In this post, I show an R script that downloads the Hadley Centre’s 1850-2009 monthly sea surface temperature (HadSST2) anomaly data, converts the ASCII file into 2 usable R data frames, calculates overall and post 1980 SST anomaly trend rates and develops a 2 panel chart that shows SST anomalies and trend rates and the % global coverage for the SST series.
Sea Surface Temperature Anomalies
I have developed a number of land & ocean temperature anomaly trend charts in previous posts, including: here, here, here and here. These posts have dealt with atmospheric temperature anomalies. In this post, I present the 1850 – 2009 sea surface temperature anomaly data record maintained by the Hadley Centre.
Hadley Centre’s HADSST2 data series is available at this link. Here’s a screen shot of the first few lines of the ASCII file.
A few things to notice about this data file:
- 2 rows per year
- 1st row includes 12 monthly and 1 annual average anomaly values
- 2nd row includes the % coverage of globe represented in monthly observation
- anomalies are with respect to 1961-1990
- source file must be processed to separate the SST anomaly data from the % global coverage data
R Script and SST Anomaly Trend Chart
My R script reads the source file, consolidates the anomaly and coverage data into 2 data frames, calculates the overall and post 1980 anomaly trend rates and prepares the 2 panel chart below. Panel a shows the monthly SST anomalies, Panel b shows the % global coverage by month.
Review of HadSST2 Anomaly Trends
Hadley Centre’s SST2 data set includes both the estimated monthly SST anomaly and the % global coverage by month. As I studied the trend charts, I first saw the overall increasing trend in SST anomalies, with declining periods in the late 1800s and the 1940s. I also noticed the sharp rise in the post 1980 period.
As I looked at the % global coverage, I first noticed the low coverage in the 1800s, followed by the sharp drops in the 1910s and 1940s. I quickly realized that the collection of ship SST data was affected by the 1st and 2nd world wars. I added rectangles to reflect these wars.
In developing the overall trend line, I was concerned that the early record was based on relatively small global coverage while the post 1980 period covered over 60% of the global.
The post 1980 SST anomaly trend rate is 0.0128 C per year, equivalent to 1.28 C per century.
R Script and HadSST2 Data On-line
To assist others who wish to verify or reproduce my HadSST2 charts, I have posted my R script and HadSST2 files as Google documents. You can access them at these links: