Visualizing Climate Change

Climate Change is Rewriting the History Books is an infographic from Climate Central that uses a heatmap design style to show how average temperatures have changed over the last 137 years.

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!

Oxford Crime Heatmaps from the BBC

As part of the BBC launching thier new show called The Truth About Crime, they have launched a new "Crime Map" website that uses heatmaps to show real crime data for Oxford, England.  The heatmaps are visual representations of all the crime data available for the 12-month period from November 2007 to November 2008.

The website is designed to allow users to explore crime patterns, discover more about potential risks and take action to prevent crime.  The site features a specifically commissioned crime map of Oxford created with data supplied by the city's emergency services.

Why are you using heatmaps?

There are a number of methods for mapping crime. Currently, the technique most often used is to map crime data according to geographic areas such as postcodes, census output areas or police 'beat-codes'. The geographic areas chosen to map crime data – such as 'beat-codes' by the police – are often done so because these services deploy their resources according to their chosen geographical areas.

However, as these geographical areas vary greatly in size, when crime data is plotted on a map it is often difficult for a member of the general public to properly see and understand which areas have high or low crime rates. A large area may seem to have more crime than a small area even though this is simply because there is more space and people in that area. A small area with high crime might be hard to spot because it is simply physically smaller on the map, and therefore harder to see.

After extensive consultation with a host of experts in this specialist area, we have decided to use 'heatmaps' to display our crime data, since these offer a clear way for us all to see patterns of crime, without requiring us to have the expert knowledge of crime data analysts, nor a prior knowledge of arbitrary geographical areas. These 'heatmaps' represent the relative amount of crime according to a sliding scale of colour (as detailed in the "Key"), and provide a sense of the area where a type of crime is happening without disclosing the exact location that it took place – so as to protect the anonymity of victims

Heat maps such as these have not previously been used to any great extent in the UK, but have been used in the USA and Canada.

A number of different maps are available showing Crime Patterns Over Time, Burglary & Theft, Violent Crime, Anti-Social Behavior and the ability to compare to your own neighborhood (if you live in the U.K.)

Halo 3 HeatMaps

A while back Bungie.net, the makers of the Halo series of games, started tracking data on their servers about how their different online multiplayer maps are preforming.  They converted the data on kills and deaths in the multiplayer games into heatmaps, and then started publishing the maps online for everyone to see.

The advantages to players are that you can see places to avoid (areas with the highest deaths), and the locations from where the most kills come from.  The map above shows the total data for the map called The Pit.  But you can narrow down the information based on the type of weapon used.  For example the map below shows the locations of the kills made with the sniper rifle.  Meaning that shooting from these locations have been the most successful.  (Also helpful if you keep getting killed by snipers and can't find them)

"Heatmaps are the Doppler Radar System of Death in Halo 3. We're tracking encounters, weapons used and their results in a given game, collecting that data and sharing it with players visually. The key here is 'the darker the red, the more frequent the deaths (or kills, depending on the parameters)'," Bungie explains in its weekly update.