Visualizing Climate

Observations from the conterminous United States for the year 2017

When choosing a place to live, people often consider climate, inter alia, but as the number of options grows, the degrees of difference lessen, trade-offs become multi-dimensional, and heuristically comparing climates becomes quite challenging.

Classification systems intended for climate comparisons like Köppen-Geiger or Trewartha are certainly more accurate than heuristic-based comparisons, but these systems may not be granular enough. To achieve a higher degree of granularity, I averaged NOAA temperature records and projected them spatially:

Fig. 1: Mercator projection of NOAA climate observations

The projection above reveals interesting regional patterns, but it is less suited for determining the degree of difference. For a comparison of climate features by degree, a multi-dimensional scatter plot may be more useful. Figure two depicts how temperature variance and rainfall vary with average temperature. To find consistently hot and dry locations in figure two, zoom-in on the bottom-rightmost corner where rainfall is zero, temperatures are highest, and variance is at a minimum.

Fig. 2: Interactive scatter plot of temperature variance, average temperature, and rainfall
Drag to rotate. Scroll to zoom. Hover for data.

Finally, I include a tabular representation of the above data which is useful for fast look-ups. As a reminder, these values are sampled from a single year, and for most purposes, it would be prudent to average climate observations over 15 or 30 years.

Fig. 3: Tabular data