Stormtalk: Situating adaptation and vulnerability with social media

EMAPS Datasprint: 24 March - 28 March 2014, University of Amsterdam
Report by Ezgi Akdag and Inte Gloerich
Subject matter expert: Jerome Duvernoy
Project leaders: Axel Meunier, Bernhard Rieder
Team members: Andreas Birkbak, Inte Gloerich, Ezgi Akdag
Map designers: Giovanni Magni, Giulia de Amicis

Note before reading: some of the visualisations are not finished yet: they still contain some Lorem Ipsum text and fake names of researchers. The actual data that is represented is correct however.

Table of Contents


EMAPS (Electronic Maps to Assist Public Science) is a research project funded by the European Commision. Through intensive weeklong research projects, called datasprints, EMAPS aims to find innovative ways and methods for investigating at the web and its uses for different publics. Partnered with University of Amsterdam, Sciences PO, Politecnico Di Milano, University of Dortmund, Barcelona Media and The Young Foundation, EMAPS has brought professionals and students together since November 2011 to employ digital methods to map out controversial issues such as ageing politics and climate change (EMAPS). This research paper was written to present the results of the project titled “Stormtalk” as part of the datasprint held in Amsterdam in March 2014, which centred around vulnerability to climate change.

Beck lists climate change, as well as financial crises, terrorist attacks, and digital freedom, as some of the global risks that the world faces today (Beck 2013 1). The United Nations Framework Convention on Climate Change defines climate change as “a change of climate which is attributed directly or indirectly to human activity that alters the composition of the global atmosphere and which is in addition to natural climate variability observed over comparable time periods.” (UNFCC). Yet, the latest IPCC (Intergovernmental Panel on Climate Change) report came up with a different definition, conceptualising it as “a change in the state of the climate that can be identified (e.g., by using statistical tests) by changes in the mean and/or the variability of its properties, and that persists for an extended period, typically decades or longer. Climate change may be due to natural internal processes or external forcings such as modulations of the solar cycles, volcanic eruptions, and persistent anthropogenic changes in the composition of the atmosphere or in land use” (IPCC). According to Beck, climate change is egalitarian in the sense that, even though its effects are diverse in different regions, it affects the whole world, not a specific part of it. Yet, while climate change affects all countries, not all countries contribute to it in the same degree (2009 37). While currently industrialising and already highly industrialised countries try to maximise their economic profits, their methods create hazardous side-effects that affect nature and subsequently the climate (38). These negative effects are spread around the world, and are in fact most severe in developing countries (37). This means that the countries who are least responsible for the environmental change are actually most prone to its adverse effects. Beside their natural susceptibility to these adverse effects of climate change, they are also least capable of dealing with them financially (37-38). Füssel argues that this means there is a “double inequity” between responsibility, capability to allocate the resources to fight the adverse effects of environmental change, and vulnerability (597). Similarly, Barnett, Lambert and Fry state that vulnerability also varies according to class, gender, ethnicity, and location as the capacities to cope with environmental change and exposure to risk change (104). Füssel argues that this inequity argument leads to the idea that the countries who are responsible for environmental changes and have the capacity to cope with the adverse effects of climate change, should bear the costs, while these resources should be allocated to the most vulnerable countries (598).

As with climate change, vulnerability is defined in various ways, which all have far reaching consequences when it comes to allocating funds (Barnett, Lambert and Fry 103). One of its many definitions is “the state of susceptibility to harm from exposure to stresses associated with environmental and social change and from the absence of capacity to adapt” (Adger 2). Several projects have focussed on building indices to as sess the vulnerability of countries to climate change since the 1990s. Barnett, Lambert and Fry underline that even though there is not a single formula for measuring vulnerability, most indices consider “one or more exposure to risks, susceptibility to damage, capacity to recover and net outcomes” (102-103). Yet, Eriksen and Kelly contend that the assessments and the comparing of vulnerability levels of countries gain more importance in allocating funds like the Adaptation Fund, allotted by the United Nations Framework Convention on Climate Change (79). Cutter stresses the need to develop a set of parameters to assess and compare the level of vulnerability of countries for policy making (7). While assessing the level of vulnerability has been gaining importance for distributing responsibilities and resources, the different parameters used for measuring, change the ranking of countries in terms of vulnerability (Füssel 599). However, Eakin and Luers state that making such a comparison between countries is hardened by several factors, including the quality of the available data, the indicators selected for the assessment, presuppositions employed in weighting of variables and the mathematics of aggregation (377). Furthermore, as Klein and Möhner underline, there is no one approach that is systematic and widely agreed upon for measuring the vulnerability of a country (16). Economic, social, environmental indicators are all elements currently employed in different degrees to assess the vulnerability of a country (Birkmann).

