Contents

Surface observations

Surface observations

%load_ext autoreload
%autoreload 2

import numpy as np

import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec
from matplotlib.colors import ListedColormap
import matplotlib.path as mpath
import seaborn as sns

import cartopy.crs as ccrs

import figure_panels
import obs_surface
import util

Available records

Station information

stninfo = obs_surface.get_stn_info('CO2')
stninfo
lat lon alt stn type institution constituent
SPO_Multi_med3_CO2 -89.98 -24.8 2810 SPO aggregate Multiple CO2
SPO_Multi_median_CO2 -89.98 -24.8 2810 SPO aggregate Multiple CO2
SPO_NOAA_insitu_CO2 -89.98 -24.8 2810 SPO in situ NOAA CO2
SPO_NOAA_flask_CO2 -89.98 -24.8 2810 SPO flask NOAA CO2
SPO_SIO_O2_flask_CO2 -89.98 -24.8 2810 SPO flask SIO_O2 CO2
SPO_SIO_CDK_flask_CO2 -89.98 -24.8 2810 SPO flask SIO_CDK CO2
SPO_CSIRO_flask_CO2 -89.98 -24.8 2810 SPO flask CSIRO CO2
HBA_NOAA_flask_CO2 -75.605 -26.21 10 HBA flask NOAA CO2
SYO_NOAA_flask_CO2 -69.0125 39.59 14 SYO flask NOAA CO2
SYO_TU_insitu_CO2 -69.0125 39.59 14 SYO in situ TU/NIPR CO2
MAA_CSIRO_flask_CO2 -67.617 62.867 32 MAA flask CSIRO CO2
CYA_CSIRO_flask_CO2 -66.283 110.517 47 CYA flask CSIRO CO2
PSA_NOAA_flask_CO2 -64.9 -64 10 PSA flask NOAA CO2
PSA_SIO_O2_flask_CO2 -64.9 -64 10 PSA flask SIO_O2 CO2
DRP_NOAA_flask_CO2 -59 -64.69 10 DRP flask NOAA CO2
MQA_CSIRO_flask_CO2 -54.483 158.967 6 MQA flask CSIRO CO2
MQA_CSIRO_insitu_CO2 -54.483 158.967 6 MQA in situ CSIRO CO2
CRZ_NOAA_flask_CO2 -46.4337 51.8478 197 CRZ flask NOAA CO2
CGO_NOAA_flask_CO2 -40.683 144.69 94 CGO flask NOAA CO2
CGO_SIO_O2_flask_CO2 -40.683 144.69 94 CGO flask SIO_O2 CO2
CGO_CSIRO_flask_CO2 -40.683 144.69 94 CGO flask CSIRO CO2
CGO_CSIRO_insitu_CO2 -40.683 144.69 94 CGO in situ CSIRO CO2
AMS_LSCE_insitu_CO2 -37.7983 77.5378 55 AMS in situ LSCE CO2
CPT_SAWS_insitu_CO2 -34.3523 18.4891 230 CPT in situ SAWS CO2
BHD_NIWA_insitu_CO2 -41.4083 174.871 85 BHD NaN NIWA CO2
LMG65_NOAA_underway_CO2 -65 -63.55 10 LMG65 NaN NOAA CO2
LMG63_NOAA_underway_CO2 -63 -61.7 10 LMG63 NaN NOAA CO2
LMG61_NOAA_underway_CO2 -61 -62 10 LMG61 NaN NOAA CO2
LMG59_NOAA_underway_CO2 -59 -63.37 10 LMG59 NaN NOAA CO2
LMG57_NOAA_underway_CO2 -57 -64.23 10 LMG57 NaN NOAA CO2
LMG65_NCAR_underway_CO2 -65 -68.58 10 LMG65 NaN NCAR CO2
LMG63_NCAR_underway_CO2 -63 -68.58 10 LMG63 NaN NCAR CO2
LMG61_NCAR_underway_CO2 -61 -68.58 10 LMG61 NaN NCAR CO2
LMG59_NCAR_underway_CO2 -59 -68.58 10 LMG59 NaN NCAR CO2
LMG57_NCAR_underway_CO2 -57 -68.58 10 LMG57 NaN NCAR CO2

