Disaster Media Coverage

« The media is an integral part of the disaster risk management process. Radio, television and the internet are amongst the fastest channels for the transmission of knowledge and information across wide areas. These means of sharing information can also reach a large number of people from different backgrounds and engaged across a variety of sectors. Media has an important function to inform, educate and empower communities with relevant knowledge to influence public action and policy towards disaster preparedness and mitigation. The media has a huge power to influence national and global public opinion, giving visibility to disaster-related issues. The media can make a real difference in the way people think about and act on disasters. » Information Kit for Media for Disaster Risk Reduction, Scaling-up Community-Based Disaster Risk Reduction in Lao PDR


With the Sendai framework for Disaster Risk Reduction (DRR) adopted in 2015, the international community of Disaster Risk Reduction has vowed to push for greater media involvement in disaster risk reduction, given their crucial role as a conduit of information to the public. But before to ask for a greater media involvement in DRR, it would be certainly useful to understand how media usually process the information on disasters. Only such analysis can help deriving appropriate guidelines for the media themselves and for the DRR stakeholders (e.g. national authorities or scientists) who use the media to communicate on a disastrous event. This is the aim of the present project. On a more fundamental aspect, this analysis will also contribute to a better understanding of the imaginary associated with the term “disaster” (Devès, 2016). Whatever a disaster is said natural or human-induced, it is considered as being “a disaster” within a given social, political, economical, historical, cultural framework, which is worth analysing (e.g. Favier et Granet-Abisset, 2005 ; Walter, 2009).

An interdisciplinary team

To undertake such an analysis, one need an interdisciplinary team. Our team is composed of geophysicists who are hazards specialists (Maud Devès and Jean-Bernard de Chabbalier from the Institut de Physique du Globe de Paris) and of geographers who are familiar with international media studies (Claude Grasland and Hugues Pécout from CIST, Univ. Paris Diderot and Marion Le Texier from IDEES, Univ. Rouen). Maud Devès is also Head of the Scientific Council of the French Platform for Disaster Risk Reduction (AFPCN) and brings about her knowledge of the DRR community needs and expectations. Claude Grasland has coordinated the project ANR Corpus GEOMEDIA (2012-2016) and is now involved as partner in the H2020 project ODYCCEUS (2017-2020) that proposes to analyse the dynamic of conflicts in public opinion.  

A project in several steps

Our team started to work together within the framework of the Earth Politics program. The present project is a second step that would allow us to pursue an original and fruitful partnership.

  • A fruitful preliminary analysis on earthquake media coverage

The first step of our common work has aimed at articulating very distinct datasets. For that, we decided to focus on earthquakes because they are punctual events that are, geophysical speaking, relatively well-located – we use the USGS catalogue. On the media side, we used the dataset collected within the framework of the ANR-GEOMEDIA (Grasland et al., 2011). We focussed, more specifically, on RSS feeds from a dozen of international journals published in French, in English or Spanish.

The treatment of earthquakes by the medias is a question under study since more than half a century both in the field of Earth sciences and in the field of social sciences, but few studies actually articulate both fields of knowledge. The first challenge was to design an automatic procedure that could allow us identifying of which particular event journals where talking about (Figure 1).

Figure 1: Identifying automatically what particular earthquake is mentioned in a given news is not trivial. Indeed, many earthquakes occur every day on the planet, some of which are aftershocks or foreshocks of bigger ones while some others are completely unrelated. And not all the news provides the same information about a very same event. However, an automatic procedure is required given the vast number of news items we are dealing with today.


1st step: Interoperability of the geophysical and media databases

This interoperability work has been undertaken with the support of a research engineer from GIS CIST (H. Pecout) and the help of a first master student (co-advised by Marion Le Texier and Maud Devès) who defended his M2 in Geography in June 2015 (see the research report Gollnick, 2015). It resulted in the publication of a first paper in an international journal of Geography (Le Texier et al., 2016) and has opened new perspectives by showing that media coverage of earthquakes responds to a global scheme due to a focalisation of the news on one indicator: the magnitude. As it is evident that earthquakes’ impacts and associated risks cannot be derived from the magnitude alone, these results confirm the hypothesis that the treatment made of the information on earthquakes by international journals does not really allows a good perception of risks by the lay public.

This first step has also led to the development of a robust automatic process able to interoperate the geophysical and media databases, but with still a significant proportion of error. That procedure cannot be improved however without undertaking a deeper analysis of the content of the news and of its evolution with space and time.

