The Feeling of Red and Blue—A Constructive Critique of Color Mapping in Visual Climate Change Communication

In this article the authors will present some new research gained by different methodologies on color symbolism in climate change visualizations. For this purpose, the authors took questions and insights derived by visual studies methods and transferred them into the test setting of a qualitative survey on impacts of color schemes in climate change maps. The core question is what impact colors have on the interpretation of global climate change visualizations. With their combined setting the authors sought to interpret maps depicting global temperature rises not from a cartographic perspective in the first place but from an iconic perspective: what does a certain color scheme applied to global data maps tell different viewers? Does color play a role in the assessments of scientific appearance, risk and danger?

With the outcome of the research the authors hope to offer a contribution to estimate the complex and sometimes conflicting impacts of color in visual climate change communication, by taking into account the following difficulties: the difference of culturally influenced types of risk perception, the gravity of the issue and the dynamic interplay between science, media and the public, that trigger different narratives and speak different languages of catastrophe (Hulme 2009, Chap. 6). Therefore, if an IPCC-map is analyzed in the following this approach is not about criticizing the message of climate science or the work of the IPCC, but to estimate the role IPCC visualizations produced for policy makers play in climate change communication and to value their productive function as communicative instruments for the “honest brokers” of climate change. Nonetheless the authors know that analyzing the complexity of climate change communication today is in itself problematic, because on a level of the post-factual many powerful actors still undermine the findings by a profound and deliberate misunderstanding of skepticism (Washington and Cook 2011). At the same time, the authors should like to emphasize right at the beginning the limits of their survey: because of the small size of the group interviewed the argument cannot be universalized to any extent; still it is possible to recognize tendencies which would then need further research.

The question was derived from humanity’s interpretation of color in the first place. It grounds on the fundamental assumption that no message is neutral, instead facts are always filtered, amplified and rhetoricised in many ways (Hulme 2009, 225). This holds also true for images also: History of art and visual studies base on the assumption that meaning in imagery can never be completely controlled—this polyphony of pictorial meaning has been studied extensively in the field of semiotics, philosophy and art history (Hatt and Klonk 2006). Art history has proved that there always is a surplus of uncontrollable meaning. The analysis of pictorial media is able to access layers of “disguised symbolism” (Panofsky 1953) beyond the factual form, which materialize in any pictorial medium (Bredekamp et al. 2015). In particular, artists and creatives in advertising, who target broader audiences, are very aware of the fact that the meaning of images cannot be restricted completely. This is especially true for color. Colors address human perception immediately, they influence the meaning of a picture. They are able to highlight, to superimpose symbolic layers, to aestheticize, to attract attention, to clarify, to alarm or to balance out what they display.

Because of the polyphony of meaning of colors, very different and powerful symbolic meanings are inevitably assigned to the subject that go beyond the natural coding of red and blue to depict temperatures but which likewise belong to the established semantics of these colors. The appliance of color in map design often is arbitrary and in most cases got ruled only by conventions. A cultural history of color here offers many conflicting paths of interpretation (Gage 1993, 2000). On a physical level the color red is less energetic than blue, opposing the temperature scheme found in map design, because from a color-psychological perspective, red is perceived to be more warm than blue, which is perceived as a cold color (Meadows 2014, 131). At the same time red has an exceptional position because it occurs only rarely in nature. Therefore, it not only became the color of honor, love and energy, but also got conventionalized as the color of highlighting and of marking danger and action at the same time, because red also stands for the destructive power of fire and blood. The highlighting dimension of the color red applies to information design, where red triggers symbolic layers such as danger, warning signal, intensity and the deviation of a norm—especially in the case of visualizing global risks. Map historian Gilles Palsky even put forward the thesis (Palsky 1996) that colors turn maps into political maps, because colors always imply meanings and interpretations in the case of scientific climate maps becomes strangely valid also. At the same time, it was stated from a history of science’ perspective that black-and-white images in science tend to give a more neutral and objective impression (Daston and Galison 2007), although it is not clear if this argument applies to map and information design in general, because color plays a crucial role for enriching the image with information.

