ArgoPy¶

Matt Heinle¶

Argo is an international research program that deploys floats (as seen below) to collect physical and chemical data in the ocean at a variety of depths.

Image source: NOAA

ArgoPy is an effort to make data collected through the Argo network more accessible to the public as well as researchers. It contains up-to-date data from Argo floats from all around the world, as well as tools to clean and visualize the data.

Using the function below, users can see a dashboard of active Argo floats.

In [1]:

from argopy import dashboard

dashboard()

from argopy import dashboard dashboard()

Out[1]:

Install ArgoPy¶

Doumentation can be found here: https://argopy.readthedocs.io/en/latest/index.html

To install: conda install -c conda-forge argopy

In [2]:

import argopy
from argopy import DataFetcher

import argopy from argopy import DataFetcher

Data acquisition¶

One of the strengths of ArgoPy is that the DataFetcher function provides easy access to up-to-date Argo data.

In [3]:

# three methods of data selection
# by region (list: [lon_min, lon_max, lat_min, lat_max, depth_min, depth_max, date_start, date_end])
data_region = DataFetcher().region([-75, -65, 20, 30, 0, 100, '2015-01', '2015-02'])

# by World Meteorological Organization (WMO) float ID 
data_floats = DataFetcher().float([6902746, 6902757])

# by profile (WMO float ID, [list of cycles])
data_profile = DataFetcher().profile(6902746, [10, 20, 30, 40, 50])

# three methods of data selection # by region (list: [lon_min, lon_max, lat_min, lat_max, depth_min, depth_max, date_start, date_end]) data_region = DataFetcher().region([-75, -65, 20, 30, 0, 100, '2015-01', '2015-02']) # by World Meteorological Organization (WMO) float ID data_floats = DataFetcher().float([6902746, 6902757]) # by profile (WMO float ID, [list of cycles]) data_profile = DataFetcher().profile(6902746, [10, 20, 30, 40, 50])

Data visualization¶

There are three kinds of plots built into ArgoPy:

1. 'trajectory', which shows the location of the float (default)
2. 'profiler' or 'dac', which show metadata histograms
3. 'qc_altimetry', which shows quality control information for each float

In [4]:

data_region.plot()

data_region.plot()

Out[4]:

(<Figure size 900x540 with 1 Axes>,
 <GeoAxesSubplot:xlabel='longitude', ylabel='latitude'>)

In [5]:

data_floats.plot('dac')

data_floats.plot('dac')

Out[5]:

(<Figure size 900x540 with 1 Axes>, <AxesSubplot:xlabel='Number of profiles'>)

In [6]:

data_profile.plot('qc_altimetry')

data_profile.plot('qc_altimetry')

interactive(children=(Dropdown(description='Float', options=('6902746',), value='6902746'), Output()), _dom_cl…

Out[6]:

<function argopy.plot.plot.open_sat_altim_report.<locals>.f(Float)>

Compatability with other packages¶

In [7]:

import pandas as pd
from matplotlib import pyplot as plt
import xarray as xr
import cartopy
import cartopy.crs as ccrs

import pandas as pd from matplotlib import pyplot as plt import xarray as xr import cartopy import cartopy.crs as ccrs

In [8]:

# convert argo data to pandas dataframe
argo_pd = data_profile.to_dataframe()

# convert argo data to pandas dataframe argo_pd = data_profile.to_dataframe()

In [9]:

argo_pd

argo_pd

Out[9]:

	CYCLE_NUMBER	DATA_MODE	DIRECTION	PLATFORM_NUMBER	POSITION_QC	PRES	PRES_ERROR	PRES_QC	PSAL	PSAL_ERROR	PSAL_QC	TEMP	TEMP_ERROR	TEMP_QC	TIME_QC	LATITUDE	LONGITUDE	TIME
N_POINTS
0	10	D	A	6902746	1	3.0	2.4	1	35.085999	0.01	1	28.636999	0.002	1	1	19.242	-59.534	2017-08-22 06:55:00
1	10	D	A	6902746	1	4.0	2.4	1	35.084999	0.01	1	28.638000	0.002	1	1	19.242	-59.534	2017-08-22 06:55:00
2	10	D	A	6902746	1	5.0	2.4	1	35.085999	0.01	1	28.636999	0.002	1	1	19.242	-59.534	2017-08-22 06:55:00
3	10	D	A	6902746	1	6.0	2.4	1	35.087002	0.01	1	28.636000	0.002	1	1	19.242	-59.534	2017-08-22 06:55:00
4	10	D	A	6902746	1	7.0	2.4	1	35.085999	0.01	1	28.639000	0.002	1	1	19.242	-59.534	2017-08-22 06:55:00
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
534	50	D	A	6902746	1	1938.0	2.4	1	34.987999	0.01	1	3.755000	0.002	1	1	19.650	-57.767	2018-03-10 06:58:00
535	50	D	A	6902746	1	1963.0	2.4	1	34.987000	0.01	1	3.721000	0.002	1	1	19.650	-57.767	2018-03-10 06:58:00
536	50	D	A	6902746	1	1988.0	2.4	1	34.985001	0.01	1	3.672000	0.002	1	1	19.650	-57.767	2018-03-10 06:58:00
537	50	D	A	6902746	1	2013.0	2.4	1	34.983002	0.01	1	3.634000	0.002	1	1	19.650	-57.767	2018-03-10 06:58:00
538	50	D	A	6902746	1	2030.0	2.4	1	34.983002	0.01	1	3.618000	0.002	1	1	19.650	-57.767	2018-03-10 06:58:00

539 rows × 18 columns

In [11]:

# plot argo data with matplotlib

fig, ax = plt.subplots()
ax.scatter(x = argo_pd.TEMP,
           y = argo_pd.PRES,
           marker = '.')
plt.xlabel('Temperature ($^\circ$C)')
ax.xaxis.set_ticks_position('top')         # label x-axis at top of plot
ax.xaxis.set_label_position('top')
plt.ylabel('Pressure (dbar)')              # pressure is an analog for depth
plt.gca().invert_yaxis()                   # invert y-axis to reflect depth

# plot argo data with matplotlib fig, ax = plt.subplots() ax.scatter(x = argo_pd.TEMP, y = argo_pd.PRES, marker = '.') plt.xlabel('Temperature ($^\circ$C)') ax.xaxis.set_ticks_position('top') # label x-axis at top of plot ax.xaxis.set_label_position('top') plt.ylabel('Pressure (dbar)') # pressure is an analog for depth plt.gca().invert_yaxis() # invert y-axis to reflect depth

In [12]:

# convert argo data to xarray dataset
argo_xr = data_region.to_xarray()

# convert argo data to xarray dataset argo_xr = data_region.to_xarray()

In [13]:

# perform calculations
argo_xr['PSAL'].mean()

# perform calculations argo_xr['PSAL'].mean()

Out[13]:

<xarray.DataArray 'PSAL' ()>
array(36.588856, dtype=float32)

xarray.DataArray

'PSAL'

• 36.59

  array(36.588856, dtype=float32)

• Coordinates: (0)
• Indexes: (0)
• Attributes: (0)

In [37]:

# convert argo data to pandas dataframe
argo_floats = data_floats.to_dataframe()

# convert argo data to pandas dataframe argo_floats = data_floats.to_dataframe()

In [39]:

# use dataframe with cartopy to create map
plt.figure(figsize = (10, 10))
ax = plt.axes(projection = ccrs.PlateCarree())
extent = [-80, -55, 12, 25]
ax.set_extent(extent)
ax.coastlines()
ax.gridlines()

plt.scatter(x = argo_floats.LONGITUDE,
            y = argo_floats.LATITUDE,
            c = argo_floats.PRES,
            cmap = 'Blues',
            marker = '.',
            transform = ccrs.PlateCarree())
plt.colorbar(label = 'Pressure (dbar)', shrink = 0.3)

# use dataframe with cartopy to create map plt.figure(figsize = (10, 10)) ax = plt.axes(projection = ccrs.PlateCarree()) extent = [-80, -55, 12, 25] ax.set_extent(extent) ax.coastlines() ax.gridlines() plt.scatter(x = argo_floats.LONGITUDE, y = argo_floats.LATITUDE, c = argo_floats.PRES, cmap = 'Blues', marker = '.', transform = ccrs.PlateCarree()) plt.colorbar(label = 'Pressure (dbar)', shrink = 0.3)

Out[39]:

<matplotlib.colorbar.Colorbar at 0x7ff00d08f2e0>