Creating an Anomaly Heatmap
On the right, the
averaged global temperature per
month, calculated from Jan 1951 to Dec 1980, is displayed.
You can navigate to the next step by using arrow keys or the
displayed buttons.
To facilitate perceiving differences, the y-axis does not start
at zero.
Next, we take a closer look at the temperatures during the year
1887 - a year with a globally
extremley high anomaly in January. This winter is also known as the
"Big Die Up", as it especially affected the cattle industry in
continental North America and led to major losses of cattle that
died as a result of the extreme weather conditions.
The anomalies of each month
are highlighted in red - The anomaly is the difference between the
actual temperature and the average temperature.
If we smooth the averaged temperature curve down to a straight line,
the differences and absolute values of the anomalies become clearer.
Next, we color code the anomaly values relative to the lowest and
highest anomaly that ever appeared between 1850 and 2022.
As the anomaly value is now coded in color, we remove the y-axis and
create rectangles of equal height.
In order to compare different years to one another or to discover
trends, the year 1887 is put in the context of the period from 1850
to 2020. Finally, we repeat steps 0 to 5 for every year and display
them all at once in a heatmap to be able to identify trends or
outliers.
If you hover over a rectangle, the corresponding year will be
shown. When clicking, you zoom into the selected year to have a
more detailed looked at the values. (Note that this occludes
other parts of the heatmap.)
This website showcases an animated sequence that was created with the help of the library
GSAP-ASEQ. GSAP-ASEQ was developed in context of a bachelor thesis at the Institute of Simulation and Graphics at the Otto-von-Guericke University Magdeburg.
The used
data was extracted from
Berkely Earth. The
source code is publicly available on GitHub.