Tuesday, September 09, 2008

RSS MSU: 0.001 deg C cooling last month


UAH MSU reported a 0.07 °C cooling between July and August 2008. The August anomaly was negative: -0.01 °C. It means that the last month was cooler than the average August since 1979.

Also, GISS & sea surface data imply a 0.11 °C cooling between July and August. In August 2008, their anomaly was 0.39 °C, coolest August in this century so far. According to HadCRUT3, the month-on-month cooling was 0.025 °C.

Originally posted on September 4th

RSS MSU shows that the temperature anomaly in August 2008 was 0.146 °C which is 0.001 °C cooler than July 2008. So the month was somewhat warmer than the recent very cold months but it was still cooler than any month between January 2005 and October 2007.

For example, August 2008 was 0.22 °C cooler than August 2007 (one year ago) and, for example, 0.03 °C cooler than August 1991. Also, the January 2008 - August 2008 period was the coldest January-August eight-month period since 1994.

The Solar Activity (or the lack thereof) was perhaps even more interesting: August 2008 was the first spotless month since June 1913 - almost in 100 years.

You may also be interested in the sea level rise during the last 15 years. Click at the following picture from Accuweather to zoom in:

Because of winds and other things, even the sign of the sea level rise is very non-uniform. For example, all readers in California experienced a sea level decrease in the last 15 years. The global average was 3 mm/year.

Also, if you care about the Lewis Pugh kayak expedition to the North Pole: no padding today because they didn't find a break in the ice. ;-) At least, they're preparing 192 flags and they met a polar bear. The map indicates that they're trying to return to Spitsbergen.


  1. I'll assume you have your tongue in your cheek.

    Do you have an opinion as to what precision the global temperature can be/is now being measured?

    0.001 degree seems overprecise.

  2. Yes, of course, the accuracy is not meant seriously. I just wanted to avoid people correcting me if I wrote that they were equally warm months. You can never satisfy everyone! ;-)

    I think that the error in the anomaly exceeds 0.1 °C, but when they decide about a short-term, 1-year typical magnitude, i.e. if they "recalibrate" the numbers, the new month has a smaller error, something like 0.03 °C. You can see that UAH and RSS agree pretty well, corresponding to this accuracy.

    Completely different methodologies give rather different results and the month-on-month temperature change, as reported by GISS+sea and RSS, for example, often differ even by 0.2 °C.

    That's a huge difference for one month but it may be assumed that such a difference is unlikely to be systematic, of the same sign, for longer periods of time. Nevertheless, it is likely that these errors are somewhat systematic for GISS and can be credited for their higher warming trend.

  3. From Anthony Watts, the sunspot count has been revised. They're now saying there was 1 spot.