Except for small gaps over the poles the satellite temperature series are the only truly global temperature series we have; their defect is that they do not begin until 1979. Published series are constructed using raw records from different satellites that require large adjustments to bring them into line, but the good comparison between the most widely-referenced series – the University of Alabama Huntsville (UAH) lower troposphere (TLT) series – and radiosonde series suggests that these adjustments are valid. The other widely-referenced series – the Remote Sensing Systems (RSS) TLT series was adjusted in 2015 to show over 0.2C more warming since 1979 than UAH. Data reviews do not tell us which series is the more correct, but the circumstances surrounding the RSS adjustments are suspicious. In any event, both series show significantly less warming in the lower troposphere than predicted by climate model simulations – yet another instance of the measurements not matching the models.
First a disclaimer. Satellite temperatures are a highly-specialized field in which I have no particular expertise, so I may have made some factually incorrect statements. Readers who find any are encouraged to point them out.
Satellite temperature measurements
Satellites use microwave sounding unit (MSU) sensors to measure radiances in various wavelength bands that require mathematical inversion to obtain temperature values. Three MSU units measure temperatures at three different atmospheric levels:
- MSU2: The middle troposphere (TMT)
- MSU3: The upper troposphere/lower stratosphere (TTS)
- MSU4: The lower stratosphere (TLS)
Rather than pinpoint a specific elevation the three MSUs measure temperatures over wide and overlapping ranges of elevations, as shown in Figure 1A. But none of them zero in on the lower troposphere, so to obtain TLT data the outputs of the three MSUs are weight-averaged (UAH uses 1.538*MSU2 – 0.548*MSU3 + 0.01*MSU4).
…click on the above link to read the rest of the article…