CO2 observations dataset

data_array = obs_surface.open_surface_data_obs('CO2')
data_array

<xarray.DataArray 'CO2' (time: 255, record: 32)>
array([[365.087 , 365.13  , 365.0911, ..., 364.925 , 364.613 ,      nan],
       [365.0233, 365.1871, 365.0878, ..., 364.9575, 364.6776,      nan],
       [364.9477, 365.042 , 364.9615, ..., 365.09  , 364.9108,      nan],
       ...,
       [408.8532, 408.9975, 408.8587, ..., 408.5017, 408.8434,      nan],
       [     nan,      nan, 408.7141, ...,      nan, 408.949 ,      nan],
       [     nan,      nan, 408.3551, ...,      nan, 408.8221,      nan]])
Coordinates:
  * time         (time) datetime64[ns] 1998-12-15 1999-01-15 ... 2020-02-14
    year_frac    (time) float64 1.999e+03 1.999e+03 ... 2.02e+03 2.02e+03
  * record       (record) object 'SPO_NOAA_insitu_CO2' ... 'CPT_SAWS_insitu_CO2'
    institution  (record) object 'NOAA' 'NOAA' 'SIO_O2' ... 'NOAA' 'LSCE' 'SAWS'
    lat          (record) float64 -89.98 -89.98 -89.98 ... -40.68 -37.8 -34.35
    lon          (record) float64 -24.8 -24.8 -24.8 -24.8 ... 144.7 77.54 18.49
    stncode      (record) object 'SPO' 'SPO' 'SPO' 'SPO' ... 'CGO' 'AMS' 'CPT'
Attributes:
    long_name:  CO$_2$
    units:      ppm

Temporal coverage

def right_bnd(y, m):
    if m == 12:
        return y+1, 1
    else:
        return y, m + 1

fig = plt.figure(figsize=(15, 8))

gsL = gridspec.GridSpec(
    nrows=1, ncols=1,
    left=0, right=0.55,
    wspace=0.05, hspace=0.01,
)
gsR = gridspec.GridSpec(
    nrows=1, ncols=1,
    left=0.55, right=1,
    wspace=0.05, hspace=0.01,
)
axs = []

records = [r for r in data_array.record.values if 'Multi' not in r and 'LMG' not in r]
da = data_array.sel(record=records)


year, mon, day = util.datetime64_parts(da.time)
year_last, mon_last = right_bnd(year[-1], mon[-1])
year = np.concatenate((year, [year_last]))
mon = np.concatenate((mon, [mon_last]))
   
time = [np.datetime64(f'{y:04d}-{m:02d}-01') for y, m in zip(year, mon)] + []


ax = fig.add_subplot(gsL[0, 0])
axs.append(ax)
cmap = ListedColormap(sns.color_palette(figure_panels.sns_palette, 1).as_hex()[::-1])
ax.pcolormesh(
    time,
    np.arange(0, len(da.record)+1, 1),
    np.where(np.isnan(da), np.nan, 1).T,
    cmap=cmap,
    shading='flat',
)
ax.set_yticks(np.arange(0, len(da.record), 1)+0.5);
ax.set_yticklabels(['_'.join(s.split('_')[:-1]) for s in da.record.values]);
ax.grid(True)
ax.grid(True, axis='x', which='minor')
ax.set_title('Southern Ocean CO$_2$ records: temporal coverage')
   
    

ax = fig.add_subplot(gsR[0, 0], projection=ccrs.SouthPolarStereo())
axs.append(ax)
ax.set_extent([180, -180, -90,  -30], crs=ccrs.PlateCarree())

# Compute a circle in axes coordinates, which we can use as a boundary
# for the map. We can pan/zoom as much as we like - the boundary will be
# permanently circular.
theta = np.linspace(0, 2*np.pi, 100)
center, radius = [0.5, 0.5], 0.5
verts = np.vstack([np.sin(theta), np.cos(theta)]).T
circle = mpath.Path(verts * radius + center)

#ax.set_boundary(circle, transform=ax.transAxes)

ax.coastlines('50m', color='gray', alpha=0.5)
ax.gridlines(alpha=0.5)
ax.set_title('Southern Ocean CO$_2$ monitoring stations')

util.label_stations(ax, stninfo.loc[records], fontsize=14, 
                    color=sns.color_palette(figure_panels.sns_palette, 2).as_hex()[0],
                   )

util.savefig('station-coverage')
_images/obs-surface_9_0.png

Comparison with shipboard data from the L. M. Gould

reference_records = obs_surface.reference_records['CO2']
reference = data_array.sel(record=reference_records).mean('record')
data_a = data_array - reference

ds = data_a.to_dataset().sel(time=slice('2012-06', '2017-11'))

ds_lmg_djf = util.ann_mean(ds, season='DJF', time_bnds_varname=None)
ds_lmg_jja = util.ann_mean(ds, season='JJA', time_bnds_varname=None)

fig, axs = util.canvas(1, 2, figsize=(6, 3.5))

figure_panels.stn_v_lat(ds_lmg_djf.CO2, axs[0, 0], include_LMG=True)
axs[0, 0].set_title(f'Surface obs: DJF CO$_2$ minus SPO (2012-2017)')
axs[0, 0].set_ylim([-0.85, 0.73])

figure_panels.stn_v_lat(ds_lmg_jja.CO2, axs[0, 1], include_LMG=True)
axs[0, 1].set_title(f'Surface obs: JJA CO$_2$ minus SPO (2012-2017)')
axs[0, 1].set_ylim([-0.85, 0.73])
util.label_plots(fig, [ax for ax in axs.ravel()], xoff=-0.02, yoff=0.04)


util.savefig('LMG-data')
_images/obs-surface_11_0.png

previous

Airborne observations

next

Analytical uncertainty at CO2 measurement stations