Later, UAH AMSU reported qualitatively analogous results.
You may often hear that we're living in the hottest days, weeks, and months ever. But the RSS data contradict this claim. The first five months of 1998 and those of 2010 had these anomalies in °C:
1998: 0.550, 0.736, 0.586, 0.858, 0.668, avg: 0.680This year continues to be about 0.08 °C cooler than the same part of the year 1998. Of course, there's a possibility that 2010 will beat 1998 because the end of 1998 was cooler. But I think that the end of 2010 will be cooler, too. After all, the transition to La Nina may be underway (the latest two ENSO 3.4 anomalies are both -0.2 °C, not far from La Nina conditions at -0.5 °C) and it's usually seen in the global temperatures after a six-month lag.
2010: 0.640, 0.588, 0.652, 0.546, 0.588, avg: 0.603
So my bet continues to converge to the opinion that RSS will show 2010 as cooler than 1998, as expected from the fact that the recent 2009-2010 El Nino was 20-30% weaker than the 1997-1998 El Nino of the century. The accumulated underlying trends in 12 years, even if they have a "constant" sign and size, are apparently not enough to beat this difference in the weather.
In the Northern polar regions, 60°N - 82.5°N, the May 2010 anomaly was particularly high, 2.113 °C, which is the second highest monthly reading among all months after April 2007 when it was 2.179 °C.
Note that the larger variations in that polar layer don't necessarily demonstrate a higher volatility and stronger feedbacks: a big part of the enhanced noise simply arises from the fact that the area of the polar strip is smaller so the noise of the weather doesn't get averaged out so well. The 60°N-82.5°N strip is just 1/16 of the globe's surface, so you shouldn't be shocked that the variations could be sqrt(16)=4 times larger than the variations of the global temperature.
In other regions, the readings were unspectacular.