Introduction
The Arctic environment presents significant challenges for risk management due to its extreme conditions and the accelerating effects of climate change. The Arctic is characterized by very low temperatures, long periods of darkness, rapidly changing weather, and unstable natural phenomena such as glaciers and avalanches (Rainville et al., 2020). Climate change in this region occurs four times faster than the global average, a phenomenon known as Arctic amplification (Rantanen et al., 2022), leading to rapid environmental transformations.
As melting sea ice opens the Arctic to increased economic activities, such as shipping and oil exploration, new safety risks arise. For instance, the opening of the Northeast Passage, or the Transarctic route, highlights the growing potential for maritime transportation (Chen et al., 2020). However, this increased activity brings a range of safety concerns that must be effectively analysed, communicated and managed. In such a high-risk and unpredictable environment, situational awareness is essential for decision-makers to anticipate and respond to evolving risks.
Situational awareness can be described as the perception, comprehension, and projection of relevant information (Endsley, 1995). Effective risk visualization is a key tool for enhancing situational awareness, and helping stakeholders navigate the uncertainty of Arctic operations. Traditional risk assessment tools, such as risk matrices, have long been used to assess safety risks in various industries (Cox, 2008). These matrices typically use a two-dimensional model, with consequences represented on the horizontal axis and probability on the vertical axis. While widely adopted for their simplicity and ease of use, these tools may present a misleading picture when applied in environments with high uncertainty, such as the Arctic (Aven, 2017).
To overcome these limitations, newer methods of risk visualisation have been developed. Examples of this are the safety-orientated bubble diagram (Abrahamsen et al, 2014). In these visualisations, not only the probability and consequences of risk are presented but also a measure of the supporting knowledge of the assessment.
The current study aims to explore how bubble diagrams can better visualise safety risks in the Arctic compared to traditional risk visualisations due to high uncertainty and a rapidly changing environment. This leads us to the following research question:
How can bubble diagrams improve risk assessment and management in Arctic environments relative to traditional risk matrices?
The first part of the paper will outline the theoretical framework. In this chapter, the theoretical basis of the paper is outlined. Following this, the concept and purpose of situational awareness are introduced. Next, the paper discusses risk visualisation and the most commonly used risk visualization tool ‘the risk matrix’ along with its strengths and limitations. Finally, an alternative method for risk visualisation is described in the form of a bubble diagram. The second part of the paper describes the methodology that is used to find relevant literature and how these sources are selected. In the third part of the paper, the selected theories are discussed and brought in relation to each other. This creates the foundation for the fourth and last part of the paper, the conclusion.
Theoretical Framework
Situational Awareness
Situational Awareness (SA) can be defined differently based on the context of the study. In this paper the SA will be defined as the “A state of working knowledge of an individual; it is how much and how accurately he/she is aware of the current situation and concerns” (Chatzimichailidou et. al. 2015 p.127). In this article, we will look at how the use of the bubble diagram can help with the increase of SA in the uncertain and rapidly changing Arctic environment. Given the SA contains the working knowledge of the individual, the preparatory work with visualising risk through the risk matrices can help with increasing the safety of the users in those conditions. This article will look further into the differences between the traditional risk matrices and the bubble diagram and explore how the usage of the bubble diagrams can help increase risk visualisation in comparison to the traditional matrix.
The theory behind situational awareness was developed by Mica Endsley in 1995. The model was originally designed for air traffic control and aircraft pilots due to the dynamically changing environment. The theory consists of a three-level model that consists of:
1. Perception of the elements in the environment.
2. Comprehension of the current situation.
3. Prediction of future status.
The first level describes the lowest level of situational awareness. It describes one’s perception of the actors involved, the environment around oneself or the available instruments. Those instruments in the arctic conditions could be the compass, GPS, satellite images of the terrain, a map or a risk matrix describing the situation. A person at the first level of situational awareness does not interpret the data, but only observes and processes the information received. The level two is constructed by the integration of the already established perception of situation and deeper understating of the information in the given terrain. This can be shown by using the map and GPS to establish the position of oneself and visualizing the distance and time needed towards the end goal. This level is argued more effective for people with a higher degree of expertise and experience in a given situation.