Our research focusses on the current situation in France. With the above in mind, this is particularly interesting because France is a developed country, and has for instance high emission levels (Burck, Marten and Bals). It is on the other hand not part of the regions of the world that are considered to be most vulnerable to the adverse effects of climate change (e.g. ND-GAIN). In this paper, we start with an analysis of climate change discourses used by politicians in France in the context of the municipality elections, which took place in March 2014. By examining the words used by the candidates when talking about climate change, we will get a sense of local concerns related to this topic. The second part of our project is geared towards the analysis of the resonance of climate change as a topic in France. We aim to analyse this with the assistance of the GASPAR database,[1] which provides detailed information about natural catastrophes that have taken place in France. Our point of departure is the hypothesis that natural disasters will increase the discussions related to climate change. To find out to which degree this is true, we will analyse comments made by users on Facebook pages of selected French newspapers. To do this, we will use Netvizz[2], built by Bernhard Rieder, and its accompanying search tool[3]. These tools allows us to extract “observational data” from Facebook. This means that we are able to analyse people’s behaviour directly, without having to rely on their own accounts of it, which are often subject to various filtering or altering factors (Rieder 347). By looking at the differences between talking about concrete weather events, such as extremely violent storms, and the broader, abstract concept of climate change, we aim to understand whether there is a correlation between them. In short, our project aims to answer the following research question: Do extreme weather events put climate change on the agenda in France? This is operationalised using the following sub-questions: (1) How is climate change ‘localised’ during the 2014 municipality elections in France? (2) How are people’s responses to extreme weather events influenced by their proximity to the events? (3) What are the differences in language usage between climate change discussions and extreme weather events discussions? and finally (4) To what extent are discussions on climate change triggered by specific events? The following sections will be grouped according to these sub-questions.

How is climate change ‘localised’ during the 2014 municipality elections in France?

Climate change is an inherently political affair; while individual people can make changes in their lives to accommodate for a lifestyle that leaves smaller ‘footprints’ on the ecological health of the world, the problem we are facing is of such vast proportions, the consequences are so incredibly entangled on a global scale, that the real stage for action seems to be that of (international) politics. While the largest part of our research answers questions that relate to user engagement with the abstract topic of climate change, the municipality elections that took place in France on 23 March 2014 allow for an insight into the issuefication of climate change on a (local) political level that acts as a backdrop behind the comment analysis on Facebook that we will introduce in the next section. It is for this reason that we will present our findings for this part of the research before continuing with the method of the second part.

For the occasion of the elections, the Reseau Action Climat[4] conducted interviews with political candidates in nine of the main cities of France. The interviews highlight different views on climate change and the way politicians address and take action against it. We compiled the interviews into city-specific lists to localise the types of concerns related to climate change across France. This means that politicians in one city from different parties were grouped to allow for the discovery of local concerns, rather than party concerns. Each list was subject to textual analysis: the climate change related words, such as vélos, bio, and jardins, were selected and counted within each city-sphere. Subsequently, these words were grouped into the categories Milieu, Place, Concept, Organisation, Object, Operation, and Political Action. This resulted in an overview of accumulated language usage as it differs between the cities in France answering the question what do politicians ‘turn climate change into’ when they speak?

politics language_def.jpg

Link to pdf.

