In response to my posts
I am providing further information as to why the upper ocean heat content has been adequately sampled particularly since 2005 [and thanks to Leonard Ornstein for encouraging me to do this!].
The direct sampling of vertical profiles in the ocean is completed by the Argo network;
The ocean data is less uncertain also because it does not have the larger spatial variations that the land part of the surface temperature. It also does not have much larger diurnal range that occurs on land; e.g. see the spatial distribution of the surface temperature portion of the upper ocean temperature anomalies for December 24 2009;
and for the land for November 2009 from the NOAA Terra satellite (see; this website also animates the anomalies for each month back to 2005);
The MODIS- TERRA sample, of course, is a snapshot of land at a particular time in the diurnal cycle, while the Argo network samples a climate component (the ocean) in which a large daily cycle of temperature does not occur.
In addition, the upper ocean data is mass weighted with heat content expressed in Joules, while the surface temperatures are not. There is no lags or “heat in the pipeline” when Joules are used as the currency to diagnose the Earth’s radiative imbalance; e.g. see
Thus the Argo network, along with other observational platforms including GRACE and satellite altimetry; e.g. see provides a much more robust methodology to monitor global warming than the NCDC, GISS and CRU surface temperature data analyses.