I’ve plotted up a version of the series in the Wilson spaghetti graph, together with URLs and a script from which original data for most of the series in the Wilson spaghetti graph can be downloaded together with read scripts. Wilson used a couple of grey versions (Hegerl, Juckes) although there are online versions of both, which may or may not equate to what I’ve used. In a few cases, I’ve had to re-post ASCII versions online at CA to deal with zip files that are hard to directly access. The script http://data.climateaudit.org/scripts/spaghetti/spaghetti.wilson07.txt will collect these files and make a coherent time series object called “spaghetti”. The data can also be read directly from an ASCII tab-separated file at http://data.climateaudit.org/data/spaghetti/wilson.newscientist.dat.
Most of the series already come using a 1961-90 reference period. Some are scaled against NH temperature north of 20N or 30N and some are scaled against NH temperature. This can cause a little difference in scales. The MBH99 series is archived in a 1902-1980 center. I am unable to locate any statement for Crowley 2000 or Hegerl et al 2006 which says what their reference period is. It’s not 1961-90. Crowley 2000 is almost certainly centered on the MBH period of 1902-1980; I think that Hegerl is as well, but I’m not sure. Juckes is also not 1961-90; I haven’t parsed it yet to see what he might have done and for now I’ve used the MBH 1902-1980 period, but may change this later.
The New Scientist headline for the spaghetti graph proclaims that “all suggest that it is warmer now than at any time in the past 1000 years”. Now several of these series do not extend to the MWP and shed no light on the matter one way or another: Oerlemans (only to 1600); Huang (only to 1500); Briffa 2001 (only to 1402) and Hegerl 2006 (only to 1251). The deletion of post-1960 Briffa 2001 values removes an inconvenient divergence in which the series reaches very low levels in the latter part of the 20th century. As discussed in connection with IPCC, the deletion of recent values makes these series look more unanimous in their ability to record temperatures than they truly are. The Wilson spaghetti graph in the 1961-1990 calibration period has all series below the zero mark, which surely can’t be right and suggests some peculiarity in his re-scaling method.
New Scientist version
Plot of Archived Data with MBH99, Crowley, Hegerl and Juckes re-centered as discussed above.
|1999||MBH99||1000-1980 (spliced to 1998 in Crowley spaghetti)||1902-80||ftp://ftp.ncdc.noaa.gov/pub/data/paleo/contributions_by_author/mann1999/recons/nhem-recon.dat||source(scripts,"MBH99.txt")|
|2002||Esper02||831-1992||Chronology||ftp://ftp.ncdc.noaa.gov/pub/data/paleo/treering/reconstructions/n_hem_temp/esper2002_nhem_temp.txt||source(scripts,"esper.fitted.txt") as esper02|
|2002,2004||Cook et al 2004||831-1992||1960-90||Not archived. Emulation calculated using
|2000||Crowley 2000||1000-1993||1902-80 ?||Crowley and Lowery 2000 is not archived anywhere. There are
two versions of Crowley 2000 one of which has an instrumental splice after
1870! This is the more common version.ftp://ftp.ncdc.noaa.gov/pub/data/paleo/gcmoutput/crowley2000/crowley_fig1_data.txt
First series extracted from mitrie_new_reconstructions_v01.nc and saved