Figure 1: language usage related to climate change by local politicians involved in the municipality elections of March 2014 in France as it appeared in interviews with the Reseau Action Climat. The analysed words were grouped into seven categories and mapped according to their resonance in each city. The bigger the bubbles, the more a word resonates in a city-sphere. The higher a city is on the graph, the more words the politicians used to express their views on climate change.

As Figure 1 shows, collectively, the politicians in Lyon spent most words explaining their ideas on climate change. Talking about places, organisations, and political action, instead of more abstract categories like milieu and concepts, shows that they focus on concrete plans to counter climate change. Nice on the other hand, spends an average amount of words, and a lot of them concern concepts like écologique and habitat. The politicians in this city are very clear in localising (the effects of, or action against) climate change, as they use a lot of places and objects in their answers. Furthermore, it is clear that Bordeaux spends the least amount of words on the least amount of categories. The words dévelopment, écologique, and ville are shared among most cities, while words like solaire, transports, actions, and, notably, pollution are unique for specific cities.

How are people’s responses to extreme weather events influenced by their proximity to the events?

The recent history of extreme weather in France

A last prerequisite before looking into the comments on news pages on Facebook is to understand the recent history of extreme weather events in France. To get a sense of the historical cases of extreme weather in different regions in France, we used the GASPAR database which contains data from 1982 on. GASPAR uses official claims to create a list of natural disasters and instances of extreme weather and categorises them according to their character. In our case, the database provides a reliable account of the weather situation in different parts of France during the two earliest months that are the subject of our research: February 15th 2010 to March 15th 2010 and September 2013. Although the GASPAR database is updated regularly, the data on February 2014, the most recent month subject to our project, was not available yet at the time of research. As the focus of our research was motivated by the storms on the Atlantic coast of France in February of 2014 (Garric), we took the Gaspar categories of phénomènes meteorologiques and the often related inondation as input to create a map of relevant events. The occurrences of these events were accumulated for each month and each département.


Link to pdf

Figure 2: a comparison between the flooding and meteorological phenomena in France between February/March 2010 (orange) and September 2013 (turquoise) as recorded in the GASPAR database. The bubbles are placed on the département that claimed the events. The sizes of the bubbles indicate the amount of events claimed by that municipality. The shades of gray of the base represent the total amount of events in that département.

From figure 2 it is clear that February/March 2010 was host to a lot of extreme weather events, especially in the form of the storm Xynthia on the Atlantic coast (Davies). This means that it will serve as an interesting comparison with February 2014. September 2013 serves as a control month; it has relatively few cases of extreme weather. This means that we will be able to compare the influence of extreme weather events. This map will allow us to compare other findings on social media interaction to actual events that might have triggered them.

stream graph_all events_def.jpg

Link to pdf

Figure 3: All floods and Meteorological events in France between 1982 and 2014, aggregated by region and year. Each stream represents a region. While the X axis indicates the year, the thickness of the streams indicates the amount of extreme weather events that occurred in that particular year in that region. The higher a region is placed in a particular year, the more extreme weather events it hosted in that year.

Using the same data as above, although this time from 1982 to 2014, figure 3 was created to provide an overview of the history of extreme weather in France. Here the weather events were grouped by year and region, and organised to create a timeline in which the total amount of events can be compared. It is clear that the recent years are not exceptionally eventful; rather somewhat uneventful. As in Figure 2, here too Xynthia shows up; in 2010 the region Poitou-Charentes, usually a quiet region, becomes pronounced as the host of extreme storms and floods. The violent storms in 1999 (BBC) place more recent years in perspective, although individual regions, such as in fact Poitou-Charentes, have been equally affected later on.

L ocating comments on Facebook newspaper pages in France

To understand if and how people relate storms and other concrete weather events to the more abstract notion of climate change, we looked at user comments on Facebook posts by French newspapers. Table 1 shows the list of national and local newspapers that were subject to our investigation. Although the local newspapers reach audiences across various regions in France, it is beyond the scope of this research to cover each département.

