April 3, 2009
The satellite data for march from Remote Sensing Systems (RSS) has been released. The results indicate that for two months in a row the data have shown a downward trend. This is by no means momentus however the longer term trend over the last five years continues to be in a downward phase. The latest graph is presented below.
As you can see from the Loess functions, the trend for the last five years shows a general downward trend. However, the period of the 2000s is still substantially higher (statistically significant) than the other decades (i.e. 1980s and 1990s). The persistent trend downward in recent years places in doubt any declaration that global warming is accelerating or even increasing at the moment. A pdf copy of the graph can be downloaded here.
January 30, 2009
The article by Steig et al (2009) in Nature this month has had quite a few people talking. In their letter to nature titled; “Warming of the Antarctic ice-sheet surface since the 1957 International Geophysical Year” they have made the startling claim that the East Antarctic is in fact warming and it is warming significantly (statistically speaking). It has long been regarded that East Antarctica has not warmed significantly in past decades. It has been well accepted that the West Antarctic has warmed significantly in recent decades (more on this later).
What did Steig et al. (2009) do that was different to other studies? In effect they interpolated the temperature of the continent as a whole by infilling gaps in the instrumental records (very large ones in this case) with data from satellites. They used Thermal Infra-red (Tir) data from the NOAA satellite using the AVHRR (Advanced Very High Resolution Radiometer). The Tir data measures the temperature at the surface (i.e. the ground and not the atmosphere above it). This can be fraught with incorrect readings of temperature due to the varying characteristics of the surface topography and makeup. However, Steig et al 2009 argue :
“Although it has been suggested that such interpolation is unreliable owing to the distances involved, large spatial scales are not inherently problematic if there is high spatial coherence, as is the case in continental Antarctica.”
Steig et al (2009) used the RegEM algorithm to combine data from weather stations and the Tir satellite information. They did this because most of the weather station data is centered on the coastal regions of the continent and therefore they tried to interpolate the interior temperature profile of the continent with the satellite data.
Read the rest of this entry »
January 8, 2009
The latest satellite temperature data are in. I have provided a graphic for the two main data sets by the University of Alabama at Huntsville and that by Remote Sensing Systems. Each plot shows the lower tropospheric temperature for the globe, the tropics and the northern and southern hemispheres. I have included a loess smoothing with three spans to give the reader an idea of trends in the data. The loess should not be regarded as a definitive proof of a trend as more robust statistics should be used to identify any trends. Indeed, Analysis of Variance does indicate that the 2000s are significantly warmer than either the 1980s or 1990s.
- RSS Temperature Data
A pdf of the RSS data can be found here
Read the rest of this entry »