Naldmann et al. 
from different sources can occur because of improper definition 
and inadequate uncertainty assignment. 
Outloo! 
The presented approach to quantify uncertainties of 
measured EOVs in a practical manner has been presented on 
che basis of temperature measurements. The parameter can be 
directly measured using a single datum that can immediately be 
zalibrated with a temperature standard. Moreover, temperature 
sensors show good long-term stability. The collected time series 
have shown that measurement results of all sensors are matching 
well even after some months. Other EOVs are however more 
challenging. Their numerical value has to be calculated from 
Jifferent parameters, all having their own uncertainties. Salinity, 
for instance, is derived from temperature, conductivity and 
pressure measurements. The respective calibration procedure 
is relatively elaborate and due to the fact that correlations 
between involved parameters exist, the uncertainty calculations 
are not straightforward. Moreover, the stability of the sensors is 
strongly affected by environmental effects, ie., biofouling. In a 
subsequent paper the collected salinity data of the measurement 
series will discuss the effect of stability issues and multi- 
parameter measurements on the uncertainty quantification of 
measured ocean variables. 
Other contributions to uncertainty quantifications have to 
ve considered as well. One example would be the pressure 
sensitivity of temperature in profiling observations, such as 
shipboard CTD and Deep ARGO observations (Uchida et al., 
2015)). Checking the time drift of temperature sensors in 
profiling float observations (Oka, 2005), would be another 
important topic, as profiling floats are not usually recovered 
and therefore post-observation calibrations for the temperature 
sensors are not possible. 
Similar initiatives to enable the quantification of 
ancertainties had been started like the US CLIVAR Working 
group on Ocean Uncertainty Quantification (US-CLIVAR, 
2020) and the International Quality-Controlled Ocean 
Database (Cowley, 2021). In the publication of Cowley et al. 
an additional aspect is mentioned that is described as the 
“Representativeness Errors”. This aspect is putting the 
uncertainty assessment in the framework of what processes 
shall be observed and what type of fluctuations can be 
expected. Because here assumptions have to be made that are 
-rontiers in Marine Science 
IE 
10.3389/fmars.2022.1002153 
telated to the used models this contribution to the uncertainty 
will probably change over time. 
With this study a contribution to the UN Decade for Ocean 
Sciences shall be made. It will be a unique opportunity to use 
established platforms like the IODE Ocean Best Practice System 
(IODE, 2022) to disseminate the ideas and methods developed 
here. A close interaction with expert groups within WMO is 
already ongoing and will provide an additional impulse bridging 
the existing gap between ocean and meteorological practices. 
Data availability statemer.t 
The datasets presented in this study can be found in online 
repositories. The names of the repository/repositories and 
accession number(s) can be found below: PANGAEA - https:// 
doi.,pangaea.de/10.1594/PANGAEBA.942643. 
Author contributions 
CW, PF, and HB conceptualized the experiment, SS was 
overseeing and contributed to the systematic application of 
metrological principles, MK and J-GF did the oceanographic 
evaluation of the complete data set, MB designed the 
experimental set-up and evaluated the sensor performance 
during the lifetime of the study, SW carried out all calibration 
tasks. RH was contributing to the data management description. 
CW wrote an initial draft of this manuscript. All authors 
contributed to the article and approved the submitted version. 
Fundirng 
The project named DAUNE was carried out using existing 
junding resources of all involved institutions. 
Acknowledgments 
The authors very much acknowledge the support by the AWI 
Diving Team located on the island of Helgoland who deployed and 
safely retrieved all instruments, did the maintenance and repair and 
handled all sensor data. We also thank Kai Herklotz from BSH, 
Germany, for his support and valuable input. 
rrontiersin.org