National newspapers Local newspapers Regions reached by local newspapers
Le Monde Le Telegramme Brest Finistère
Le Figaro Le Telegramme Quimper Finistère
Libération Charente, Charente-Maritime, Dordogne, Gers, Gironde, Landes, Lot-et-Garonne, Pyrénées-Atlantiques
Le point Ouest France Vendée, Sarthe, Orne, Morbihan, Mayenne, Manche, Maine-et-Loire, Loire-Atlantique, Ille-et-Vilaine, Finistère, Côtes-d'Armor, Calvados
L'express Le Telegramme Finistère, Morbihan, Côtes-d'Armor
Le nouvel Observateur Le Dauphiné Libéré Grenoble Ain, Hautes-Alpes, Ardèche, Drôme, Isère, Savoie, Haute-Savoie, Vaucluse
La Croix La Montagne Puy-de-Dôme, Allier, Corrèze, Cantal, Creuse, Haute-Loire, Haute-Vienne
20 Minutes La Voix du Nord Nord, Pas-de-Calais
Rue89 La Dépêche du Midi Ariège, Aveyron, Haute-Garonne, Gers, Lot, Hautes-Pyrénées, Tarn, Tarn-et-Garonne, Lot-et-Garonne, l'Aude
  L'Alsace Haut-Rhin, Bas-Rhin, Territoire de Belfort, Doubs, Haute-Saône
  Le Parisien Oise, Paris, Hauts-de-Seine, Seine-Saint-Denis, Val-de-Marne, Seine-et-Marne, Yvelines, Essonne, Val-d'Oise
  Nice-Matin Alpes-Maritimes
  La Provence Bouches-du-Rhône, Vaucluse, Alpes-de-Haute-Provence
  Le Républicain Lorrain Metz Moselle, Meurthe-et-Moselle, Meuse, Vosges
Table 1: the selected newspapers and their respective reach. Newspapers in the first column are distributed nationally.

These Facebook pages were scraped using the Netvizz application. Subsequently, using the Comment Analysis tool, we were able to analyse the content of the comments. To be able to make several different comparisons, the files were put through the tool individually, as well as grouped by year as a total, and grouped by year and reach (national or local). After manually reading through comments and testing queries for their effectiveness, we decided on the following queries as the base of our research:
  • [tempete]
  • [climat OR rechauffement]
While the French terms tempête and réchauffement use accents, the tool was programmed to recognise the spelling with and without accents. The query climat will in this tool also return longer words such as climatique. The first query is often used to refer to storms, and will thus create strong links with the weather events at hand at a certain time. We coined the term stormtalk for the conversations captured by this query. The second query is designed to return comments on climate and climate change, referred to in this project as climatetalk. Figure 4 and 5 show that the relationship between mentions of storms and mentions of climate change differs greatly between national and local newspapers: in national newspapers the two seem to be linked, whereas in local newspapers this correlation is much smaller. This first insight lead us to go on to the investigations described below.


Figure 4: a representation of the resonance of [tempete] (blue) and [climat OR rechauffement] (red) in February 2014 in national newspapers as returned by the Comment Analysis tool.


Figure 5: a representation of the resonance of [tempete] (blue) and [climat OR rechauffement] (red) in February 2014 in local newspapers as returned by the Comment Analysis tool.

For further analysis, we aim to situate stormtalk and climatetalk geographically. To be able locate the comments, we looked at the départements in which the newspapers are distributed and grouped them according to this. Using the Comment Analysis tool, the percentages of stormtalk and climatetalk in the comments of each researched month were calculated separately. These percentages were graphed on a map of France to allow for comparison between the different subjects, and between different regions of the country. Figure 6 illustrates the relative percentages of climatetalk and stormtalk in the total comments in February 2014. It is clear that the region hardest hit by the February storms, Brittany and other parts of the Atlantic coast, also generated the most comments on either topic. While the subject of climate and climate change resonates well in this region, the storms resonate even more in almost all départements. It is only in the region around Paris and the more Western and Southern départements that climate change takes the lead in the conversation. These are all regions located far away from the destructive forces of the storms.


