Anna Lombardi, Copernicus Climate Change Service, about using Datawrapper in different organizations

Introduction

[00:00:04] Anna Lombardi: Hi everyone, my name is Anna and I'm currently based in the U.K. where I work as a climate data visualizer for the European Climate Agency Copernicus. And the first time I actually came across Datawrapper, it was back in 2016. After a few years in academic research, I was attending a master's in science communication. And this tool was listed as a possible tool for data visualization. There was a free version, it was very user friendly, so I had a go. And this tool has become part of my professional life ever since in one way or another. And today, in these 10 minutes, I would like to give you a brief overview of how and why I've been using data visualization while working in two very different organizations.

The Times newspaper, where I worked as a data and graphics journalist for five years, and Copernicus, where I now help scientists visualize climate data. So when I first joined the Times in 2018, at first I thought I would be able to swim in these open waters easily, but actually I soon discovered that I really needed a sort of life vest, to avoid drowning in all the requests that were coming my way every day.

I really needed a toolbox I could easily go to when needed, and Datawrapper definitely has become part of this toolbox for me. I've been using it on a daily basis from day one. Actually, when I joined, The Times already had a company license and a custom design fitting its own style guide, which made our life a lot easier because it meant, it was extremely easy for us to create several charts and maps every day in a uniform style, all ready to be featured on our website or to be picked up for the print edition of the paper as well.

So over five years, I've created over 3,000 charts and maps. And of course, across the team, we have produced many more, for daily news stories, but also investigations and features. I've really used it on a daily basis. And today I would like to give you a few examples of our work. And if you want, for every chart that I will be showing, there will be a tiny QR code next to them, so if you have a phone and you want to scan it, basically you can interact with the chart yourself in real time while we talk through them.

Four examples of work at The Times

[00:02:39] Anna Lombardi: So this first example is to show you how a chart can help you find stories, especially when working in media. This chart was done in 2022. It was a very hot summer in the U.K., and the news editor asked me to look for some data to back this story, and if possible, to find some extra news lines. So I've been looking at rainfall data for the first six months of every year since 1931, and instead of simply looking at the numbers in a dry CSV file, I've plotted them in a scatterplot instead, as you can see here.

And a clear message came out straight away. Actually, 2022 has been the driest year in 46 years, which was quite a strong statement, a strong news line. And actually, it ended up being a top story for the next day; it was featured on the front page of the next day's edition as well. And this is all to show you how a chart and a data visualization can really help you find some very interesting stories behind a list of numbers.

In this case, the story was about skiing in the Alps and the lack of snow in recent years. We wanted to give some extra context to that story. So we looked at temperature data to see how countries across the Alps were warming up over time. And we created this table. We used the heat map option that is available in Datawrapper to actually recreate these climate stripes for every country. And we fixed the bottom row to be the global average. So any country you look for, you can basically quickly compare it with the global trend. And we fixed the top four countries to be actually four countries across the Alps. So it was really fitting nicely, the story. But adding also the search button at the top left, actually, users could also look for any country they wanted to know about. So this extends the scope of the chart outside the news story that it was sitting in.

And this was again a chart for the same story. We wanted to show the glacier mass change over time in Europe. This could have been a simple column or bar chart, but I really wanted to give an extra visual element that makes the chart stand out a little bit more. And to give this idea of melting ice, I've used the scatterplot. I've used a light blue color to give a sense of ice and also some custom lines drawing from the top, just to give an extra bit of data visualization that gives a bit more sense of this ice melting over time.

And, over my five years at the Times, we've also experimented a lot with a mix and match of various chart types. And in particular, we tried, where possible, to also mix them in bigger and more complex infographics to give some extra context to the stories that we're sitting in. This specifically was used in a story about deforestation in the Amazons, and we've used a choropleth map to give a sense of the deforestation by region in the area, and a stacked column chart on the top right to actually give the main causes of deforestation, again, across time.

Live-updating charts

[00:06:25] Anna Lombardi: Of course, another way we've been exploiting Datawrapper is to track real time trends and changes. We've all been through a global pandemic, and we were tracking the number of cases and hospital admissions every day. At the very beginning, we tried to, maybe in a very naive way, to track and update the charts manually every day, updating the numbers. We soon realized it wasn't sustainable, especially in the long term.

So what we did was to write a few scripts in R and link our charts through the Datawrapper API. Which meant we could easily update them on a daily basis whenever new data were coming out. And we used the same approach on various different occasions. We wanted to track, for example, political polling data, weather warnings, ... Any story where datasets were basically updated periodically, we adopted this approach to try to automate it, to automate the process.

