Data Processing

After sourcing our data, we used Breve to identify any major issues with the data before manually rearranging our data to be suitable for our visualizations. Our first dataset is on surface temperatures in Greenland. This dataset included data points once every month from 1750 until 2013 from countries all over the world, but we cleaned the data to only include Greenland because we are interested in the melting of glaciers. Instead of having one data point per month, we went with one data point per year and made that data point July 1st of every year. This made our dataset much smaller and easier to handle. Each data point gives us the surface temperature so we are able to see the temperature over time. After cleaning our data, our variables were as follows:

• Year: starts in 1750 and is from July 1st of every year
• AverageTemperature: global average land temperature in celsius
• AverageTemperatureUncertainty: the 95% confidence interval around the average
• Country: country

Our second dataset showed the potential impacts of sea level rise on countries from around the world. When cleaning the data, we decided that we only wanted to look at land area, population, and GDP affected because those are more relevant to the human-centered narrative we want to tell. The data was originally in a complicated table across multiple sheets so we had to clean it to be in CSV format.

To clean the data we made columns with the variables above and copy and pasted the corresponding values from the table into our CSV so that it was in a format that could be handled by Tableau for visualizations.

To see the before and after of our data cleaning, visit the Data Critique for options to preview and download the original and cleaned datasets!

To create our data visualizations, we used Tableau and R. Our visualization choices were guided by three key factors: the data types we needed to represent, how many variables each graph would display, and how each visual element supported our story’s flow. Tableau was very useful for bar charts and maps to visualize the sea level rise dataset and R was helpful for using scatter plots to visualize the Greenland temperatures over time.

Presenting the Narrative

Our team chose to use WordPress to create our website due to its ease of access and its ability to present our information in a well organized format. We chose a lighter theme and used dark fonts while also including Earth tones to match the overall theme of the website. We chose light and dark contrasts to better increase readability and contrast between the text and background. With each figure and data visualization we will include an alt text or caption to make the entire project more reader friendly and ensure its accessibility to a wide range of readers. Lastly we ensured all links worked and easily allowed readers to access information within the website.