This March clocked in as the second warmest March on record when compared to the 20th century average, according to newly released data from the National Oceanic and Atmospheric Administration. NASA data published last week came to the same conclusion, comparing temperatures to a 1951-1980 baseline.
The NOAA data shows the planet was 1.9°F (1.05°C) above the 20th century average for March, the first time any month has breached the 1°C threshold in the absence of El Niño. This March is the latest freakishly hot month following three years in a row of record heat.
NOAA and NASA baselines don’t really tell the whole story. How much the world has warmed since pre-industrial times is a crucial measuring stick for international climate talks and a more accurate representation of how much climate change is altering the planet.
Using the baseline of 1881-1910, a new, more dire picture of global warming emerges. This March was 2.4°F (1.3°C) above the pre-industrial average by that measure. More notably, this March marks a whopping 627 months in a row of warmer than normal temperatures. If you were born after December 1964, you’ve never experienced a month cooler than average on this planet.
To understand what that looks like, take a peek at the global temperature chart below. Each month is represented by a box. Cool blues have been disappearing, replaced by a wave of unending heat. Climate change is likely to continue the streak of warmer than normal months into the foreseeable future as temperatures keep marching upward.
There are a few things about this design worth commenting on:
- I'm accepting the data from NASA and NOAA at face value, because this blog discusses the visual communication and design of data. However, I can't find the data they used. It would be helpful if they provided the final data they used to build the visualization. Provide a spreadsheet with the data as an act of data transparency.
- This looks like a great use of the Conditional Formatting capability in Microsoft Excel. If it wasn't designed in Excel, it easily could be.
- Heatmaps or choropleth maps (ie. using color hue, density, shading, opacity, or saturation) are impossible for the reader to differentiate the exact difference between the values. You can get a general impression, but this is at the bottom of the Scale of Graphical Perception
- In a heatmap, the designer chooses the minimum and maximum values, and the data dictates all of the actual color saturations shown for each month. The minimum would be solid blue, the midpoint would be white and the maximum value would be solid red.
- In this case, the chart colors start near the average, not the minimum value. So, all of the months from 1881-1910 are very close to white because these are the values that were used to calculate the midpoint
- The maximum temperature value is 1.9° higher than the average, and that value is shown as the fully saturated red color. This is a choice by the designer, and makes the average temperatures in the highest months visually appear as very dramatic.
- This is the default setting for a function like Conditional Formatting. It takes the maximum value in a given dataset, and correlates that to the maximum color saturation. It's up to the designer to decide if this default setting is correct for visualizing the data.
- The article suggests that the trend will continue with warmer temperatures in the future, so an alternate choice the designer could make is to set the maximum color saturation to something like 10°F over the Baseline. The current temperatures would look much less dramatic, but it would allow for future higher temperatures to be displayed on the same scale.
The choices a data visualization designer makes has a huge impact on how the data is perceived by the audience!