Figure with multiple AXES, set up manual!

Example from a longitudinal survey in a estuary with vertical profiles of CTD

For further information, check

import numpy as np
import matplotlib.pyplot as plt
from scipy.interpolate import griddata

# load pre processed data
diretorio = r'd:\GUTO\1_Trabs\1_Aestus\Itamaraca\Campanha20040715\rotinas\\'
arquivo = 'ctd_longitudinal_rdz.txt'

d = np.loadtxt(diretorio + arquivo)

# variables in 'd'
# dist level sal temp ox ox turb

# separating the variables to make it clearer...

distancia = d[:,0]
nivel = d[:,1]
salinidade = d[:,2]
temperatura = d[:,3]
oxigenio_sat = d[:,4]
turbidez = d[:,6]

# correcting some oxigen values
idx = np.where(oxigenio_sat>100)
oxigenio_sat[idx]=100

# to draw the bottom 'mask'
dist_ref = -100
distancias = []

# the distance along the survey
for di in distancia:
    if di > dist_ref:
        distancias.append(di)
        dist_ref = di

distancias2 = np.copy(distancias)        

# to find the maximum depth of every CTD profile
profundidades = []
for di in distancias:
    p_prof = np.min(nivel[distancia == di])
    profundidades.append(p_prof)

# closing the polygon in a not elegant fashion
distancias.append(distancias[-1])
profundidades.append(np.min(profundidades)-.5)

distancias.append(distancias[0])
profundidades.append(np.min(profundidades))

distancias.append(distancias[0])
profundidades.append(profundidades[0])

plt.fill(distancias, profundidades, 'b')

[<matplotlib.patches.Polygon at 0x2359414c760>]

Interpolating

points = d[:,0:2]

xi = np.linspace(np.min(distancia), np.max(distancia), 50)
zi = np.linspace(np.min(nivel), np.max(nivel), 20)

xx, zz = np.meshgrid(xi, zi)

salinidade_i = griddata(points, salinidade, (xx, zz), method='linear')
temperatura_i = griddata(points, temperatura, (xx, zz), method='linear')
oxigenio_sat_i = griddata(points, oxigenio_sat, (xx, zz), method='linear')
turbidez_i = griddata(points, turbidez, (xx, zz), method='linear')

# stretch the zi a little for aesthetics purposes
zi_2 = np.linspace(np.min(nivel)-1, np.max(nivel)+.5, 20)
xx, zz_2 = np.meshgrid(xi, zi_2)

Assembling the figure with multiple axes

Despite more laborious than just using the grid spacing of the subplots, this allow better positioning

For example, the positioning of the colorbars as quite anoying... here we can put whereever we want

plt.rcParams.update({'font.size': 14})
fig = plt.figure(figsize=(6,7))

px = .1
py = .75
dx = .8
dy = .23
dint = .02

ax1 = fig.add_axes([px, py, dx, dy])
ax2 = fig.add_axes([px, py-(dy+dint), dx, dy])
ax3 = fig.add_axes([px, py-(dy+dint)*2, dx, dy])
ax4 = fig.add_axes([px, py-(dy+dint)*3, dx, dy])

pxcb = .91
dxcb = .01
fc = .02

cbax1 = fig.add_axes([pxcb, py+fc, dxcb, dy-fc*2])
cbax2 = fig.add_axes([pxcb, py-(dy+dint)+fc, dxcb, dy-fc*2])
cbax3 = fig.add_axes([pxcb, py-(dy+dint)*2+fc, dxcb, dy-fc*2])
cbax4 = fig.add_axes([pxcb, py-(dy+dint)*3+fc, dxcb, dy-fc*2])


c1 = ax1.contourf(xx, zz_2, salinidade_i, cmap='rainbow')
ax1.fill(distancias, profundidades, color=(.7, .7, .7))
ax1.plot(distancia, nivel, '.k', markersize=2)
ax1.set_ylim([-8.5, 0])
ax1.set_xlim([-2.75, 6])
ax1.set_yticks([-6, -3, 0])
ax1.set_ylabel('Depth (m)')
ax1.set_xticklabels('')

cbax1 = plt.colorbar(c1, cax=cbax1, ticks=[0, 10, 20, 30])


c2 = ax2.contourf(xx, zz_2, temperatura_i, cmap='rainbow')
ax2.fill(distancias, profundidades, color=(.7, .7, .7))
ax2.set_ylim([-8.5, 0])
ax2.set_xlim([-2.75, 6])
ax2.set_yticks([-6, -3, 0])
ax2.set_ylabel('Depth (m)')
ax2.set_xticklabels('')

cbax2 = plt.colorbar(c2, cax=cbax2, ticks=[25, 25.5, 26, 26.5])

c3 = ax3.contourf(xx, zz_2, oxigenio_sat_i, cmap='rainbow')
ax3.fill(distancias, profundidades, color=(.7, .7, .7))
ax3.set_ylim([-8.5, 0])
ax3.set_xlim([-2.75, 6])
ax3.set_yticks([-6, -3, 0])
ax3.set_ylabel('Depth (m)')
ax3.set_xticklabels('')

cbax3 = plt.colorbar(c3, cax=cbax3, ticks=[60, 80, 100])

c4 = ax4.contourf(xx, zz_2, turbidez_i, cmap='rainbow')
ax4.fill(distancias, profundidades, color=(.7, .7, .7))
ax4.set_ylim([-8.5, 0])
ax4.set_xlim([-2.75, 6])
ax4.set_yticks([-6, -3, 0])
ax4.set_ylabel('Depth (m)')
ax4.set_xlabel('Distance (km)')

cbax4 = plt.colorbar(c4, cax=cbax4, ticks=[0, 15, 30])

ax1.text(-4, -.3, '(a)')
ax2.text(-4, -.3, '(b)')
ax3.text(-4, -.3, '(c)')
ax4.text(-4, -.3, '(d)')

ax1.text(-2.5, -8, 'Salinity (psu)')
ax2.text(-2.5, -8, 'Temperature (°C)')
ax3.text(-2.5, -8, 'Oxygen (%)')
ax4.text(-2.5, -8, 'Turbidity (FTU)')

ax1.text(-2.5, .5, 'Shelf')
ax1.text(-.2, .5, '|')
ax1.text(.5, .5, 'Channel')
ax1.text(2.5, .5, '|')
ax1.text(3.5, .5, 'Estuary')

plt.show()