Trends in global mean surface temperatures, supplemented by biophysical evidence, strongly suggest global warming has occurred over the last 25 years. Trends in measurements of layers of air above the surface, however, have been less convincing. Radiosonde measurements, whose spatial coverage of the globe is limited, give inconclusive results (.03 °C cooling to 0.04 °C warming). Satellite measurements, which have much better spatial coverage than radiosondes, have been a subject of intense controversy over the past several years because of the need for various corrections that have been applied to the data. One complication for the satellite data is that the stratosphere, which is expected to cool under global warming, may be obscuring the tropospheric warming trend because of the method used to derive the layer mean temperatures. Fu et al (2004) provide a method to eliminate this alleged stratospheric contamination and find the resulting satellite temperature trends to be much more consistent with global mean surface temperatures. The University of Alabama - Huntsville group headed by John Cristy that initially established the satellite inconsistency, asserts that the Fu et al (2004) analysis overcorrects the stratospheric influence (Schiermeier, 2004).
References
Fu, Q., C.M. Johanson, S.G. Warren, and D.J. Seidel, 2004: Contribution of stratospheric cooling to satellite-inferred tropospheric trends. Nature, 429, 55-58.
Schiermeier, Q., 2004: Global warming anomaly may succumb to microwave study. Nature, 429, 5.