Level three describes the highest level of situational awareness. This level gives the person the possibility of visualizing and projecting the changes and development of the elements in the future. This level is highly dependent on the accuracy of level one and two (Stanton, et. al. 2001).
Risk visualisation
Visualizations are increasingly valuable in today’s text-heavy world, as they can communicate complex information more effectively than words alone. The best visualization is one that clearly explains the data without over-complicating it. However, adding more details can provide a more nuanced understanding, but this can also make the visualization more complex. Safety managers must strike a balance between making the data understandable without oversimplifying critical information, as emphasized by High-Reliability Organization (HRO) Theory. One of the key principles of this theory is the reluctance to simplify; oversimplifying complex risks can lead to a false sense of security and potentially disastrous outcomes (Weick & Sutcliffe, 2015).
Risk matrices
Risk matrices are one of the most common graphical visualisations of risk. Often consisting of a 3×3 or 5×5 matrix, whereby the probability of the event occurring is presented on the y-axis and the severity of the consequences is presented on the x-axis. A risk matrix provides a straightforward approach to comparing risks. The matrix is often divided into three colours, green representing acceptable risks, yellow representing risks that could potentially be accepted with the use of risk mitigating measures, and red representing risks that are unacceptable. Risk matrices require little to no experience in risk or safety science to understand. Providing an easy overview of the most pressing risks within the organisation to a wide variety of stakeholders and decision-makers (Cox, 2008).
Figure 1 Example of consequence/likelihood matrix adapted from NEK IEC 31010:2019, p. 115
The simplicity and straightforwardness of the risk matrix are also its downside. By defining risks as a combination of consequences and probabilities, the matrix uses expected values as its foundation. However, this approach to risk assessment has its limitations. For instance, a risk with a likelihood of 5 and a consequence of ‘e’ is considered equivalent to a risk with a likelihood of 1 and a consequence of ‘a’. These risks are clearly not the same (Aven, 2017). Another way to illustrate this issue is by examining the probability distribution of a particular risk. The expected value represents the centre of gravity of the probability distribution (Aven, T., & Thekdi, S. 2021). In the figure below, both risks have the same expected value, but their characteristics are very different. The ‘exponentially distributed risk’ has a high probability of low consequences but also a low probability of severe consequences, whereas the ‘normally distributed risk’ has a higher probability of medium to low consequences. This distinction is important for decision-makers and should not be overlooked.
Figure 2 Comparison of Normal and Exponential Distribution with the same Expected Values (fictitious data).
Bubble diagram
A bubble diagram is an alternative visualisation technique for presenting risk. It represents risk by presenting three dimensions: probability, consequences, and uncertainties. In the current paper, we build upon the bubble diagram presented by Abrahamsen et al. (2014).
Figure 3 Visualisation of risks through probabilities, prediction intervals for the consequences and strength-of-knowledge adapted from Abrahamsen et al, (2014).
In this diagram, the severity of the consequences is shown on the x-axis, typically reflecting a 90% confidence level or prediction interval for the consequences. This means that the assessor is 90% certain that the real consequences of the event will fall within the specified range. The y-axis represents the probability of the event occurring, where the likelihood is expressed as the assessor’s degree of belief, based on the background knowledge available.
In addition to the 90% confidence interval for the consequences, bubble diagrams incorporate uncertainty into the visualization by adjusting the size of the bubbles. The bubble size reflects the Strength of Knowledge (SoK), which shows the reliability of the information used to estimate the probability and consequences. A larger bubble indicates a lower strength of knowledge, meaning more uncertainty surrounds the risk estimates. Conversely, a smaller bubble suggests that the assessor has strong supporting data or consensus among experts, leading to higher confidence in the probability and consequence predictions.
Discussion
Safety management in the Arctic presents a unique set of challenges due to its extreme environmental conditions, including harsh weather, long periods of darkness, and remote locations. In such an environment, situational awareness is critical (Rainville et al., 2020). Situational awareness allows decision-makers to perceive and understand their surroundings and anticipate future events, enabling them to respond accordingly to emerging risks.
