries.dig" (which was published as part of the PB2002 plate model of Bird, 2003, Geosphere). -When prompted to name the new nodal velocity file, enter "vEarth5-049.out". -When prompted to name the new nodal forces file, enter "fEarth5-049.out". -When prompted to name the new torque balances file, enter "qEarth5-049.out". (At this point, the run may be left unattended.) (8) Obtain our graphical post-processing program FiniteMap (and associated files) from: http://peterbird.name/oldFTP/FiniteMap if you wish to produce graphical maps like those seen in the figures of the paper. (9) Obtain our scoring post-processing program OrbScore2 from: http://peterbird.name/oldFTP/neotec/SHELLS/OrbScore/ if you wish to reproduce the scoring of the model, as reported in the paper. (But note that scoring datasets were not authorized for publication, and thus are not included here.) 1. 2007jb005460-SHELLS.f90.txt This is the Fortran90 source code for thin-shell dynamic neotectonic finite-element program SHELLS. Version of 29 August 2006, including important new features added for this paper. 2. 2007jb005460-Earth5R.feg.txt This is a finite-element grid representing the global lithosphere, including its faults which divide it into plates (and subdivide some plates into blocks within orogens). For each node, the values given in SI units are: longitude (degrees East), latitude (degrees North), elevation, heat-flow, crustal thickness, mantle lithosphere thickness, density anomaly of compositional origin, & geothermal curvature due to transient cooling. To edit this grid (or simply to visualize it) you will want to use our program OrbWin, from: http://peterbird.name/oldFTP/OrbWin/ 3. 2007jb005460-iEarth5-046.in.txt This is an input-parameter file, corresponding to Table 1 in the paper.\ (Also note that Table 2 of the paper shows how we determined optimal parameters.) This is the first of a group of 4 input files needed to reproduce Earth5-049. Most lines in this group of files do not change; however, note changes in: -line 1: title line (changed to reflect number of this run: ..."-046") -line 14: value of parameter ICONVE (= 3, but irrelevant due to following...) -line 15: value of parameter TRHMAX (=0, preventing any distributed basal traction) 4. 2007jb005460-iEarth5-047.in.txt This is an input-parameter file, corresponding to Table 1 in the paper.\ (Also note that Table 2 of the paper shows how we determined optimal parameters.) This is the second of a group of 4 input files needed to reproduce Earth5-049. Most lines in this group of files do not change; however, note changes in: -line 1: title line (changed to reflect number of this run: ..."-047") -line 14: value of parameter ICONVE (= 6; use inferred basal tractions) -line 15: value of parameter TRHMAX (= 20.E6; limit raised to 20 MPa) 5. 2007jb005460-iEarth5-048.in.txt This is an input-parameter file, corresponding to Table 1 in the paper.\ (Also note that Table 2 of the paper shows how we determined optimal parameters.) This is the third of a group of 4 input files needed to reproduce Earth5-049. Most lines in this group of files do not change; however, note changes in: -line 1: title line (changed to reflect number of this run: ..."-048") -line 14: value of parameter ICONVE (= 6; use inferred basal tractions) -line 15: value of parameter TRHMAX (= 20.E6; limit raised to 20 MPa) 6. 2007jb005460-iEarth5-049.in.txt This is an input-parameter file, corresponding to Table 1 in the paper.\ (Also note that Table 2 of the paper shows how we determined optimal parameters.) This is the last of a group of 4 input files needed to reproduce Earth5-049. Most lines in this group of files do not change; however, note changes in: -line 1: title line (changed to reflect number of this run: ..."-049") -line 14: value of parameter ICONVE (= 6; use inferred basal tractions) -line 15: value of parameter TRHMAX (= 20.E6; limit raised to 20 MPa) 7. 2007jb005460-Earth5R-type4AplusA.bcs.txt This is the first of two versions of the boundary-conditions file appropriate for use with finite element grid Earth5R.feg. This version includes "extra" velocity boundary conditions in plate interiors, and is used for steps (4), (5), (6). 8. 2007jb005460-Earth5R-type4A.bcs.txt This is the second of two versions of the boundary-conditions file appropriate for use with finite element grid Earth5R.feg. This version does not include any "extra" velocity boundary conditions, and is used only in the last step (7). 9. 2007jb005460-qEarth5-048.out.txt This is a "torque report" file with content similar to Table 4 (and Figures 15-16) in the paper. It will be needed if you jump directly to step (7), because normally it would be produced by running step (6).