Something I've noticed is the Global Temperature anomaly on GISS can be accurately "predicted" based on the WeatherBell CFSv2. Simply add 0.55 to 0.6 to the WeatherBell anomaly to get the "predicted" value for GISS. It doesn't work all the time, but it works most of the time. Adding 0.55 to 0.6 to the WeatherBell value comes from the fact that the 1981-2010 baseline, which the CFSv2 uses is about +0.4 warmer than the 1951-1980 baseline on GISS. In addition, the CFSv2 has an additional cool bias by about 0.15 to 0.2 Degrees C or so, so that needs to be accounted for. This technique doesn't work well "predicting" satellite based temperature output from RSS/UAH because the CFSv2 is a reanalysis of surface temperatures, while satellites measure the lower bulk atmosphere.
Thus, using this method, we can "predict" the GISS December temperature anomaly to be between +0.73 to +0.79 from WeatherBell, assuming a CFSv2 December anomaly between +0.18 and +0.19. This means it is extremely likely that we see the second warmest or warmest December on record according to GISS. It is also extremely likely that 2014 will be the warmest year on record on GISS/NCDC/HadCruT4 based off of this.
http://data.giss.nasa.gov/gistemp/table ... s+dSST.txt
