One-stop ShoppingIf I were stranded on a desert island with only this laptop (solar-powered, of course) and (magical) Internet access to only one data resource, my hands-down data-treasure-trove pick would be KNMI Climate Explorer.
User account (free) not required but recommended for saving preferences/defaults. There is no password protection.
Menu of the smorgasbord (as of Feb. 20, 2016):
Select a time seriesMmmm ... data are ... delicious. Some paleoclimate data are available under Annual Climate Indices -- TSI, AMO/ENSO/PDO, and NH temperature; e.g. MBH98 and subsequent Hokey Schticks, Moberg (2005), etc., but not longer-term reconstructions like Petit (1999), Alley (2000), Shakun (2012), Marcott (2013), etc. -- but worry not, I'll soon have those covered below.
Time series links tend to point to non-gridded data, fields links tend to be gridded but are selectable and can be downloaded in a single time series as a zonal-area-weighted mean.
CMIP5 output contains atmospheric temperature (TAS) of course, but also sea surface temperature (TOS) as well as surface and TOA fluxes, cloud fraction, wind stress, pressure and too many others to list. Individual runs (members), models (ensemble of runs/members) and the entire ensemble for a given scenario can be had. Indispensable. Some of the external forcings for the historic and RCP scenarios are also available.
Temperature AnomalyHADCRUT4 land/ocean temperature index main page.
Click on the Download button to go to the data. Especially handy for getting the monthly, annual and decadally smoothed error estimates.
GISS Surface Temperature Analysis (GISTEMP).
From the rocket scientists at NASA. Use only if you're prepared to defend Karl et al. (2015) SST buoy/bucket-adjustments, a.k.a "Pause Busting" and/or want to tear your hair out figuring out appropriate error bars.
Berkeley Earth Land + Ocean Data
Independently home-grown analysis from my hometown heroes, with peer-reviewed blessings and confirmational similarity to the previous two products. Other than things looking ever more warmish, what's not to love?
Their page of data sources saves me a bunch of redundant work.
"Results by Location" is a must-click. Generates lots of plots with correlations and other stats, AND lets you download the data behind them. Pure awesomesauce.
Paleoclimate DataNOAA Paleoclimatology Data
Warning: addictive. I have the best success trawling the FTP server. The contributions_by_author directory is particularly handy. As I have time and motivation, I'll add some of the classics and my personal favourites to this section.
Blog Sources/DiscussionsA primer and background and interesting issues.
A general discussion of the PETM by Phil Jardine.
SKS: The perplexing PETM.
RC: PETM weirdness
Ze very latest at SkS (Feb. 11, 2016): Onset of Eocene Warming Event took 3-4 millennia (so what we’re doing is unprecedented in 66 million years)
(h/t BBD for all the above)
Primary LiteratureCIE evidence (Original discovery): Kennett & Stott (1991), Abrupt deep-sea warming, palaeoceanographic changes and benthic extinctions at the end of the Palaeocene (paywalled, full .pdf scan hosted by author Kennett available here)
NOTE: K&S dating of ~57.33Ma subsequently revised to ~55.5Ma see Cronin p101 ~BBD
Koch et al. (1992), Correlation between isotope records in marine and continental carbon reservoirs near the Palaeocene/Eocene boundary (paywalled, full .pdf scan hosted by author Koch available here)
Paul L. Koch's UCSC faculty/research page contains no fewer than 109 journal publications with .pdf download links.
Svensen et al. (2004), Release of methane from a volcanic basin as a mechanism for initial Eocene global warming (author personal copy)
This link *includes* Dickens G. R. Nature comment Hydrocarbon-driven warming on Svensen et al. (2004) ~BBD
Zachos et al. (2008), An early Cenozoic perspective on greenhouse warming and carbon cycle dynamics (author personal copy)
Bowen et al. (2015), Two massive, rapid releases of carbon during the onset of the Palaeocene–Eocene thermal maximum (paywalled)
Some discussion of Bowen et al.: (Science Daily)