A methodology to uncertainty quantification of essential ocean variables

Abstract

The goal of this study is to provide a universally applicable procedure for a systematic evaluation of in situ measured data from single sensors regarding quantifying the uncertainty of the measurement results. As determining uncertainty for an environmental parameter also depends on the parameter itself, the focus here will be set on the variable water temperature in the first place. A separate analysis for salinity and other data will follow in later publications. With this first of a series of planned manuscripts on different parameters, we aim at providing a common understanding of how measurement uncertainty on single sensor measurements can be derived. Using an experimental in situ set-up with 6 different standard CTD sensors of two different brands, we created a four month-long, high-quality data set to be used to develop a reliable method for quantifying measurement uncertainties. Although the CTDs were deployed in a mooring in a coastal environment the described method can be extended to other deployment configurations as well. The described procedures have evolved as a stepwise process that takes the different perspectives of the involved authors into account, as well as the special conditions for environmental measurements, which are collected while the observed volume/area is undergoing a constant change. By sharing the ideas with other stakeholders, the basic concept can be extended to other observing programs and to other essential ocean variables.

Kelp forest community structure and demography in Kongsfjorden (Svalbard) across 25 years of Arctic warming

“…PAR and turbidity data were quality controlled according to an adapted ARGO standard for stationary sensors (Fischer et al., 2021 ; Waldmann et al., 2022 ). To analyze long‐term changes and trends over the sampling period, the residuals of the observed weekly PAR and turbidity values to the expected PAR and turbidity values were calculated, using the means of the observed weeks across all years as expected values.…”

“…Homogeneity of variances (Levene's test) and normal distribution (Shapiro-Wilk test) could not be achieved through data transformation. The effect of Species and Depth on the individual parameters were tested in separate non-parametric Kruskal-Wallis tests and by pairwise Wilcoxon rank sum tests with Bonferroni correction to reveal differences between species.2.8.4 | Turbidity and PAR time seriesPAR and turbidity data were quality controlled according to an adapted ARGO standard for stationary sensors(Fischer et al, 2021;Waldmann et al, 2022). To analyze long-term changes and trends over the sampling period, the residuals of the observed weekly PAR and turbidity values to the expected PAR and turbidity values were calculated, using the means of the observed weeks across all years as expected values.…”

Challenges, limitations, and measurement strategies to ensure data quality in deep-sea sensors

“…Many of the manually observed "dips" in measured oxygen (Obs4) also result in a difference between the redundant measurements above the set threshold. Similar effects are observed for Drake 137. systems is detailled in (BIPM, 2008), and a very simplified application of this method to in situ temperature measurements is proposed by (Waldmann et al, 2022).…”

Arctic kelp forest decline -- a consequence of melting glaciers?

“…PAR and Turbidity time series: PAR and turbidity data were quality controlled according to an adapted ARGO standard for stationary sensors (Fischer et al, 2021;Waldmann et al, 2022). To analyze long-term changes and trends over the sampling period, the residuals of the observed weekly PAR and turbidity values to the expected PAR and turbidity values were calculated, using the means of the observed weeks across all years as expected values.…”