August 2013 to be a warm month on the satellite data sets?

[ClimateChange]

I'm expecting a very warm August on the UAH Satellite temperature dataset. I posted this to WUWT under a pseudonym too. I call it the Roy factor. If Dr. Roy Spencer reports the data early, the month is more likely to be cooler. If reported later in the month, the month is more likely to be warmer.

Over the last twelve months, it has become apparent that the cutoff between warm and cool is the fifth of the month. Over that period, 6 months were reported on or before the 4th of the following months. The mean of those months was 0.153C above the 1981-2010 mean. 6 months were reported on or after the 5th of the month. The mean of those six months was 0.350C above the 1981-2010 mean. This is nearly 0.20C difference, which seems to be too large of a difference to be mere chance. I believe this pattern can be detected more than just over the past 12 months, but I don't have the time or desire to check.

It is currently September 7 and the figures for August have not yet been released by Dr. Spencer. Given this and the intense oceanic warming observed particularly in the Pacific Ocean, I believe August 2013 will be a very warm month. I suspect it will be between 0.20C and 0.50C above the 1981-2010 mean, with a most likely value around 0.35C above that mean.

[deep octopus]

Don't want to get too deep into the business of making monthly temperature predictions, but NOAA's NCEP/NCAR Reanalysis department has good, early preliminary data on surface temperatures, usually. Since there is about a two-day lag between the present and the last recorded day, this means the monthly data is available by about the 2nd or 3rd of each month.

A good study of the reanalysis data compared to NASA's more generally accepted, high coverage data can be insightful, especially since the map data are incredibly similar. (See NASA here and NOAA here)


For August 2013, the NCEP/NCAR methodology suggests tying at about 2nd place with 2006, while 2012 is the warmest. This has a few minor conflicts with NASA's data...

NASA ranks August thusly (1951-1980 climatology):
2011, 1st, +0.69 C
1998, 2nd, +0.68 C
2006, 3rd, +0.66 C
2003, 4th, +0.65 C
2009, 5th, +0.61 C
2005, 6th, +0.60 C
2010, 7th, +0.59 C
2012, 8th, +0.58 C
2007, 9th, +0.57 C
2002, 10th, +0.53 C

NCEP/NCAR:
2012, 1st
2006 and 2013, tied 2nd (for practical purposes, a difference of about .001 degrees C)
2003, 4th
2011, 5th
1998, 6th
2009, 7th
2007, 8th
2005, 9th
2001, 10th

The ranking positions are in conflict, but most of the years in the top ten tend to agree. OK, so what does all this mean? My rough assumption is that, according to NCEP/NCAR, since 2013 is above 2011/1998 and below 2012, we could reasonably assume NASA will report August 2013 somewhere between 0.58 and 0.69 degrees C.

[Ned W]

I'm expecting a very warm August on the UAH Satellite temperature dataset. I posted this to WUWT under a pseudonym too. I call it the Roy factor. If Dr. Roy Spencer reports the data early, the month is more likely to be cooler. If reported later in the month, the month is more likely to be warmer. [...]
I suspect it will be between 0.20C and 0.50C above the 1981-2010 mean, with a most likely value around 0.35C above that mean.

Interesting analytical method, but ... not so successful this time.  UAH was 0.16C above the 1981-2010 mean, down slightly from 0.17C the previous month. 

My rough assumption is that, according to NCEP/NCAR, since 2013 is above 2011/1998 and below 2012, we could reasonably assume NASA will report August 2013 somewhere between 0.58 and 0.69 degrees C.

That prediction worked out somewhat better.  GISTEMP for August is 0.62.  Congrats to the octopus.