While scientific research continues to enhance our understanding of the processes driving earthquakes, volcanoes, landslides, and other natural hazards, uncertainties persist about when these events will occur, their duration and magnitude, and their impact on lives and property. The editors of a new book, published by the American Geophysical Union in December 2016, on analyzing and modelling uncertainties in natural hazards answer questions about the relevance of research in this field.
With ongoing advances in Earth sciences why can we not be more certain about natural hazards?
Perhaps a better question is to ask why can we not be ‘less uncertain’ about natural hazards, acknowledging that reaching certainty is a steep climb in an ever-changing and complex world. To be sure, uncertainty is often reduced as science advances, with commensurate improvements in society’s ability to understand, predict, and prepare for these events. Ironically, however, sometimes new findings reveal there is more going on than we thought we understood, requiring a closer look under the hood. In other cases, improved computing power allows us to dig much deeper and explore model behavior and uncertainties that might have been overlooked in the past.
How does a better understanding of uncertainty feed into modeling of natural hazards?
Understanding uncertainty helps identify priority areas for research; for example, in order to best allocate research dollars, a researcher might choose to identify which input parameters in a model are least certain and could affect model outputs most greatly, and focus research efforts on reducing uncertainty in these parameters. Understanding the type of uncertainty can also help researchers choose research methods. For instance, where uncertainty is high and the probabilities of various inputs are not known, scenario planning can be a useful tool.
Why is an interdisciplinary approach so important when considering uncertainty and risk?
Hazard sources are blind to disciplinary boundaries, compelling needs for creative thinking across conventional academic problem-solving to comprehend process interactions and develop predictive tools. Also, as the editors of this book compared notes across our various disciplines (which include geosciences, forestry, volcanology, and engineering), we found that some approaches were used in more than one discipline. For instance, the extent of both wildfires and volcanic ash clouds are modeled using techniques that vary forecasted wind speed and direction, and yield a probability surface. We hope this volume fosters cross-pollination and that new applications will be found for techniques that are currently used in one discipline that would help address uncertainty in other disciplines.
How does a better understanding of uncertainty help societies plan and prepare for natural hazards?
The types of natural hazards that detrimentally affect societies are often sporadic, rare events that are difficult to predict (such as large earthquakes, wildfires, landslides, and volcanic eruptions). While the exact timing and location of these events will probably never be possible to predict, models that incorporate uncertainty can give societies an idea of the probability of events of various sizes. Where probabilities and consequences are high, societies may invest in mitigation measures, such as retrofitting buildings to resist to damage from earthquakes. It makes sense to target such investments where the potential for saving lives and other highly valued resources is greatest.
How does climate change affect uncertainty in natural hazards?
Climate change will undoubtedly affect temperature and precipitation patterns, which will directly affect the location and magnitude of mass wasting events. In addition, changes in temperature and precipitation will in turn will affect the distribution of vegetation and moisture content of dead fuels, which will affect the behavior and extent of wildfires as well as severe erosion following fires. Climate changes will depend not only on how the atmosphere, oceans, and biosphere react to increased greenhouse gas concentrations, but also on human decisions regarding emissions levels. Uncertainty regarding these factors increases the farther into the future we look.
Where are additional data or modeling efforts needed to improve natural hazard risk assessment?
In short, everywhere. In a research sense, this means broader interdisciplinary collaboration to develop measurement and assessment techniques that are reliable and portable. In a mitigation sense, it means applying these techniques across geographic locations to help raise awareness and prepare. Collectively the chapters in this book help shed light on some of these high priority areas.
Natural Hazard Uncertainty Assessment: Modeling and Decision Support, 2016, 360 pp., ISBN: 978-1-119-02786-7, list price $149.95 (hardcover), $119.99 (e-book)
—Karin Riley, Rocky Mountain Research Station, US Forest Service and College of Forestry and Conservation, University of Montana; email: email@example.com; Matt Thompson, Rocky Mountain Research Station, U.S. Forest Service; Peter Webley, Geophysical Institute, University of Alaska Fairbanks and Volcanic Ash Detection, Avoidance and Preparedness for Transportation Inc.; and Kevin Hyde, College of Forestry and Conservation, University of Montana
Riley, K., M. Thompson, P. Webley, and K. Hyde (2017), Reducing uncertainty in hazard prediction, Eos, 98, https://doi.org/10.1029/2018EO071397. Published on 28 April 2017.
Text © 2017. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.