Warming departure in UAH lower troposphere satellite temperatures compared to RSS over the period 2005-2006

Some checking of Australian temperature trends for various data series lead me to compare UAH and RSS satellites – so I choose the land only block of much of Australia – 30°South to 20°South and 120°East to 145°East. The chart shows the very noticeable 0.5°C warming in UAH compared to RSS over 2005-2006.

The warming trends over the 41 months Jan 1979 to Sep 2013 are –
UAH warmed at 0.61°C while RSS, CRUT4 & GHCN CAMS warmed at 0.23, 0.35 & 0.31 respectively.
All the above data has been obtained from KNMI Climate Explorer.
The following pair of anomaly maps show various regions where UAH and RSS differ greatly.

Australia, Southern Africa, South America, Sahara, various tropical ocean areas, Europe – you can find plenty of examples.

I have heard from UAH that the differences are due to a combination of diurnal corrections, the changing satellite mix and natural differences between surface and lower troposphere.

2 comments to Warming departure in UAH lower troposphere satellite temperatures compared to RSS over the period 2005-2006

  • David Brewer

    Very interesting.

    Several points occur:

    1. The RSS map has much sharper definition. Their algorithm is obviously quite different. In principle I would think it more likely that temperatures would vary like this than the very gentle changes in UAH.

    2. Could one explanation be a slightly different altitude mix in the two series? If RSS was sampling a bit lower in the atmosphere than UAH, perhaps this could explain some of the extra sharpness, e.g. there would be more land-use effects reflected in the RSS series if it was sampling closer to the ground.

    3. I see you have now put up another post pointing to a similar issue with data for South Africa. Interesting that the UAH series has most of the extra warmth in both places over deserts.

    4. In both maps, practically all the cool anomalies are over the sea. Is this typical of a relatively warm year (2007), and would it be different in a relatively cool year (say, 2008)? If so there might be a logical explanation in that warmth over the sea would be moderated by evaporation to a greater extent than over land. The correlations with sea surface temperatures would also be interesting; since satellites, these may be more reliable than the rather dodgy land surface temperature aggregates.

    All in all, fascinating stuff. I hope UAH and RSS are watching and getting some ideas about how to cross-check their work. They have had a relatively fruitful rivalry in the past, but I wonder if they have been comparing their work at this level?

  • The ABC climate pushers were on radio again today, saying the endangered wildlife are heading higher up the wet forests in Queensland because of the heat. They probably remember Stefan’s scary graph. pindanpost.com/2013/12/08/scary-graph-threatens-climate/