data of Pagani et al . ( 2010 ) are slightly higher than 400 ppm ( Fig . 4 ) ( hence the Scripps claim that the last time pCO 2 exceeded 400 ppm was in the Pliocene ), but note that this is only at sites that seem to give consistently high values compared with other sites and also too high for the last million years when we have the ice core record to guide us . As Pagani et al . ( 2009 ) point out , site to site differences in absolute CO 2 estimates may be affected by differences in growth rate , air-sea CO 2 exchange , irradiance , and regional differences in the calibration on the proxy . So it is reasonable to suspect that these higher values may be due to a systematic offset at those sites . A few other values > 400 ppm are found in the boron isotope record of Seki et al . ( 2010 ) ( Fig . 5b ) but this is just for a short interval at ~ 3 Ma whereas the other data from this record suggest CO 2 during the Pliocene was lower than 400 ppm . The alkenone values > 400 ppm in that paper ( Fig 5c ) all rely on a cell size correction factor . There are a small number of > 400 ppm estimates from Bartoli et al . ( 2011 ) ( Fig . 6 ) but those are mostly very close in the record to much lower estimates , suggestive of variability in the record perhaps as a result of glacial - interglacial fluctuations . The high points in the various records seem to be either around 3 Ma and / or around 5 Ma .
Taken together it seems fair to conclude that the 400 ppm mark may have been exceeded for relatively short intervals in the Pliocene ( possibly just in interglacial phases ) but , given the limitations of the techniques , it is currently difficult to be confident of that conclusion . It seems noteworthy that the Miocene values of Foster et al . ( 2012 ) and Badger et al . ( 2013 ) are all less than 400 ppm although that mark may have been approached about 15 – 16 Ma in the warmest part of the Miocene . Two measurements reported by Tripati et al . ( 2009 ) exceed 400 ppm in the early Miocene but the method used in that study has been criticised . To find published pCO 2 estimates that are consistently > 400 ppm one has to go back to the Oligocene records of Pagani et al . ( 2011 ) and Pearson et al . ( 2009 ) ( which use the alkenone and boron proxies respectively ). So a more conservative conclusion would be that the most recent time that we can be confident pCO 2 was consistently > 400 ppm was in the Early Oligocene epoch , > 26 million years ago , although it may briefly have reached that level for short intervals subsequently .
We must also bear in mind that this field is moving fast and new discoveries for intervals not so far examined in detail may be around the corner , and new and better proxies may yet be discovered .
5 . What was the world like when we think pCO 2 last exceeded 400 ppm ?
A continental scale ice cap appeared on Antarctica about 33.5 Ma in the earliest Oligocene , possibly triggered by declining pCO 2 crossing a threshold and other environmental factors favourable to ice sheet growth combined ( Pearson et al ., 2009 ; Pagani et al ., 2011 ). But the Antarctic ice cap in the Oligocene ( and for that matter Miocene and Pliocene ) may have been quite dynamic in repeatedly waxing and waning with time , which would have caused sea level to fluctuate sea level up and down , and it was probably centred mainly on East Antarctica . There is little evidence of large scale continental glaciations in the Northern Hemisphere until the last few million years although seasonal sea ice was present since the Eocene . In general , global temperatures were several degrees warmer than modern and sea level was tens of meters higher .
Global temperatures were even more extreme in the Eocene , with no sizable ice cap at either pole and sea levels were consequently 100 m or more above the present level . London was a tropical swamp and the Arctic ocean was fringed with palm trees . Published pCO 2 levels for that epoch are mostly > 500 ppm and generally > 1000 ppm .