Link to pdf

Figure 6: a geographical illustration of stormtalk and climatetalk in February 2014 by département. The bubbles represent the percentages of the climatetalk and stormtalk in the totals comments. While the red bubbles show the climatetalk, blue bubbles represent the stormtalk. The smaller bubble of these two is always placed on top of the bigger one. The shades of gray shown on the basemap represent the total amount of comments made in that département.

Going back in time, figure 7 represents the same information, but this time from February 15th to March 15th in 2010. As some of our selected newspapers were not present on Facebook at the time, they have been taken out of this representation. Some of the regions did have a Facebook presence, however, none of their comments returned through our query. These regions are shaded on the map, but they do not have bubbles on them. This map illustrates a similar trend, although less pronounced this time because of the lack of data; the regions close to where the impact of the storms was highest are dominated by stormtalk, while many of the départements around Paris, far away from the storms, are dominated by climatetalk.


Link to pdf

Figure 7: a geographical illustration of stormtalk and climatetalk between 15th of February and 15th of March 2010 by département. The basemap is shaded to represent the total amount of comments in that département. The bubbles grow according to their percentage of total comments. The red bubbles show climatetalk, the blue bubbles show stormtalk. The smaller bubble of these two is always placed on top of the bigger one.

A last location comparison is made regarding the resonance of climatetalk between the storm-heavy month of February 2014 and the, weather-wise, quiet month of September 2013. Figure 8 shows the percentages of comments that were related to climate and climate change in each département. Although some regions vary in their amount of climatetalk, the overview of the map indicates that there is no strong correlation between the occurrence of extreme weather events and the occurrence of online commenting on climate change. In the coastal regions climatetalk is resonating more than in other regions, such as in Île de France and other regions close to Paris. Comparing this with Figure 2, it becomes clear that this is not related to historical proneness to extreme weather events; both regions are similarly shaded in the earlier map.


Link to pdf

Figure 8: a geographical illustration of climatetalk. A comparison is made between comments made regarding climate change in September 2013 (blue), and those made in February 2014 (yellow). The basemap is shaded to represent the total amount of comments in that département. The half-bubbles grow according to their percentage of total comments.

What are the differences in language usage between climate change discussions and extreme weather events discussions?

While the concern of the previous method was to localise the resonance of talk about storms and the climate, the method we will discuss in this section is aimed at facilitating a more textual analysis. The basic method of data collection is similar to the previous one; we queried the Comment Analysis tool with [tempete] and [climat OR rechauffement] to get two sets of data surrounding the two main concepts. The tool has a functionality that returns words that are used in close proximity to it. This is the basis for the following analyses. After looking at the lists of words that the tool created with this functionality, we decided on four categories to group the words into: Weather, Climate, (Inter)national/Politics, and Personal/Local. Figure 9 maps the most frequently used relevant words for each dataset grouped according to these categories. The most striking difference between the two data sets is the size of the Personal/Local and (Inter)national/Political categories. While the stormtalk comments, on the right side of the map, are dominated by personal and local words like dégats, courage, and espoir, the climatetalk side of the map is dominated by international and political words like gouvernement, négociation, and responsabilité. It has to be noted that a part of the words in this category are irrelevant: they relate to the political situation in the Ukraine, and were included in this data set because of the usage of terms like climat politique. Notably, the climatetalk data set only contains one weather related word, while this category is the second biggest in the stormtalk dataset.

stormtalk vs climatetalk_def.jpg

Link to pdf

Figure 9: This map illustrates the resonance of stormtalk and climatetalk in February 2014. Words that are frequently used close to the keywords are grouped into the categories Weather, Climate, (Inter)national/Politics, and Personal/Local. The size of the bubbles reflects the frequency the word occurred. Different versions of the same word (such as pay and pays) have been grouped together.