2nd step: Understanding of the news content and of its evolution with space and time

This qualitative work has been undertaken with the help of a second master student (co-advised by Maud Devès and Marion Le Texier) who defended his M2 in Geophysics that time in June 2016 (see the research report Mahi, 2016). For that, we focus on the RSS feeds on a longer time-period (from April 2014 to December 2015). We analyse, earthquake by earthquake, the spatial and temporal distribution of the media response and the evolution of the content of the news with time.

The first challenge has been to identify the events of interest to the media. Indeed, all earthquakes do not get to be discussed by international journals. Le Texier et al. (2016) showed that the media intensity accorded to an event was a question of magnitude but there are other important factors that cannot be eluded. This part of the project aims at understanding more finely what are these others factors and what mechanisms come into play as a crisis develops.

Indeed, medias are key players at all stages. Before the event, they can play a role in prevention by ensuring people are aware of the risks they are exposed to and of the behaviour they should adopt depending on the hazards they might encounter. During the event, they inform people by providing continuous updates on the evolution of the situation and by diffusing appropriate behaviour tips. After the event, they can provide a follow-up on the process of reconstruction. At all times, they can contribute to cultivate the risk culture and to disseminate a knowledge useful to prevent disastrous impacts in the future.

Of course, all this stands in theory. In practice, it raises many questions.

  1. Is there identifiable logic in earthquake media coverage (beyond talking about all large magnitude earthquakes)? In other words, do the medias treat similarly all earthquakes of similar size? To which extent does it depend on the level of material or human impacts? Does it depend on the country that is impacted? What is the role left to science, to national or international politics?
  2. Is that logic relevant with respect to risk prevention and management strategies? Does it follow the big steps of DRR management (i.e. disaster/response/rehabilitation-reconstruction/prevention-mitigation/preparadness) or does it focus on the most sensational, and probably most polemical, part of it? What places does it give to various stakeholders (citizen, policy maker, aid worker, civil security officer, etc.)?
  3. And, eventually, how could then be improved to fulfil DRR requirements? How could the persons or institutions who provide the primary information to the media could participate in making the process better?

We have undertaken a preliminary analysis in the last year and we are now able to confirm the existence of an identifiable logic in earthquake media coverage. The results obtained by Mahi (2016) confirms that media coverage intensity is not always correlated with magnitude (figure 2) and that the time distribution of the media coverage depends on the event (there can also be competition between two events occurring a same day) but it also shows that earthquakes of similar geophysical origins (e.g. subduction vs intracontinental earthquakes) are eager to provoke a similar evolution of the media coverage response, an evolution that seem to follow the big steps of DRR management.

Figure 2: The discrepancy between the intensity of media coverage and the biggest earthquake magnitude of the day can be explained by various factors (competition between distinct earthquakes or with another international event, level of human or material impacts, geopolitical importance of the impacted country, etc.).

  • What we plan to do now

  • 1) Understanding the mechanisms underlying earthquake media coverage variations

The questions that have just been raised are at the core of the project we propose here. We have identified clues of an underlying logic to earthquake media coverage. We now need to work at understanding better the trends we observe.

2) Extending our approach to another type of hazard

To complete this proof of concept work, we would like to extend the previous approach to another type of the so-called “natural hazards” and “physical catastrophe”. We suspect indeed important differences in the coverage of events of this type by media, according to their space-time properties and, more precisely, their capability to be anticipated or not. Consider for example the difference between an earthquake, where the media coverage is developed only after the catastrophe and a volcanic eruption that is covered by media during several weeks or months before the major catastrophe. We can therefore really ask if the media coverage of earthquakes is representative or not of the way media cover other physical catastrophes?

Volcanic hazards are a good target because of their variable phenomenology and their planetary extension. The volcanic activity has been widespread over our time period of interest (e.g. Mount Sinabung in Indonesia on October 13, 2014 ; Pico do Fogo volcano in Cape Verde on November 28, 2014 ; Sangeang Api volcano in Indonesia on May 31, 2014 ; Bardarbunga volcano in Iceland on September 14, 2014 ; Mount Kelud volcano in Java on February 14, 2014 ; Tungurahua volcano in Ecuador on August 31, 2014 ; Pavlof volcano in Alaska on November 15, 2014 ; Le Piton de la Fournaise volcano in France on June 21, 2014, etc.).


Marion Le Texier, Maud Deves, Claude Grasland, Jean-Bernard de Chaballier (2016), “ La couverture médiatique des séismes à l’ère numérique”, L’Espace Géographique, 1, p. 5-24.