Toward best practice framing of uncertainty in scientific publications: A review of Water Resources Research abstracts

Abstract

Uncertainty is recognized as a key issue in water resources research, among other sciences. Discussions of uncertainty typically focus on tools and techniques applied within an analysis, e.g., uncertainty quantification and model validation. But uncertainty is also addressed outside the analysis, in writing scientific publications. The language that authors use conveys their perspective of the role of uncertainty when interpreting a claim—what we call here “framing” the uncertainty. This article promotes awareness of uncertainty framing in four ways. (1) It proposes a typology of eighteen uncertainty frames, addressing five questions about uncertainty. (2) It describes the context in which uncertainty framing occurs. This is an interdisciplinary topic, involving philosophy of science, science studies, linguistics, rhetoric, and argumentation. (3) We analyze the use of uncertainty frames in a sample of 177 abstracts from the Water Resources Research journal in 2015. This helped develop and tentatively verify the typology, and provides a snapshot of current practice. (4) We make provocative recommendations to achieve a more influential, dynamic science. Current practice in uncertainty framing might be described as carefully considered incremental science. In addition to uncertainty quantification and degree of belief (present in ∼5% of abstracts), uncertainty is addressed by a combination of limiting scope, deferring to further work (∼25%) and indicating evidence is sufficient (∼40%)—or uncertainty is completely ignored (∼8%). There is a need for public debate within our discipline to decide in what context different uncertainty frames are appropriate. Uncertainty framing cannot remain a hidden practice evaluated only by lone reviewers.

Practice and perspectives in the validation of resource management models

“…Uncertainty is not among the top fifty words in this Methods topic, although it is expected to be an important concept in scenario studies. This finding can be related to the dominant uncertainty framing attitudes in scientific publications, which are shown to be expressing the findings as facts and defending the results based on a validation 43 , without an explicit discussion of uncertainty. Among the scenario-focused publications, Agriculture and Hydrology studies are distinguished from Hydrology and Climate Change studies, where the latter is defined by frequent words such as climate change, land, scenario and future.…”

A framework for characterising and evaluating the effectiveness of environmental modelling

“…What constitutes adequate treatment of uncertainty is highly problem-specific, but it is generally recognised that formal uncertainty evaluation provides indirect benefits such as increasing the depth of analysis (see Guillaume et al, 2017). Uncertainty quantification can provide better understanding of forecast accuracy (see NRC 2012), which provides for more informed evaluations of risk and reliability.…”

Uncertainty in and around biophysical modelling: insights from interdisciplinary research on agricultural digitalization

“…This includes framing and communicating such understandings effectively. Otherwise, decision-makers and other users may be less or over-confident about modelled information, which can lead to poor or detrimental outcomes [11,12,[84][85][86]. Our relational perspective on modelling uncertainties, therefore, stresses that social, biophysical and technical contexts matter.…”