Satellite data shows warming trend

Charlie Nelson
February 2007

As shown in other articles on this site, temperature measurements show clear evidence of global warming.  The (land and marine) surface measures show a statistically significant increase of 0.6 degrees Celsius per century since 1880.  Upper-air Radiosonde Balloon measures show a statistically significant increase of 0.8 degrees Celsius per century since 1958.

MSU Satellite data, which starts in 1979, has shown no statistically significant increase in temperatures over the period 1979 to 2002.  However, the analysis in our earlier article showed that the data was consistent with the warming trend observed in the longer time-series records.

Since 2002, the upward trend in the satellite data has edged closer to statistical significance.

Until 2005, the global atmospheric temperature as measured by satellites displayed no statistically significant trend that is, the chance that the upward trend was a fluke was more than on in 20.  It was getting close to that level of statistical significance but has now broken through that threshold.  There is a statistically significant upward trend of 0.9 degrees per century (Chart 1) and the chance that this is a fluke result is one in 50.

This is a very important milestone.  The lack of a statistically significant trend in the satellite data has been used by skeptics as evidence that global warming is a hoax.  But as the measurements have only been made since 1979 the time series has until now been too short to separate trend from random variation.

Chart 1


Now skeptics have no data to back their case and some have resorted to a very unscientific argument - that there has been no net warming since 1998.  This is a spurious argument as the range of variability of annual temperatures masks trends over short periods of time such as 20 years.  The year 1998 was exceptional in that it was affected by the upward trend caused by the greenhouse effect plus a strong El Nino plus the peak of the sunspot cycle.  There has been net warming since 1997 and since 1999 - but not since 1998.  In a few years, the sunspot cycle will again peak and when we have an El Nino event near that peak, the 1998 record will be broken.

What argument will skeptics come up with then?