Risk visualisation is important for improving situational awareness. Effective risk visualizations help stakeholders grasp complex and uncertain risks. Creating effective visualizations requires striking a balance between providing enough information, without overwhelming the reader (Aven, 2017). Risk matrices, while commonly used for their simplicity and ease of interpretation, fall short in more complex environments like the Arctic (Weick & Sutcliffe, 2015). Risk matrices focus on the probability and consequences of risks but fail to account for the uncertainties inherent in unpredictable environments. As a result, risk matrices can oversimplify the risk picture, making them inadequate for Arctic safety management.
The bubble diagram provides a more nuanced approach by incorporating three dimensions: probability, consequences interval, and the Strength of Knowledge (SoK). This allows the diagram to show a level of uncertainty surrounding the assessment. By providing more detail on uncertainties, bubble diagrams improve situational awareness and offer decision-makers a clearer understanding of the risks they face (Abrahamsen et al, 2014). This is particularly important in the Arctic, where reliable data can be scarce, and risks often come with significant uncertainty.
The bubble diagram is not without its downsides. Its primary drawback is its increased complexity compared to traditional risk matrices. The inclusion of the SoK dimension adds a layer of detail that can make the visualization more challenging to interpret, especially for stakeholders unfamiliar with the risk science. Additionally, because the risks are presented in a multi-dimensional format, it can be difficult to directly compare different risks. This complexity can hinder quick decision-making in situations where time is of the essence.
In the figure below, an example of a bubble diagram is given for a field trip in Arctic conditions. In the upper right corner, we see the risk of an avalanche. The risk assessor is 80% certain that the consequences of the event will result in severe consequences (between 4 and 5). The assessor considers the strength of Knowledge for this event to be medium. And the likelihood of the event occurring to be within three and four. The second risk, hypothermia is placed in the middle of the graph and has a larger bubble, this means that the assessor considers the SoK to be low. The risk of a polar bear encounter has a very small bubble, indicating that there is strong supporting knowledge related to the probability and the consequences. The consequence interval on the other hand is wide, indicating that the assessor is not willing to be more certain than that the consequences of the event occurring extend from low consequences to high consequences.
Figure 4 Example of a bubble diagram in an Arctic safety context adapted from Abrahamsen et al, (2014).
In the Arctic, the operationalization of the Strength of Knowledge (SoK) is particularly important due to the high levels of uncertainty and the limited availability of reliable data. In practice, SoK can be assessed by examining the quality, consistency, and quantity of data available for risk assessments in specific contexts, such as weather conditions or avalanche risks.
For instance, in the case of weather data, SoK could be measured based on the agreement between different weather stations. If all stations consistently report similar conditions, this would suggest a high SoK, indicating confidence in the forecast. Conversely, if the stations show significant discrepancies, or if there are fewer data points, the SoK would be lower, indicating greater uncertainty in the weather forecast. The amount of data available and the variety of data sources (e.g., satellite observations, ground stations, and atmospheric models) would further contribute to SoK (Aven 2017). In areas with sparse weather stations or inconsistent reporting, the uncertainty would increase, affecting the confidence in risk assessments.
Similarly, for assessing avalanche danger, SoK can be determined by the number and recency of measurements taken in a specific area. High SoK would be reflected in frequent, up-to-date measurements that show consistent readings across various points in the area. In contrast, if the measurements are outdated, infrequent, or show considerable variation, the SoK would be lower, signifying less confidence in the risk assessment. This approach helps capture the degree of certainty in the available data and makes the bubble diagram a valuable tool for visualizing not only the probability and consequences of an event but also the confidence in the supporting data.
Conclusion
The Arctic environment presents a unique risk profile characterized by high levels of uncertainty and the potential for extreme consequences. Situational awareness is critical in such a dynamic and unpredictable environment, as it allows decision-makers to perceive and understand risks as they evolve. These challenges call for a more nuanced approach to risk visualization than what traditional risk matrices can provide. Bubble diagrams provide a compelling alternative, offering a more nuanced visualization that incorporates not only the probability and consequences of risks but also the strength of knowledge (SoK) (Abrahamsen et al, 2014). This approach allows for a more nuanced presentation of the risk picture.
