Dr Roy Spencer was quick off the mark this month, releasing the University of Alabama-Huntsville (UAH) global average lower tropospheric temperature anomaly for September 2013 on October 3rd.
The anomaly refers to the difference between the current temperature reading and the average reading for the period 1981 to 2010 as per satellite readings.
September 2013: Anomaly +0.37 degrees Celsius
This is the joint 3rd warmest September temperature recorded since the satellite record was started in December 1978 (34 May observations). The warmest September to date over this period was in 2010, with an anomaly of +0.45 degrees Celsius.
As background, five major global temperature time series are collated: three land-based and two satellite-based. The most high profile satellite-based series is put together by UAH and covers the period from December 1978 to the present. Like all these time series, the data is presented as an anomaly (difference) from the average, with the average in this case being the 30-year period from 1981 to 2010. UAH data is the earliest to be released each month.
Spencer, and his colleague John Christy at UAH, are noted climate skeptics. They are also highly qualified climate scientists, who believe that natural climate variability accounts for most of recent warming. If they are correct, then we should see some flattening or even reversal of the upward trend within the UAH temperature time series. To date, we haven’t (click for larger image).
That said, we also haven’t seen an exponential increase in temperature, which would be required for us to reach the more pessimistic temperature projections for end of century. However, the data series is currently too short to rule out such rises in the future.
One of the initial reasons for publicising this satellite-based data series was due to concerns over the accuracy of terrestrial-based measurements (worries over the urban heat island effect and other factors). The satellite data series have now been going long enough to compare the output directly with the surface-based measurements. All the time series are now accepted as telling the same story (for a fuller mathematical treatment of this, see Tamino’s post at the Open Mind blog here).
Note that the anomalies produced by different organisations are not directly comparable since they have different base periods. Accordingly, to compare them directly, you need to normalise each one by adjusting them to a common base period.