From an information graphic perspective, data visualization in a scientific sense must fulfill specific quality criteria. In informational design colors denote data, they represent numbers, measurements. At the same time, they denote frontiers within the data because somebody has to decide where to cut the data into zones. With maps one also can highlight certain findings and readings of the data. Informational designers learn certain rules to apply color to maps: A key finding should be recognizable by color spaces. By applying color to a map it is assumed that it gets possible to display more data layers and to make visible more details in the data in a clearer and more direct way (Bertin 1967). If certain color schemes are used it is possible to display the data in a very exact way. The image becomes readable. Since the early twentieth century, guidelines exist at a level of practical knowledge about the use of colors in a “good” and “effective” way and how to apply this to data graphics and maps (Bertin 1967; Tufte 1990; Palsky 1996; Stauffer et al. 2015). As data visualizers know, graphics are capable of containing only a limited amount of information.

Summing up, the case with red and blue is tricky. Colors can trigger symbolic and iconic levels at the same time: In the case of climate science graphs, this lead to the interpretation that the pure informative level of the intensive colors red and purple in contrast to blue in order to code warming can very strongly direct attention to all possible notions of red in connection to the motive: alarm, devastating heat, fire and destruction, meaning that the hot color schemes not only work on a rational level but at the same time heat up the feelings of their beholders and depict future as catastrophe (Schneider 2016; Horn 2014).

The interpretation of the IPCC map (Fig. 1) by Schneider (2016) using methods of image analysis from a humanity’s perspective, came to the following results: It contains alerting effects by the chosen fiery red color spectrum; this impression is increased by the extra color of a bright purple added to the end of the color scheme, where one would expect a darker red. The color purple marks the highest temperature increase on the map. It adds alertness in particular because it carries connotations of unnatural danger, alarm, and abnormal exceeding of extreme values (“the new normal of abnormalities”). If this color scheme is used to present already occurred global warming, the map suggests a burning future of extreme disasters that does not call for action but evokes feelings like powerlessness.

Open image in new window

With the survey undertaken here the authors want to change this analysis into a constructive critique, because it is not about interpreting a certain map as “good” or “bad”, but to gain a better understanding in the role of color in relation to climate change maps. Therefore, the authors are interested to ask in a more general way if color plays a crucial role for the impact of global warming imagery, how would a politically constructive and empowering map look like? Or is it just not possible to change the imagination of climate change by modifying color?

Those interpretations, derived from a cultural reading, the authors wanted to study more deeply by changing the methodology from visual studies to a qualitative survey and by at the same time altering the color scheme of an IPCC global temperature map. If color changes perception on an emotional and interpretative level, different colors would have to lead to different interpretations. So what does it change for the understanding and impact of the map to apply colors different to the conventional color scheme of blue (cold) and red (hot)? Therefore, for test purposes, the authors chose a prominent global map from the IPCC AR5 report. In the process of recoloring this map it gets possible to analyze its color scheme and to provide alternatives.

For their study the authors decided to change a single global temperature map from the Summary for Policy Makers-Report of the Intergovernmental Panel on Climate Change, Working Group I (see Fig. 1). The map represents global temperature changes between the years 1901 and 2012 based on measurement data. It visualizes data of global temperature observations from the last 100 years by using a color scheme that diverges from blue to orange to different shades of red, the highest temperature rise is given in a bright purple. The map played an important role in the recent report, because it proves that global warming is already happening in the present, it is not a distant future anymore. Its importance was confirmed during the press conference of the IPCC on the occasion of the publication date of the last report in September 2013. It was one of the few visualizations that got highlighted during the press conference to emphasize the urgency and the need to act politically because of global warming. At the same time global red maps in general might be called iconic for global warming—only that this time for the first time it was not the future of global warming presented in a bright red color scheme but the presence in comparison to the past.

2.1 Survey Redesign of the IPCC Map

For their survey the authors decided to change the map’s color scheme only by basing on the color maps from Brewer et al. (2003) and the web portal www.colorbrewer.org, a resource that was designed at the GeoVISTA Center at Penn State, USA, where map designers can try out optimized color schemes adapted to their nature of data. In this case the authors had to look for alternative color schemes for diverging data, because this is the nature of the data given here: data that allows the emphasis of a quantitative data display that progresses outward from the reference period of the data range—that is from degrees below zero to degrees plus zero. We retained the number of data classes from the original map, that is 13 different temperature bands ranging from −0.6 °C to 2.5 °C, with a slight change in step size above 1.25 °C (from that point on the steps are changing from 0.2 to 0.25 °C). We maintained the general design and coloring of the map: white squares mark geographical regions where the data set is not robust enough to tell about temperature changes during this period of time (e.g. above the poles, the Pacific Ocean, central Africa and small areas in Asia and South America). Dark little crosses in each colored square highlight those regions where observations permit an especially robust estimate.

