c program gcsv10_export.f
c fortran version of bob berner's GEOCARBSULF BASIC code
c gcsv10 stands for GeoCarbSulfVolc algorithm for 2010
c THIS CODE CAN READ EITHER THE 2006/7 proxy-CO2 data file
c OR THE 2010 DATA FILE.
c 8/17/2010 -- confirmed the VNV parameter as 4, though Berner asked for 5,
c as motivated by Aaron Taylors thesis -- but other Taylor source is vnv=3
c this code developed in MacOS 10.4, using gcc compiler,
c takes 90minutes to run on a 2006-vintage Macbook Pro with 2 GHz Intel Core Duo processor
c takes 45minutes to run on a 2009-vintage Macbook Pro with 2.53 GHz Intel Core 2 Duo processor
c -- looped over range of input parameters, set up to fit proxy data for CO2 ppm
c 11/22/06 JJP
c proxy data has uncertainties that are sometimes large. To make the data-fitting robust,
c compute chi-squared using an uncertainty in the log-domain.
c xf77 -o /Users/jjpark/bin/gcsv10_export gcsv10_export.f /Users/jjpark/Plotxy/plotlib.a
c
c code uses plotit, a standalone fortran/C plotting subroutine in Xwindows
c this code can be found on JPark's website http://earth.geology.yale.edu/~jjpark/Code/plotit.tar
c instructions for PLOTIT at http://earth.geology.yale.edu/~jjpark/Code/plotit.html
c
c code updated from the code used for the paper
c
c Royer, D. L., Berner, R. A., and Park, J., 2007, Climate sensitivity constrained by carbon
c dioxide concentrations over the last 420 million years: Nature, v.446, p.530-532.
c
c and used for this paper
c
c Park, J., and D. L. Royer, Geologic constraints on the glacial amplification of Phanerozoic
c climate sensitivity, American J. Science, V.311, 2011, P. 1-26, DOI 10.2475/01.2011.01
c
c there are many test computations in Park and Royer (2011) that were effected
c with modified versions of this gcsv10_export.f code
c
c Output of this code is written to two files
c outoutout.dat is a simple table of CO2proxy data values averaged in 10My intervals
c out_gcsv10.dat is a large ASCII file (>50 Mb) that contains the GEOCARB carbon-cycle simulations
c for all parameter choices. This allows the user to examine the statistical behavior of the
c carbon-cycle simulations offline this code (e.g. in gcsv_ppdf.f)
c
c code converted from BASIC to run grid search over parameters
c Newton-Raphson damping applied here, to avoid NaN.
c divergent transition over 380-350My is interpolated better,
c correcting earlier convergence error
c the code reads a file proxydata.txt that contains proxy CO2 PPM data
c the GEOCARBSULF carbon-cycle model is computed at time steps of 1-My from 570Ma to present
c the proxy data ranges only from 420Ma to the present
c
c Five nested loops comprise the heart of the code, from outermost to innermost
c DeltaT = temp increase with CO2 doubling
c ACT = dimensionless activation-energy for silicate weathering
c FERT = plant CO2-fertilization coefficient
c LIFE = liverwort-based weathering factor -- less than vascular plants
c GYM = gymnosperm-based weathering factor (1=angiosperm weathering parameter)
c gymnosperms (e.g. conifers) might be better at weathering than angiosperms
c
c for each choice of parameters, we compute the carbon flux balance between terms
c that depend on silicate weathering (fBBS) and carbon burial/degassing (fB),
c each normalized with common factors derived by Bob Berner many years ago
c The carbon-cycle balance depends on CO2 level,
c because the weathering of silicates accelerates with warmer temps induced by greenhouse gases
c
c the code solves for the CO2 level that achieves the balance of carbon fluxes
c The CO2-dependent parameter fB is calculated at each time step from carbon and
c carbon isotope mass balance and values of all other parameters, based on
c geological and biological data, that affect the carbon cycle. Then the value of
c RCO2 (CO2 concentration normalized to the mean value for the past 1 million
c years = 250 ppm) is calculated by inversion from a complex expression for fB based on the
c greenhouse effect, CO2 fertilization of plant weathering, solar evolution, and
c changes in land temperature).
c
c for detailed exposition of the theoretical GEOCARB models, see
c
c Berner, R. A., The Phanerozoic Carbon Cycle: CO2 and O2, 150pp. Oxford University Press, 2004.
c
c and a collection of GEOCARB and GEOCARBSULF papers published over the past decade by
c Berner and Colleagues.
c
c Berner, R. A., 2006a, GEOCARBSULF: A combined model for Phanerozoic atmospheric O2 and CO2:
c Geochimica et Cosmochimica Acta, v.70, p.5653-5664.
c
c Berner, R. A., 2006b, Inclusion of the weathering of volcanic rocks in the GEOCARBSULF model:
c American Journal of Science, v.306, p.295-302, doi:10.2475/052006.01.
c
c Berner, R. A., 2008, Addendum to Inclusion of the weathering of volcanic rocks in the
c GEOCARBSULF model (Berner, R. A., 2006, v.306, p.295-302) American Journal of Science,
c v.308, p.100-103, doi:10.2475/01.2008.04.
c
c Berner, R. A., 2009, Phanerozoic atmospheric oxygen: New results using the GEOCARBSULF Model:
c American Journal of Science, v.309, p.603-606, doi:10.2475/07.2009.03.
c
c The code computes the aggregate data fit of GEOCARBSULF for each choice
c of the combined parameters, expressed as the fractional variance residual when expressed
c in the log domain without uncertainty weighting.
c description of the proxy data can be found in
c
c Royer et al, CO2 as a primary driver of Phanerozoic climate, GSA Today, March 2004.
c
