During the MOSAiC expedition 2019-2020 atmospheric thermodynamic profile measurements have been conducted from a meteorological (Met) Tower on the sea ice, as well as via collocated radiosondes that were launched approximately every six hours from aboard Polarstern. While the radiosondes lack the lowermost 10 m above the sea ice, the Met Tower profile can be used to fill this gap (observations at 0, 2, 6 and 10 meters). This is a blended data product that merges the Met Tower profile (data version 3.4, doi:10.18739/A2PV6B83F) in the minute of the radiosonde launch with the radiosonde profile aloft (data version 3, doi:10.1594/PANGAEA.943870). Parameters included are temperature (T), relative humidity (RH), wind speed and -direction, and air pressure. The aim of this product is two-fold: (1) To provide comprehensive atmospheric profiles for each radiosonde launch, that additionally retain the lowermost meters of the atmospheric boundary layer above the sea ice and (2) to remove potential unrealistic T/RH values from the radiosonde profiles that can emerge in the lowermost 100 m due to the influence of the ship on the measurement. Examples for the latter are occasional warm anomalies due to the heat island effect of the ship, or elevated, vertically confined peaks that can arise from the ship's exhaust plume. The potential effect of the exhaust plume on the T profile is estimated by comparing the radiosonde at 30 m height to the concurring Polarstern meteorological observation (doi:10.1594/PANGAEA.935263 - doi:10.1594/PANGAEA.935267). Given the geometrical constellation of the Polarstern observation towards the bow of the ship and the sounding launch platform at the aft of the ship, and depending on the wind direction relative to the ship, it can be assumed that at least one of the T measurements is less impacted from the ship exhaust than the other, and is retained. In a next step, the 10 - 30 m height segment in T and RH is filled with a linear interpolation between the Met Tower at 10 m and the radiosonde observation at 30 m. When identified, remaining T/RH peaks in the lowermost 100 m of the profile are removed and filled with a linear interpolation from below to above the peak. T/RH flags are provided to indicate where the profiles have been manipulated from the original data, and to indicate the reason for missing data in the profile. Compared to the original profiles, this blended product adds value and quality control in the lowest 100 m, which makes it better suitable, for example, for boundary layer analyses.

This dataset is a merged data set from the MOSAiC 20192020 expedition, in which two concurrent atmospheric profile measurements are blended together. One is collected from a Met Tower and covers the 0-10 m altitude range, one is a radiosonde data set which covers the altitude range above 10 m . Both base data sets are already published and referenced to. The added value of the blended data set is that it combines the two to generate one comprehensive profile for each radiosonde launch, that adds additional information in the 0 - 10 m height range, and occasional unrealistic radiosonde values in the 10 - 100 m height range have been detected and replaced by improved estimates. This makes this blended data set better suitable for applications that rely on the lowermost 100 m of the atmospheric profiles (e.g., for boundary layer studies).