A Bayesian approach for assessing the boundary between desirable and undesirable environmental status - An example from a coastal fish indicator in the Baltic Sea

Östman, ÖrjanSwedish University of Agricultural Sciences,Sveriges lantbruksuniversitet,Institutionen för akvatiska resurser,Department of Aquatic Resources (author)

Ecological indicator approaches typically compare the prevailing state of an ecosystem component to a reference state reflecting good environmental conditions, i.e. the desirable state. However, defining the reference state is challenging due to a wide range of uncertainties related to natural variability and measurement error in data, as well as ecological understanding. This study propose a novel probabilistic approach combining historical monitoring data and ecological understanding to estimate the uncertainty associated with the boundary value of an ecological indicator between good and poor environmental states. Bayesian inference is used to estimate the epistemic uncertainty about the true state of an indicator variable during an historical reference period. This approach replaces the traditional boundary value with probability distribution, indicating the uncertainty about the boundary between environmental states providing a transparent safety margin associated with the risk of misclassification of the indicator's state. The approach is demonstrated by applying it to a time-series of an ecological status indicator, 'Abundance of coastal key fish species', included in HELCOM's Baltic Sea regional status assessment. We suggest that acknowledgement of the uncertainty behind the final classification leads to more transparent and better-informed decision-making processes.

Olsson, JensSwedish University of Agricultural Sciences,Sveriges lantbruksuniversitet,Institutionen för akvatiska resurser,Department of Aquatic Resources(Swepub:slu)48535 (author)
Sveriges lantbruksuniversitetInstitutionen för akvatiska resurser (creator_code:org_t)
Sveriges lantbruksuniversitet

About the subject

NATURAL SCIENCES: NATURAL SCIENCES; and Other Natural Sc ...; and Other Natural Sc ...

Kungliga biblioteket hanterar dina personuppgifter i enlighet med EU:s dataskyddsförordning (2018), GDPR. Läs mer om hur det funkar här.
Så här hanterar KB dina uppgifter vid användning av denna tjänst.