Creating the NASA GISS animated climate spiral in R (CC220)

In this episode Pat will use R to adapt a figure made by one creator to look like one made by another creator. He'll share his thought process for thinking about what code needs to be adapted to achieve the desired effect. This will all be done using animated climate spirals of monthly temperature anomalies by month using NASA's GISS data using tools from the ggplot2 and gganimate R packges. This figure shows the deviation in annual global mean temperatures from the normalized temperatures of 1951 to 1980 as a line plot. The lines are colored according to the size of the temperature anomaly. He uses transition_reveal, transition_manual, coord_polar, geom_line, scale_color_viridis_c, and a smattering of arguments from the theme function to create this provocative visual. All of this is done in R with the help of RStudio.

#ggplot2 #gganimate #Rstudio #climatechange #R #Rstats

You can also find complete tutorials for learning R with the tidyverse using...

0:00 Introduction
6:02 Change apperance of background
7:47 Remove title
8:40 Reformat month labels
12:25 Change color to be by temperature rather than year
15:36 Modifying appearace of gridlines
20:08 Adjusting the limits along the radius
21:19 Adjusting the gaps in the gridlines
24:32 Animating visualization

[NATS]

Hi Pat, fantastic instruction once again, thank you!

[yaqinguo]

Cool!!! I prefer to NASA one. I love this kind of color. Looks more attractive i think.

[kylelima]

I prefer NASA's climate spiral due to the color choices, the overall aesthetics, and the clean modern look. I think the color changing year is a great addition as well as having the month labels much easier to read (you don't have to read June upside down). Additionally, having January at 1200 seems more logical and is a great improvement too. The only things I don't love about NASA's figure is the green circle indicate where the 0 degree change is (which gets lost in the climate spirals and becomes difficult to see), and the fact that the March label gets overlapped with one year's data (maybe push the labels father out).