Switching chart types for a vis at Copernicus

[00:07:34] Anna Lombardi: And then last November, I left The Times and I joined the European Climate Agency, Copernicus. And here I have the great pleasure to work with a lot of brilliant scientists, who often focus a lot on their research and a little bit less on their data visualization.

So Datawrapper came in very handy to show them the difference that just a little bit more love for your data visualization can make in communicating your research and your key results. And to give you an example here, this was supposed to be a column chart. This was a chart done a few months before I joined, and it was showing temperature change over time, and in particular, the top 30 warmest months on record globally.

And the key message we get from this chart is that July 2023 was actually the warmest. But, my question was, is there any other key message that we can get from this dataset? So I plotted them in a different way. And using a scatterplot, again, we have the key message that July 2023 was the warmest month on record, but we also see straight away that actually the 10 warmest months all occurred over the past seven years, which is another interesting message coming out from the data. So really, the use of the chart, and the chart choice that you make can really change the messages that go through. And Datawrapper, I think, is very good because you can easily switch from one type of chart to another, for the same data, and really check what works best, without the need of complex coding.

Another way Datawrapper was very useful in these first few weeks in the new job was to preview some data and to draft and prototype a more complex tool. In particular, the team wanted to create a dashboard that people can use to navigate climate data. And they asked me to think about navigation for this new tool, what I had in mind.

So I created a simple line chart again in Datawrapper, but I've added a few buttons at the bottom. And these allow you to basically navigate the data, and to clearly see different decades or specific years and compare them together. So instead of a chunky data set, now you can actually look for different bits of it, using the chart.

And this was really the starting point for the development of the dashboard that we recently launched. It's called Climate Pulse. You can look it up online. And it's all free and open. And again, it all started from that simple chart with buttons to actually show different things.

So to wrap it all up, this is all I've been through today in terms of what you can do with your data visualization and how you can use different types of charts to deliver your message, in the clearest possible way.

So thank you very much. And please do keep in touch. You can find me on X, LinkedIn and I've also added my work email. Thank you very much.

Q: Examples of the two organizations?

[00:11:14] Michi (host): Thank you so much for that talk, Anna. It was really fascinating. I have a question for you, if that's alright. If you had to select two projects, one from each work environment at The Times and also now Copernicus, that in your opinion really illustrate the difference in doing data viz work at these organizations, what would they be if you're able to also reveal those perhaps?

[00:11:43] Anna Lombardi: I would say probably the COVID tracker was the key project we've been basically working on for months, unfortunately, but we had to keep it updated for a few months. And I think the strength, in that case, of the Datawrapper API was really clear because without that, as I said, manually it would have been impossible to track all the changes.

And here, I think this dashboard that we created, again. It's a nice way to provide people with quite complex data sets and let them navigate it at their own pace, look for the details they're interested in, without just throwing a chunk of data at them all at once.

So giving that extra option of navigation — navigate it for yourself and look at what interests you more — inspires curiosity in this topic, which is so complex.

Q: Workflows built around Datawrapper?

[00:12:47] Michi (host): Yeah, I think we saw an example that you had with the buttons, that it just makes the whole thing also significantly cleaner and easier for someone, that isn't maybe familiar with this data and the complexity, to just zoom in on bite sized chunks. 

We have a question from Elana here, regarding your work at The Times. You mentioned the use of the API for some automation, but I'm wondering if there are any particular workflows or processes that were built around the use of Datawrapper there. The team is producing so much great work, both in terms of quality and quantity, that I'm very curious if you have any insight you can share about how that worked.

[00:13:25] Anna Lombardi: Yeah, it's a great question. Working in a newsroom is never easy because of the number of requests you get every day. We trained our team internally. So whenever someone new joins, we give them a full overview of Datawrapper, our custom style and also some guidelines — what to do and what to avoid for each different type of visualization.

We also tried, over time, to involve the daily news reporters more and more every day in trying to do as much as they could. If they have a simple data set that doesn't really need much analysis, can they actually try and give it a go themselves, play around a bit with Datawrapper.

They could always come to us for questions if they got stuck, but quite a few actually picked it up and saved us some work. So yeah, just to share the workload a bit more with the broader newsroom as well. And I think the more we went through it, the more people actually got involved and excited about using this new tool.

[00:14:40] Michi (host): Nice. Thank you very much. There's no more additional questions at the moment, but Anna did share her contact details. So if you have any more follow up comments or feedback or questions, please do get in touch with her there. Thank you very much, Anna.

[00:14:55] Anna Lombardi: Thank you. Thanks for having me. Thank you very much.