(49) Analysis of the Atlantic seafloor depth/age relationship
In this example we show an example of data analysis using grids of seafloor depth and age for a region in the south Atlantic. Dumping separate x, y, z triplets with grd2xyz lets us paste the output back via gmtconvert to make binary tables of age,depth,depth. Here, depth is repeated in order to use blockmode for modal depth estimation and xyz2grd for mapping the data density. We image the density of (age,depth) points, overlay the modal depths and their robust uncertainty bars, and compute and plot two models for the expected depths as a function of age (see legend). Note we place most of the legend twice to achieve the thin-on-thick pen effect in the legend.
#!/usr/bin/env bash
# GMT EXAMPLE 49
#
# Purpose: Illustrate data analysis using the seafloor depth/age relationship
# GMT modules: blockmode, gmtmath, grdcontour, grdimage, grdsample, makecpt,
# subplot, basemap, legend, colorbar, plot, xyz2grd
#
gmt begin ex49
# Pull depth and age subsets from the global remote files
gmt grdcut @earth_relief_02m -R30W/5E/30S/5S -Gdepth_pixel.nc
gmt grdcut @earth_age_02m -R30W/5E/30S/5S -Gage_pixel.nc
# Flip to positive depths in km
gmt grdmath depth_pixel.nc NEG 1000 DIV = depth_pixel.nc
# Obtain depth, age pairs by dumping grids and pasting results
gmt grd2xyz age_pixel.nc -bof > age.bin
gmt grd2xyz depth_pixel.nc -bof > depth.bin
gmt convert -A age.bin depth.bin -bi3f -o2,5,5 -bo3f > depth-age.bin
# Obtain modal depths every ~5 Myr
gmt blockmode -R0/100/0/10 -I5/10 -r -E -Q depth-age.bin -bi3f -o0,2,3 > modal.txt
# Create density grid of (age,depth) distribution
gmt xyz2grd -R0/100/0/6.5 -I0.25/0.025 -r depth-age.bin -bi3f -An -Gdensity.nc
# Make CPTs for ages and depths
gmt makecpt -Chot -T0/100/10 -H > t.cpt
gmt makecpt -Cabyss -T0/7 -H -I > z.cpt
gmt subplot begin 2x1 -Fs15c/11.3c -Sc
# Image depth distribution, modal depths, and competing predictions
gmt grdimage density.nc -Q -Ct.cpt -JX15c/-11.3c -Bxaf+u" Myr" -Byaf+u" km" -c
# Compute Parsons & Sclater [1977] depth-age curve (in km)
# depth(t) = 0.35 * sqrt(t) + 2.500, t < 70 Myr
# = 6.4 - 3.2 exp (-t/62.8), t > 70 Myr
gmt math -T0/100/0.1 T SQRT 0.35 MUL 2.5 ADD T 70 LE MUL 6.4 T 62.8 DIV NEG EXP 3.2 MUL SUB T 70 GT MUL ADD = ps.txt
gmt plot ps.txt -W4p,green
gmt plot ps.txt -W1p
# Compute Stein & Stein [1992] depth-age curve (in km)
# depth(t) = 0.365 * sqrt(t) + 2.6, t < 20 Myr
# = 5.651 - 2.473 * exp (-0.0278*t), t > 20 Myr
gmt math -T0/100/0.1 T SQRT 0.365 MUL 2.6 ADD T 20 LE MUL 5.651 T -0.0278 MUL EXP 2.473 MUL SUB T 20 GT MUL ADD = ss.txt
# Plot curves and place the legend
gmt plot ss.txt -W4p,white
gmt plot ss.txt -W1p
gmt plot -Ss0.4c -Gblue modal.txt -Ey+p1p,blue
gmt plot -Ss0.1c -Gwhite modal.txt
gmt legend -DjRT+w5.5c+o0.25c -F+p1p+gbeige+s <<- EOF
S 0.5c - 0.9c - 4p,green 1.2c Parsons & Sclater (1977)
S 0.5c - 0.9c - 4p,white 1.2c Stein & Stein (1992)
S 0.5c s 0.4c blue - 1.2c Modal depth estimates
EOF
gmt legend -DjRT+w5.5c+o0.25c <<- EOF
S 0.5c - 0.9c - 1p 0.75c
S 0.5c - 0.9c - 1p 0.75c
S 0.5c s 0.1c white - 0.75c
EOF
# Image depths with color-coded age contours
gmt grdimage depth_pixel.nc -R30W/5E/30S/5S -JM? -Cz.cpt -c
gmt plot -W1p @ridge_49.txt
gmt grdcontour age_pixel.nc -A+f14p -Ct.cpt -Wa0.1p+c -GL30W/22S/5E/13S
gmt colorbar -Cz.cpt -DjTR+w4.7c/0.4c+h+r+o0.85c/0.35c -Baf+u" km" -F+p1p+gbeige+s+c0p/10p/4p/4p
gmt subplot end
rm -f age_pixel.nc depth_pixel.nc age.bin depth.bin depth-age.bin density.nc modal.txt ps.txt ss.txt z.cpt t.cpt
gmt end show