Cooperation with taz

It’s the new Twenties and everything is on social media – so are the newspapers. Instagram has become an essential platform for sharing quick news bites and unfiltered engagement with its followers. The course was taught in cooperation with the newspaper taz. The section climate taz (klima.taz) has been running its own Instagram account for some time, which it uses to communicate about the climate crisis to its audience.

The Challenge

The first part of the challenge was the medium: Contemplating the leading question of: How to bring data visualization to Instagram? – the course was a laboratory for environmental data journalism within the limits of Instagram’s small space and ultra short interaction time. Todays users tend to passively follow thematic channels rather than actively looking up generall news. So content has to be impactful, short and easy to grasp.

The second part was the data: The task was to find a suitable data set, prepare it and tell a story with it using visual story telling. In other words to apply the magic of transforming raw numbers into forms of insight.

The Data

I chose www.ourworldindata.org as my data source. The website provides valuable data sets that can be easily downloaded and processed. Thematically, I chose to look at natural disasters over the long term, to which the site devotes an entire section. I was inspired by a feature on the „Billion Dollar Disasters“ in the weather section (!) of the Tagesthemen by Carsten Schwanke (ARD). The dataset „Economic damage by natural disaster type“ seemed appropriate to tell a story that also has a highly relevant message for our future.

You can find a .zip file wih the data at the end of this documentation.

Concept

After a first superficial look at the data, I developed the following concept: Natural disasters are one burdening type of how climate change is experienced by people and the economy. The economic damage occurs worldwide – without a break but with a trend. Let us look at the severity of the global economic damage caused by natural disasters between 1988 and 2018.

Data Processing

The data processing took several steps. First, I reduced the data set in Excel to the types of natural catastrophes that are relevant for Germany - remove volcanic eruptions, for example. Then I selected an appropriate time period of the data. Data sets basically spanned from 1900 to 2019, but there were many data gaps, especially in the earlier years.

In addition, I tried initial visual plots in RAWGraphs, and discovered that displaying such a long time period in the medium of an Instagram post was rather inefficient. For these reasons, I additionally narrowed the dataset to a historical flashback from 1988 to 2018.

Data Visualization

Data structure

After grasping the data structure, I went to the datavizproject to research what forms of data visualization could be used for the structure at hand. It's an excellent site, with vivid visualizations, descriptions, and data input examples.

My first choice was the area bump chart. Here I made the bitter mistake of falling in love with a form. After several discussions with Francesca I had to kill my darling. This was only possible with the help of radical iterative variant creation.

In the end it turned out that the data structure could be best visualized in a so called stacked area graph or chart. On the datavizproject it says: The Stacked Area Chart is similar to the simple Area Chart, but here it uses multiple data series that start each point from the point left by the previous series. It is useful for comparing multiple variables changing over interval.

Visual storytelling and language

For visual storytelling I looked at pudding’s Instagram account. I like their strong and expressive visual language and the bold use of fonts and colors fitting for each data story.

Inspired I chose the approach of a colorful visualization with an own color mix for each disaster type. Fontwise I used the “Vollkorn” font, a well-developed serif typeface that evokes the feeling of a print newspaper, to emphasize the historical news aspect.

Since I was looking at economic damage worldwide I used international money bills on the whole background, at the same time the area charts got an opaque color, which kind of erases the wealth. The trend of the decade between 2008-2018, where the damage increased rapidly, is further highlighted in a brighter shade of the color and commented in text telling the percentage of the overall cost.

Final Post

This is the final version which was submitted to Francesca and the taz editorial team.

Conclusion

This project was a very rewarding challenge for me. Data visualization and data journalism as a field of work is highly demanding as it requires several skills at a high level: Data literacy, storytelling abilities and visual design know-how.

With this in mind, I enjoyed working solo – going through all steps independently and on my own responsibility, consolidating the workflow of a DataViz project in the long term. Moreover, it was very nice to be able to work so close to the real working life of data journalists. At the same time, the support from Francesca and the Klimataz editorial team was impeccable and the joint feedback rounds with fellow students very helpful - many thanks to all of you!