What I do not like about the viz:
- The legend is too long! Every year has its entry in the legend. Since the years cannot be distinguished by color anyway (most of them are grey), listing them is counterproductive.
- Different statistic measures like Median, average and interquartile Range and standard deviations are all part of the chart but are really hard to distinguish by color. They only vary in shades of grey (no pun intended).
- There is some clutter. The gridlines could be removed. The title “date” for the x-axis is not necessary since the name of the month already implies that this axis shows dates. The 1. day for each month can be removed. It is not really clear why December appears twice. The creation of two data points with only one day in between (31.12 and 1.1) seems pointless since the data will only show minimal change between these dates.
What I like about the viz:
- I like the highlighting for 2012 in green color. 2012 seems to be the year with the lowest Arctic ice extent so far – therefore the highlighting.
- I also like the salient orange color for the most recent year of 2018. This emphasizes that the most current year is likely to supersede the year 2012 as the year with the lowest Arctic ice extent.
The key message I want to convey
I want to show the decrease of sea extent over time. To achieve this, I chose a title which already frames the narrative of the viz (Ice turning into water). The subtitle then names the exact number of decrease in Million SQ KM (1.95). The number is then again picked up in the viz. The box around the number spans over all bars (each bar resembles a year) of the first visualization, indicating that the number refers to the development over all years.
The second message I want to convey is the cycle of decrease and increase of the ice extent within each year. I used a second chart type, a line chart, to show this development.
The Design choices I made
- The color choice was the most difficult one. On first thought, blue is often associated with cold and following the logic, red is associated with increasing temperature. But in the case of this viz it did not really work for me. First of all, ice is white and not blue and second, the temperatures are still very low – even though they are rising. Therefore associating low temperature with a red color which generates pictures of high heat in our head felt counterintuitive. In the end, I decided to create a continuous color palette from white to dark blue. White for the ice and blue for water from the melted ice. Since this color choice might at a first look unusual (blue often being associated with cold) I felt that I need to explain the colors in the title as well. Therefore, the color-coded title. Ice turning into water.
- To show the melting of the ice I opted for a bar chart colored by a continuous color palette from white to dark blue. This creates the imagery of melting ice. Over the years the columns transcend from a deep white into first a lighter blue and then a darker blue. The bar chart is split into multiple columns, one for each year. By hovering over the respective columns in the bar chart, the tooltips show the ice Arctic extent for the selected year.
- The line chart in the lower half of the viz shows the decrease and increase of the Arctic sea ice within a year. I found it interesting that the ice extent fluctuates by a big margin within a year. The annotation calls out the peak (March) and the bottom (September) for the Arctic sea ice extent.
- The two visualizations interact with each other. Clicking on a year in the upper bar chart will highlight the corresponding year in lower line chart and vice versa.