Monday 30 April 2012

EGU 2012

In 2011 I attended the European Geophysical Union (EGU) in Vienna for the first time as a new PhD student. In all of the presentation and poster sessions for natural hazards, only a handful dealt with multiple natural hazards. And of this handful, only a few discussed or mentioned the cascading or linked effect of multiple hazards.

Kevin Fleming as manager of the MATRIX project outlaid a plan to research many issues associated with multi-hazards. This included assessing their relationships, their cascading effects and the effect of temporal scale on vulnerability in successive hazard events.

Melanie Kappes was also in attendance to present the findings of her PhD on multiple hazards affecting mountain environments.

This year at the EGU, there was a whole session devoted to ‘Multi-type hazard and risk assessment’. It shows a shift in some researchers’ thinking towards a more holistic and less isolated way of dealing with natural hazards. I was very pleased to see it producing interesting presentations and posters, and also well attended. Of particular note in the session, was Joel Gill’s presentation on ‘Reviewing and visualising interaction relationships for natural hazards’. His presentation outlined a clear way of showing the relationships between hazards, representing how they affect the likelihood and predictability of the secondary hazard and how well these relationships are understood in the literature. I look forward to his work being published as I feel it will provide a key reference in any research studying linked and cascading hazards.

However, as I discussed with some of the attendees and presenters there at the time, there were very few presentations which dealt with the interconnectedness of hazards. Every presentation in the session dealt with multiple hazards; however, very few explored the connections and cascading effects of multiple hazards. It was postulated that perhaps a more consistent way of separating these research areas was needed in the literature.

It is encouraging that multiple hazard research has become more popular and we do need more research into this area. but in this blog I will now make a differentiation between multiple hazard research (where more than one hazard is studied, but each is considered mostly separately) and what I will refer to as multi-hazard research (where connections, relationships and cascading effects between hazards are the main focus). When I come to publish my research, I will make this distinction clear, and perhaps over time with more people researching multi-hazards, the difference between them and multiple hazards will become clearer and more established.

Friday 27 April 2012

Cascading terms in the literature



The term 'cascading' hazards is used inconsistently in the literature, often in passing. It will be used in this blog to describe the phenomenon whereby one hazard triggers another, which triggers another and so on, so that the situation worsens. Cascading hazards are also referred to in the literature as the 'domino/avalanche effect', 'cascading failures', 'catastrophe/disaster chains', 'triggered events' or similar terms (Helbing and Kuhnert, 2003; Delmonaco et al, 2007).

Shi (2005) separates disaster chains into two types: simultaneous and serial chains:
  • Simultaneous chains are when multiple hazards occur in a cluster, at the same time and space, causing several disasters concurrently (Shi, 2005).
  • Serial chains or synergistic events are a succession of disaster events caused by a single hazard with the resultant disasters happening in turn (Shi, 2005; Marzocchi et al, 2009).

Delmonaco et al (2007) outline two ways of assessing cascading hazards.
The first is by examining an individual chain of events, where one event triggers another, and try to assess the probability values in order to create risk maps (Delmonaco et al, 2007). This approach is input data demanding and ‘the complexity of the hazard chains can be overwhelming’ (Delmonaco et al, 2007).

The second approach is to assess risk for coincidences of hazards, even without assuming a direct link among them (Delmonaco et al, 2007). This is a more ‘robust’ and less data demanding approach that can give a rough estimate of risk posed by cascading hazards (Delmonaco et al, 2007). However, it is questionable whether this ‘overlay’ approach actually conveys a realistic representation of cascading effects.

Why study cascading and linked multi-hazards?


At least one natural hazard affects approximately 19% of the world's landcover and affect over 50% of it's population (Dilley et al, 2005). These natural hazards often cause damage to human infrastructure, casualties, economic losses, homelessness, and many other detrimental effects. Hazard risk assessments currently exist to identify areas at risk from individual natural hazards. However, many areas of the world are at risk of more than one hazard type. According to the Hotspots report, over 790 million people are exposed to more than one natural hazard globally (Dilley et al, 2005).

Multi-hazard risk assessments (MHRA) are used to assess the risk from more than one hazard affecting an area. The approach used to assess risk from these individual hazards varies between countries and studies. Generally, the risk from individual hazards are assessed and the results summed to calculate the risk from multiple hazards. This is the simplest and least data demanding method of assessing the risk from multiple hazards. However, this approach does not account for the connections between hazards.

Hazards are not independent phenomena. They are natural processes, and as such are complex and affect each other. Often an initial hazard event will cause or trigger a subsequent natural hazard to occur. This is a cascading hazard event. The figure below demonstrates some of the existing links between natural hazards.


Cascading hazards are mentioned in the literature, but descriptions and explanations are rare. Perhaps due to the monodisciplinary of research, cascading hazards have not been investigated to great extent. Natural hazard literature tends to concentrate on specialist, isolated hazards or topics. Little research has been conducted on the broad topic of cascading hazards until recently. The importance of cascading hazard research has been highlighted in recent years because of multi-hazard events of international importance (e.g. Japan earthquake and triggered tsunami, March 2011).

Introduction to blog

I am a PhD student at Southampton University studying cascading and linked multi-hazards. This is a relatively new field of research which has recently received more attention. I set up this blog to work through my ideas related to this area of research and hopefully receive some feedback from others with similar interests or greater expertise. With it being such a new area of research, collaborating and exploring these ideas together could help us understand these phenomena better.