This post reviews an Excel chart that misrepresents the CO2 – temperature anomaly relationship. The chart developer shows 2 time series on a single chart using Excel’s double Y axis capability. This is a poor charting technique which distorts the data. The developer makes things worse by only showing 15 years of data from a data set that stretches back over 100 years.
I have extended the data period to 1880 – 2008 and prepared 2 R based charts, a trend chart and a scatter plot of CO2 versus temperature anomaly. My charts provide a more complete picture of the CO2 – temperature relationship than the original Excel chart.
Google document links to the data and R script files are provided so that readers can prepare their own CO2 – temperature charts in R, Excel or other tools to assess the situation for themselves.
Australian Senator Steve Fielding wrote a post on Climate Change that included the chart below. I assume it was developed in Excel 2003 because of the tell-tale default gray fill.
The post states “.. global temperatures have remained steady over the past 10-15 years despite skyrocketing man made carbon emissions. ”
Senator Fielding has focused on a short time period and has ignored the role of El Nino -La Nina events. 1998 saw a major El Nino event, raising temperatures while 2008 saw a significant La Nina, lowering temperatures. I have covered El Nino -La Nina in previous posts, here, here and here.
Andrew Bolt, a columnist with the Herald Statesman, picked up on Senator Fielding chart and wrote a post, asking Al Gore to explain the chart. Bolt’s post generated 224 comments, obviously a lot of people are interested in the CO2 – temperature relationship.
Let’s Look at the Long Term Data Trend
Senator Fielding’s chart only shows the 1995 – 2009 period, the last 15 years. Here’s an R based chart that looks at CO2 and temperature anomalies for the 1880 – 2008 period, 128 years.
This longer term chart gives a different perspective on the situation. It shows a fluctuating temperature pattern with a clearly rising trend for both temperature and CO2.
To build my CO2 – temperature chart, I augmented the Mauna Loa CO2 data series with Law Dome data and used the Hadley Centre temperature anomaly series (data link) to extend the same variables that Senator Fielding used. I have used annual time series in my analysis so that I could take advantage of the law Dome CO2 data in the 1880 – 1959 period before the Mauna Loa Observatory started collecting monthly CO2 measurements.
The Hadley temperature anomalies have fluctuated widely in the past, however, it is clear that there is a definite upward trend in the overall rate. The CO2 trend is also upward.
Showing parallel trend charts is one way to evaluate the CO2 – temperature relationship, a scatter plot of CO2 versus temperature anomalies provides another view that helps us to see the relationship. Here’s my scatter plot and linear regression of the CO2 versus temperature anomaly.
I have color coded the data to distinguish the Law Dome and Mauna Loa CO2 observations as well as Senator Fielding’s study period (1995-2009). My plot includes the regression relationship which shows that temperature anomaly has increased 0.0092 C for each 1 ppmv increase in CO2.
The CO2 level in 1880 was 291 ppmv, in 2008 it increased to 385 ppmv, approximately 94 ppmv in 128 years. Based on the CO2 – temperature anomaly regression, the 94 ppmv increase in CO2 would generate a 0.86 C increase in temperature anomaly.
Google Documents Data and Scripts
Climate change is an important and controversial topic that warrants careful analysis. I encourage the charts & graphs, data visualization community to participate in the dialog. To help, I am building a series of CSV files to make it as easy as possible for readers to prepare your own climate related visualizations.
My data files are available as Google spreadsheets which can be copied and pasted or exported as CSV files. Here are the links:
- Annual average CO2 (1880 – 2008); Law Dome & Mauna Loa
- Annual average Hadley temperature anomaly(1880-2008)
- Annual Law Dome CO2 (1880 – 1958)
My R scripts are available as Google documents here:
- R script for CO2 – Temperature trend chart
- R script for CO2 – Temperature scatter plot and regression
- R script to consolidate Mauna Loa and Law Dome CO2 data
- R script to download and prepare annual Hadley anomaly data
I encourage data visualization practitioners to contribute their assessments to this complex and important global issue.
Please let me know if you have any difficulty accessing the Google document data or R script files. I’m looking for a way to make the data files and scripts easily accessible. Hopefully, Google documents works for everyone.