For the purpose of testing a broad range of different color schemes, the authors created five alternative options for the global climate map, which could then be tested against the original IPCC-map (see Fig. 2).

Open image in new window

1. 1.

   Map I (Blue/Red/Purple): original color scheme of the IPCC.

    
2. 2.

   Map II (Blue/Black, “less color”): In this case a map scheme varying from blue (cooling) to black (heating) was developed, in order to analyze how perception would change if colors were reduced as far as possible. One has to note that it was not possible to dispense with color completely because of the diverging nature of the data between cooling and warming temperatures. This is why the authors created a color scheme that uses as less color as possible by diverging from blue to dark gray.¹

    
3. 3.

   Map IV (Purple/Green, “anti-intuitive”): The authors purposely chose a contra-intuitive map scheme ranging from purple (cooling) to green (heating). By symbolizing heat with green and cooling with a warm color they wanted to underrun the impact of intuitively chosen colors in the field. The color spectrum stands in contrast to the highly conventionalized color scheme in map design that range from blue to red to symbolize cool resp. warm temperatures.

    
4. 4.

   Map V (Blue/Red, “cleared red without purple extremes”): For this version the authors tried to optimize the existing color scheme by cutting out the purple part of the legend and by finding a clearer graduation in red hues. The authors will discuss this version and offer a constructive critique of the original map in the “discussions”.

    
5. 5.

   Map VI (Green/Purple: “anti-intuitive-inversed”): Here the color scheme of Map IV was reversed. Now the color scheme runs from green (cooling) to purple (heating).

    

In the first place the authors also tried out brownish colors, different shades of red, and a reversed scheme of red and blue. Testing with the different versions proved that the survey got to complex by using more than one anti-intuitive color scheme.

The authors prepared a tabular form with ten short questions relating to the maps and two questions at the end asking for a general estimation of climate change science. Four different selections of six color-coded maps had been prepared, including the original map of the IPCC, varying in choice and order. Figure one shows the six color maps.

As study participants a sample of 22 persons from Germany was recruited. Most of the participants had studied at Universities but did not have anything to do with climate research in their job. The study procedure was conducted by an interviewer who started with explaining the general meaning of the map to a singular interviewee in the first place. After the general meaning was explained the interviewer started to pose the prepared questions regarding colors and their impact. Each interview took approx. 20 min to complete. The interviews were not undertaken by the authors but transferred to a research assistant who was not involved in the research process.

again by rating the answers on a scale from 10 (strong) to 1 (weak/small).

For the interpretation the authors averaged all figures in relation to the frequency of presented image stimuli and the specific number of participants who were presented a certain selection of images. It becomes obvious that for certain maps tendencies are more significant than for others. Those are marked with bold lettering.

In the second part of the interview participants were asked to concentrate on singular maps. Significantly more interviewees rated map III and V as a “call for action” compared to map I. A general concern was felt most clearly for map III. The most alarming figure is map I. This is also true for the impression of “powerlessness” and “fear”. Supplemented impressions brought up by the interviewees affirm this reading: Map I was said to be disillusioning and discouraging. Map III was said to be confusing but to comprise also positive connotations, while map V was evaluated to be calmer, more concise and to allow more hope.

When asked for their free associations for the maps I, III resp. V,² interviewees connected map I with terms like alarm, uncontrolled heat, fire, drought, boiling over (9 out of 15 interviewees). The impression of heat and drought was also connected to map III (6 out of 17) but here more associations of desert zones were brought up (4). Map V was described with the associations of burning embers and lava (2 out of 7). At the same time the interviewees estimated map III (8 out of 15) and map V (4 out of 7) in comparison to map I as being less threatening and allowing more space to think about future solutions for human life on Earth. Two persons mentioned that in map III in comparison to map I differences in geographical impacts were better readable which were not recognizable in map I. The different associations can be summarized for map I as giving an impression of the Earth irreversibly steering into a future of extreme drought and heat zones that destroy human life, while in map III and V the problem is still shown in its severity but might leave space for hope and solutions.

