PAMOS (Portable Atmospheric Measurement box On Sea) is an air monitoring device especially designed for commercial vessels within the project SOOP (Shaping an Ocean Of Possibilities for science-industry collaboration). The instrument runs fully autonomously without frequent onboard maintenance to use commercial vessels for collecting atmospheric data in remote regions. This will help to close the data gaps over the oceans. The PAMOS samples meteorological data (temperature, pressure, relative humidity, wind speed and direction, precipitation), position, aerosol parameters (particle number concentration, particulate matter, black carbon), and trace gases.
To process PAMOS raw data, we use the python package shipspy [1]. It provides the commands 'pamos' and 'sections' which we use in the following. Here, we provide the settings used for shipspy. Furthermore, we provide additional scripts used for the final version of the data sets published on Pangaea [2].
The PAMOS with serial number 001 was deployed on the HX cruise vessel MS Fridtjof Nansen for one journey from Reykjavik to Greenland and back in the period 24 July - 7 August, 2024. The raw data was processed with shipspy [1]:
shipspy pamos -i 2024_nansen_greenland -o 2024_nansen_greenland.nc -a shipspy_setting/pamos_variables.yaml -c shipspy_settings/laststring_header -f shipspy_settings/pamos_quality_flags_greenland.yaml -e shipspy_settings/pamos_additional_attrs.yaml
shipspy sections -i 2024_nansen_greenland.nc -o 20240724_20240807_msfridtjofnansen_pamos.nc -s shipspy_settings/nansen_2024_greenland_sections.txt -a shipspy_settings/pamos_nansen_2024_greenland_global_attrs.yaml
Afterwards, we added the positions from the ship GPS for reference and exported an additional csv file for Pangaea. The respective script for the final processing is additional_scripts/pamos_greenland_final_cleanup.py.
The PAMOS with serial number 002 was deployed on the HX cruise vessel MS Fridtjof Nansen for the winter season 2024/2025 in the period 30 November - 26 March, 2025. During this time, the Nansen conducted eleven round trips from Ushuaia to the Antarctic Peninsula. Only data south of 60˚S is included. The raw data was processed with shipspy [1]:
shipspy pamos -i 2024_nansen_antarctica -o 2024_nansen_antarctica.nc -a shipspy_setting/pamos_variables.yaml -c shipspy_settings/laststring_header -f shipspy_settings/pamos_quality_flags_antarctica.yaml -e shipspy_settings/pamos_additional_attrs.yaml
shipspy sections -i 2024_nansen_antarctica.nc -o 20241130_20250326_msfridtjofnansen_pamos.nc -s shipspy_settings/nansen_winter2425_antarctica_sections.txt -a shipspy_settings/pamos_nansen_winter2425_antarctica_antarctica_global_attrs.yaml
Afterwards, we added the positions from the ship GPS for reference and exported an additional csv file for Pangaea. The respective script for the final processing is additional_scripts/pamos_antarctica_final_cleanup.py.
Plotting example for the whole campaign. Gray shaded areas indicate periods when the pump was not running.
import xarray as xr
import numpy as np
import matplotlib.pyplot as plt
data = xr.open_dataset("20240724_20240807_msfridtjofnansen_pamos_shipposition.nc")
pump_switch = data["pump_flag"].dropna(dim="time").diff("time") != 0
switch_times = data.time[1:].where(pump_switch, drop=True).values
fig, axs = plt.subplots(4,1,figsize = (10,9), sharex = True)
ax1 = axs[0].twinx()
ax2 = axs[1].twinx()
ax3 = axs[3].twinx()
for ax in axs:
if data["pump_flag"][0] == False:
ax.set_facecolor("gainsboro")
for i in np.arange(len(switch_times))[::2]:
ax.axvspan(switch_times[i], switch_times[i+1], color = "white")
else:
for i in np.arange(len(switch_times))[::2]:
ax.axvspan(switch_times[i], switch_times[i+1], color = "gainsboro")
plotting_dict = {"t_air": {"c": "tab:purple", "a": axs[0], "alpha" : 1},
"p_air": {"c": "tab:green", "a": ax1, "alpha" : 1},
"number_conc": {"c": "tab:brown", "a": axs[1], "alpha" : 1},
"pm10": {"c": "tab:orange", "a": ax2, "alpha" : 1},
"co2_conc": {"c": "tab:cyan", "a": axs[3], "alpha" : 1},
"co_conc": {"c": "tab:olive", "a": ax3, "alpha" : 1},
"bc_375nm": {"c": "darkviolet", "a": axs[2], "alpha" : 0.2},
"bc_470nm": {"c": "blue", "a": axs[2], "alpha" : 0.2},
"bc_528nm": {"c": "green", "a": axs[2], "alpha" : 0.2},
"bc_625nm": {"c": "red", "a": axs[2], "alpha" : 0.2},
"bc_880nm": {"c": "orange", "a": axs[2], "alpha" : 0.2}}
bc_mean = (data["bc_375nm"] + data["bc_470nm"] + data["bc_528nm"] + data["bc_625nm"] + data["bc_880nm"])/5
for v in plotting_dict.keys():
col = plotting_dict[v]["c"]
ax = plotting_dict[v]["a"]
alpha = plotting_dict[v]["alpha"]
data[v].plot(ax = ax, c = col, lw = 1, alpha = alpha, label = v)
if not ax == axs[2]:
ax.tick_params(axis = "y", labelcolor = col)
ax.set_ylabel(ax.get_ylabel(), color=col)
ax.set_xlabel("")
bc_mean.plot(ax = axs[2], c = "tab:red", lw = 1, label = "average")
axs[2].set_ylabel("black carbon conc. [kg m-3]")
axs[2].set_xlabel("")
axs[2].legend()
axs[3].set_xlabel("time")
axs[0].set_title("MS Fridtjof Nansen, Reykjavik - Greenland - Reykjavik")
plt.tight_layout()
Select first trip to Antarctic Peninsula (section = 0). Here, the meteorological variables are shown including there quality flag. Again, grey shading indicates that the pump was switched off.
import xarray as xr
import numpy as np
import matplotlib.pyplot as plt
data = xr.open_dataset("20241130_20250326_msfridtjofnansen_pamos_shipposition.nc")
data = data.groupby("section")[0]
pump_switch = data["pump_flag"].dropna(dim="time").diff("time") != 0
switch_times = data.time[1:].where(pump_switch, drop=True).values
vars_to_plot = ['quality_meteorology', 't_air', 'p_air', 'rh', 'wspd_mean', 'wdir_mean', 'precip_rate']
fig, axs = plt.subplots(7,1,figsize = (12,16), sharex = True)
for ax in axs:
if data["pump_flag"][0] == False:
ax.set_facecolor("gainsboro")
for i in np.arange(len(switch_times))[::2]:
ax.axvspan(switch_times[i], switch_times[i+1], color = "white")
else:
for i in np.arange(len(switch_times))[::2]:
ax.axvspan(switch_times[i], switch_times[i+1], color = "gainsboro")
for ax, var in zip(fig.get_axes(), vars_to_plot):
data[var].plot(ax = ax)
ax.set_xlabel("")
axs[-1].set_xlabel("time")
plt.tight_layout()
[1] Köhler, L. (2025). shipspy (Version 1.1.0) [Computer software]. https://github.com/shipspy-development/shipspy
[2] tba