Streamgraph visualization of Global Warming

Streamgraphs are stacked area graphs displaying different time series upon one another (see Lee Byron introduction of streamgraphs). In this visualization I used a streamgraph to plot the time course of temperature changes measured in different zones on the earth. The data is from NASA and it is available here.

Streamgraphs allow plotting all data at once. However, if the data is complex it is difficult to understand what the graph is showing. The dataset I chose contains temperature for each year from 1880 to 2016 for zones (Globe, North Hemisphere, South Hemisphere, and the zones between latitudes 24N-90N, 24S-24N, 90S-24S, 64N-90N, 44N-64N, 24N-44N, Equator-24N, 24S-Equator, 44S-24S, 64S-44S, 90S-64S). Given the amount of zones it was difficult to find a way to represent the data clearly with a streamgraph. Therefore I chose to represent the different zones on the earth also as a areas on a circle on the side of the streamgraph. The lines crossing the earth are the latitudes representing the zones at which the temperatures were measured. I used a blue bar scrolling through the circle/earth to identify the zone on the streamgraph.

Is temperature rising?

I found particularly interesting the increase of temperature in the zone between the 60th and 90th parallel south (60S-90S). From being the lowest temperature it became the zone with the highest increase in temperature across the years. It is interesting that there is no land there but mainly ocean and ice, which is likely to increase in temperature more rapidly than earth. Would this also be a mechanism of mother nature to protect us from warming up the planet too much? ... it should work, up to the point that we've melted all the ice.

On github I posted the code to reproduce this visualization and the data. On my blog I wrote a short explanation of the code I used to make the visualization. This page instead has an interactive version of this visualization.

Temperature rise