Table 4.

Overview of targets and key performance indicators related to the level of the data and especially to the quality of the collected data.

Nr.	Target	Description of the Target and Key Performance Indicators to Be Achieved
3.1	Perceptibility	The variation of the parameter to be measured is larger than the limit of detection of the sensor and lower than its saturation point. In addition, the resolution of the sensor is sufficiently high to observe subtle changes in the environment that are of interest to the user. For example, temperature changes in oceans of 0.001 °C are highly relevant in the study of climate change. Moreover, the sensor must be sufficiently selective so that the impact of interfering environmental parameters is sufficiently small (i.e., a limited cross-sensitivity to other parameters).
3.2	Signal-to-noise ratio	The level of the desired signal is sufficiently higher than the background noise (signal > average background level plus 3 times the standard deviation) so that small changes in the trends that are of interest to the user can be observed [63].
3.3	Sensor errors	The sensor does not generate responses that have no physical meaning such as outliers, drifts, bias, or uncertainty. Such behaviors must be identified and corrected [64,65].
3.4	Completeness	The minimum data capture and time coverage, without considering the losses of data due to regular calibration and normal maintenance of the device, should be as high as possible and preferably higher than 90% [66].
3.5	Meaningfulness	Data contain relevant information needed to answer a specific question of the user or to solve a specific problem of interest to the user.
3.6	Data structure	The structure of the data file must be sufficiently simple so that a software (e.g., Microsoft Excel) or a user can read the data. In addition to a clear structure in the organization of the data, the measurements are also supposed to be at equidistant time intervals.