Only one interviewee was skeptical about climate change research, all others rated their trust in science very high (8–10). The evaluation of the maps of the skeptical person did not allow to draw conclusions about the impact of colors.

The authors are aware that the survey undertaken has limitations in size, diversity of interviewed people (mainly living in Berlin with a bias towards academic and arts professions) and the survey setup with a limited number of testable color schemes. However, it can be seen as a test case for the general question how to investigate emotional impacts and interpretations of color schemes in climate change maps by laypeople quantitatively for the goal to constructively criticize and redesign climate change visualizations. In the following the authors will interpret the results we achieved by the survey for each map individually and then turn to some general remarks.

A discussion of the free associations that were told by the interviewees in relation to Map I, III and V in the second part of the survey, is also fruitful. The associations of map I (Blue/Red/Purple) were most significant. Individual interviewees told us the following impressions: it is too late to act; it is too warm everywhere; the impression of over boiling; feelings of fatalism and being trapped; deadly danger; drought; fire; threatening scenario for the future; extreme heat; no trees, no growth of plants anymore; devastation; the color purple was connected to the notion of “rotten” but also to the radioactive pollution symbolized on maps after Chernobyl. Map V was said to produce an image of fire and lava, but being less frightening and alarming. The same was said for Map III.

At this point the authors should like to emphasize what they could not find out by their survey. The authors were not able to get significant numbers in relation to differences in the scientific appearance of the maps. Many interviewees said that the scientific nature of the map, being a data map, would not be touched by different color schemes. This might be the reason why they rated the different maps similarly. Secondly, it would be interesting to find out about the correlation of a particular aesthetic (“beautiful”) impression of a certain map and the evaluation of its intuitiveness. This might give some more detailed insights in the good rating of map III, because this map drops out in the “gestalt” it draws of global climate change. Here we do not see a red world map but a yellow map with red continents.

In their study the authors took arguments and interpretations derived from an art historian and visual study’s approach and tried to test them with qualitative interviewing methods. For this purpose, a main geographical visualization of present climate change taken from the recent summary for policy makers published by the IPCC was altered according to its color spectrum. With a set of questions 22 interviewees were asked to compare different colorations in respect to the impact of color on the comprehension of the map, its credibility, the gravity of climate change as a risk, free associations to the color and different emotional reactions to the map.

The authors found some of the interpretations derived by image analysis from a humanities’ approach (Schneider 2016) confirmed by the survey, but also new insights could be brought up. The original map used by the IPCC in the Summary for Policy Makers (Map I)—as the results from the survey also suggest—is the most alerting by at the same time calling up disillusioning associations of powerlessness, fear, global heat waves and devastation. Several interviewees criticized it for its problematic nuanced differentiation of colors but at the same time suggested to be chosen for the purpose of influencing policy makers to act. The blue-yellow-red color map version (Map III) scored considerably better in several aspects like the readability, and by at the same time not appearing as disillusioning as the original IPCC version (Map I).

The authors already stated that the aim of the study is not a simple criticism of the original map, but the attempt to take into account the complexity of visual climate change communication, because there cannot be an all-in-one solution that works similarly for all possible recipients of the map. Still, the authors think that their study has shown that in order to balance out possible emotional reactions with the comprehensibility and the key message of the visualization, it might be useful to establish a dialogue integrating manifold expertise—art history, visual studies, interface design, computer graphics to assess the polyphony of meaning that visualizations can unfold beyond science.

If the authors could recommend a new map design for the original IPCC map (Map I), they would propose to take less colors and equally distributed interval values, around zero an interval with a color close to white, and a blue-white-red color map (because of its standardization) with visually equally distributed colors, and which avoids too dark colors at the highest temperature differences.

For future work, it might be interesting to incorporate maps of possible climate futures and their employment of red color schemes into a broader study design. It would be fruitful to study laypeople’s perceptions and interpretations of the different future RCP scenarios in relation to the color scheme used for the map visualizing past observed warming. In addition, it would be necessary to study a larger and more culturally diverse group of participants, and to see what arguments can be universalized. But it would be even more fruitful to study a group of different policy makers on the level of NGOs, stake holders or different administrators in relation to color perception and policy implementation.