In response to the central research question “What advantages do bubble diagrams offer over traditional risk matrices in assessing and managing Arctic safety risks?” this paper has demonstrated that bubble diagrams provide a broader and more nuanced description of the risk picture. They allow decision-makers to better understand the uncertainties associated with risks in an Arctic environment.
Based on our findings, we recommend that organizations operating in the Arctic integrate bubble diagrams into their risk management frameworks. This integration would involve training stakeholders on the interpretation of bubble diagrams and gradually transitioning from risk matrices to this more nuanced tool.
Future Research should focus on refining the operationalization of SoK within bubble diagrams, ensuring that this tool is both practical and effective in real-world Arctic operations. Further studies could explore the application of bubble diagrams in other extreme environments, contributing to a broader understanding of how this visualization can be adapted and optimized across different contexts.
References
Abrahamsen, E. B., Amundrud, Ø. A., Aven, T., & Gelyani, A. M. (2014). Safety oriented bubble diagrams vs. risk plots based on prediction intervals and strength-of-knowledge assessments: Which one to use as an alternative to risk matrices? International Journal of Business Continuity and Risk Management, 5(3), 197–213. https://doi.org/10.1504/IJBCRM.2014.066159
Albrechtsen, E., & Indreiten, M. (2021). Arctic safety. Safety Science, 137, 105165. https://doi.org/10.1016/j.ssci.2021.105165
Aven, T. (2017). Improving risk characterisations in practical situations by highlighting knowledge aspects, with applications to risk matrices. Reliability Engineering & System Safety, 167, 42-48. https://doi.org/10.1016/j.ress.2017.05.006
Aven, T., & Thekdi, S. (2021). Risk science: An introduction. Routledge. ISBN 9781000435030
Chatzimichailidou, M. M., Stanton, N. A., & Dokas, I. M. (2015). The concept of risk situation awareness provision: Towards a new approach for assessing the DSA about the threats and vulnerabilities of complex socio-technical systems. Safety Science, 79, 126–138. https://doi.org/10.1016/j.ssci.2015.05.012
Chen, J., Kang, S., Chen, C., You, Q., Du, W., Xu, M., & Chen, J. (2020). Changes in sea ice and future accessibility along the Arctic Northeast Passage. Global and Planetary
Change, 195, 103319. https://doi.org/10.1016/j.gloplacha.2020.103319
Cox, L.A. Jr. (2008). What’s wrong with risk matrices? Risk Analysis, 28, 497–512. https://doi.org/10.1111/j.1539-6924.2008.01030.x
Endsley, M. R. (1995). Toward a theory of situation awareness in dynamic systems. Human Factors, 37(1), 32–64. https://doi.org/10.1518/001872095779049543
NEK IEC. (2019). NEK IEC 31010: Risk management – Risk assessment techniques (p. 115). [Figure B.17]. Standard Norge.
Rainville, L., Wilkinson, J., Durley, M. E. J., Harper, S., DiLeo, J., Doble, M. J., Fleming, A., Forcucci, D., Graber, H., Hargrove, J. T., Haverlack, J., Hughes, N., Hembrough, B., Jeffries, M. O., Lee, C. M., Mendenhall, B., McCormmick, D., Montalvo, S., Stenseth, A., Shilling, G. B., Simmons, H. L., Toomey, J. E., & Woods, J. (2020). Improving situational awareness in the Arctic Ocean. Frontiers in Marine Science, 7, Article 581139. https://doi.org/10.3389/fmars.2020.581139
Rantanen, M., Karpechko, A. Y., Lipponen, A., Nordling, K., Hyvärinen, O., Ruosteenoja, K., & Ziese, M. (2022). The Arctic has warmed nearly four times faster than the globe since 1979. Communications Earth & Environment, 3(168), 1-8. https://doi.org/10.1038/s43247-022-00498-3
Stanton, N. A., Chambers, P. R., & Piggott, J. (2001). Situational awareness and safety. Safety Science, 39(3), 189–204. https://doi.org/10.1016/S0925-7535(01)00010-8
Weick, K. E., & Sutcliffe, K. M. (2015). Managing the unexpected: Sustained performance in a complex world. John Wiley & Sons, Incorporated.
Leave a Reply