Zooming in on the stormtalk in February 2014, we created figure 10 using the same method, although only using the query [tempete] this time to create a comparison between national and local newspapers. The largest category in the national news sphere contains words about international and national places and politics. This means that the storms are talked about in a more disconnected and dislocated manner than in the local news sphere. Here, the Personal/Local category is the largest and contains words that show personal compassion or losses. Another interesting element of this map is the relative sizes of the Climate and Weather categories within each news sphere. Relatively, the Weather category is much larger than the Climate category in the local news sphere. In the national news sphere both categories are approximately the same size. This means that words relating to climate take up a bigger part of the national stormtalk than of the local stormtalk when compared to words relating to the weather; climatetalk plays a bigger role in the national news sphere.


Link to pdf

Figure 10: This map illustrates the resonance of stormtalk in February 2014, compared between national and local newspapers. Words that are frequently used close to the keywords are grouped into the categories Weather, Climate, (Inter)national/Politics, and Personal/Local. The size of the bubbles reflects the frequency the word occurred. Different versions of the same word (such as pay and pays) have been grouped together.

To what extent are discussions on climate change triggered by specific events?

The next map shifts our focus from stormtalk to climatetalk. The map shows the shared and unique words that occur in the climatetalk of September 2013 and February 2014 in the national news sphere. These words and their frequency were again gathered using the Comment Analysis tool, this time only querying [climatique OR rechauffement]. This query was slightly changed from climat to climatique to lose the overrepresentation of the Ukraine in the earlier dataset. The map in figure 11 and 12 shows the overwhelming presence of unique words. Only five words are shared between the two time frames in all categories combined. This indicates that discussions on climate change do not consist of a large base conversation. The individual words indicate that the conversation is driven by the influence of particular events, such as the publication of the IPCC report or explorations in outer space. The Climate category within this climatetalk map has the largest number of individual words and can thus be seen as the most discussed subtopic.


Link to pdf


Link to pdf

Figure 11 and 12: this map visualises the unique and shared words used in September 2013 and February 2014 in national newspapers. The size of the half-bubbles indicates the frequency with which the word was used.


In this project, we examined climate change discourses used by political candidates of municipality elections in 2014. Additionally, we attempted to analyse the perception of vulnerability to climate change when exposed to natural disasters. By examining stormtalk and climatetalk in the comments posted on Facebook pages of national and local newspapers we found a way to look at this perception of vulnerability.

The words chosen by politicians illustrate that climate change is conceptualised in a highly localised manner. Although some words are shared between many of the cities, none are shared across the board. Even more telling is the amount of words that is shared only between a small number of cities, or in fact is unique to one specific city. Another way of interpreting this localised manner of speaking about climate change is by looking at the types of words used. While the category of Concepts is among the least used, the highly local category of Places is among the most used. This insight into the political conceptualisation of climate change as a local concern proves interesting when looking at how Facebook users deal with the topic.

Our initial hypothesis, which assumed that the natural catastrophes increase the discussions about climate change was indeed falsified by our research. The findings indicate that people do not necessarily link natural catastrophes to climate change; climatetalk was stable across an eventful and a quiet month, and climatetalk was relatively more present in regions that were not affected directly by the natural catastrophes than in those that were. Through our textual analysis of climatetalk and stormtalk we found there is a great difference in language usage between the two; while climatetalk uses more words related to large, abstract concepts like international politics and the climate, stormtalk uses more words that are personal, local, and directly weather related. This indicates that in fact, the two discussions are not intrinsically linked or interconnected. Extreme weather events do not cause affected people to express feelings of being more vulnerable to climate change. When zooming in on climatetalk, we found out that the language differs greatly between two months. This leads us to the conclusion that the discussion is not ongoing, but rather triggered by events. The nature of these events do not strictly have to be related to climate change or extreme weather, as the discussion in September 2013 illustrates.

In this research project we focused on a specific type of natural event, namely storms, to understand whether there is an established link between occurrences of extreme weather and occurrences of discussions on climate change. While we could not find a clear link between climatetalk and stormtalk, it is possible that other types of natural catastrophes do trigger conversations about climate change. As our inquiry only focussed on this specific weather event, it is not possible to claim that natural catastrophes do not have any effect on climate change perception or discussion. To be able to analyse the relation between natural catastrophes and climate change perception more extensively, the question of whether other types of natural catastrophes trigger climate change talk should be explored in future research. Secondly, as our data set for the issuefication of climate change among politicians consisted of one database of interviews, it is certainly feasible and interesting to extend the methods we used to a larger data set. Future research in this direction could answer questions related to political perceptions of climate change, and comparisons between this and the perceptions of the public or action groups.


Adger, W. Neil. "Vulnerability." Global environmental change 16.3 (2006): 268-281.

“Article 1: Definitions”.United Nations Framework on Climate Change.Web. 31 March 2014. <>.

Barnett, Jon, Simon Lambert, and Ian Fry. "The hazards of indicators: insights from the environmental vulnerability index." Annals of the Association of American Geographers 98.1 (2008): 102-119.

Beck, Ulrich. World at risk. Polity, 2009.

[.....]. “The Digital Freedom Risk: too fragile an acknowledgment.” Open Democracy Accessed on 04 Apr. 2014. <>.

Birkmann, Jorn. "Indicators and criteria for measuring vulnerability: Theoretical bases and requirements." Measuring vulnerability to natural hazards: Towards disaster resilient societies (2006): 55-77.

Burck, Jan, Franziska Marten and Christoph Bals. “The Climate Change Performance Index.” Germanwatch and CAN Europe, 2014. Web. 04 Apr. 2014. <>.

Cutter, Susan L. “The vulnerability of science and the science of vulnerability.” Annals of the Association of American Geographers 93.1 (2003): 1-12.

Davies, Lizzy. "Fatal Storm Sweeps through France." Guardian News and Media, 28 Feb. 2010. Web. 31 Mar. 2014. <>.

Eakin, Hallie, and Amy Lynd Luers. “Assessing the Vulnerability of Social-Environmental Systems.” The Annual Review of Environment and Resources. 31 (2006): 365-394.

Eriksen, S. H., and P. Mick Kelly. "Developing credible vulnerability indicators for climate adaptation policy assessment." Mitigation and Adaptation Strategies for Global Change 12.4 (2007): 495-524.

“Fifth Assessment Report”. IPCC-Intergovernmental Panel on Climate Change. Web. 31 March 2014. <>.

Füssel, Hans-Martin. "How inequitable is the global distribution of responsibility, capability, and vulnerability to climate change: A comprehensive indicator-based assessment." Global Environmental Change 20.4 (2010): 597-611.

Garric, Audrey. "Tempêtes, Pluie Et Chaleur: L'hiver De Tous Les Records En France." Le Le, 04 Mar. 2014. Web. 31 Mar. 2014. < []]>.

Klein, Richard JT, and Annett Möhner. "The political dimension of vulnerability: implications for the Green Climate Fund." IDS Bulletin 42.3 (2011): 15-22.

"ND-GAIN Index." ND-GAIN Index. Notre Dame Global Adaptation Index, n.d. Web. 04 Apr. 2014. <>.

"Objectives." EMAPS. EMAPS, 2012. Web. 01 Apr. 2014. <>.

Rieder, Bernhard. “Studying Facebook via Data Extraction: The Netvizz Application.” University of Amsterdam. (2013): 1-10.

“Storms Death Toll Rises.” B BC. 29 Dec. 1999. Web. 31 Mar. 2014. <>.

[1] Information on the Gaspar database can be found at:

[2] Netvizz is a tool that allows the user to gain insight into (user interaction with and organisation on) Facebook. The tool can be found at:

[3] The Comment Analysis tool, also built by Bernhard Rieder, allows the user to query Netvizz files and gives overviews of the resonance of that query over time. The tool can be found at:

[4] The Reseau Action Climat is the French association concerned with the action taken by various other actors to counter climate change. Their homepage can be found at: The interviews used in our research can be found here:
Topic revision: r2 - 11 Apr 2014, InteGloerich
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback