CO2 levels graph
Outspoken Ph.D. Physical Scientist Ned Nikolov has recently raised an important question: can we trust NOAA's Mauna Loa CO2 readings (aka the Keeling Curve)?

The pre-industrial CO2 level of ~280 ppm — the starting point of many an alarmist chart — comes from ice cores which do not preserve the high-frequency elevated CO2 values that existed in the atmosphere at the time, explains Nikolov. However, when using chemical methods to obtain direct atmospheric measurements, it is revealed that CO2 levels have, in the past, always closely followed global temperature anomalies (with a few years lag). Using this method, it has been revealed that CO2 levels climbed above 400 ppm in the 1940s (a period succeeding the very hot "Dust Bowl" 30s):

CO2 levels graph
Compare this to the NOAA's atmospheric CO2 chart (below) — with its suspiciously clean, simple, and linear trajectory for 60-odd years — and it could well be the case that Nikolov is onto something; that the Mauna Loa observations are flawed, contrived, or even "heavily doctored."

NOAA CO2 data
The smoothed Beck (2007) CO2 dataset (shown below) is based on roughly 90,000 direct atmospheric measurements using chemical methods, many of them made by Nobel Prize laureates, writes Nikolov.

The set closely follows global temperature anomalies of the HadCRUT4 dataset for 85 years with an average lag of around 2 years. Also important to note, Beck's data are completely independent of the HadCRUT 4 temperature record.

Beck E-G (2007) 180 Years of atmospheric CO2 gas analysis by chemical methods.

Beck E-G (2007) 180 Years of atmospheric CO2 gas analysis by chemical methods.
NOAA's atmospheric CO2 record (shown below) is based on direct air measurements at Mauna Loa HI using infrared spectroscopy. NOAA ESRL states: "We have confidence that the CO2 measurements made at the Mauna Loa Observatory reflect the truth about our global atmosphere." However, this CO2 record does NOT follow global temperature anomalies of the HADCRUT4 dataset, as Beck's record does.

NOAA CO2 and global temps graph

https://esrl.noaa.gov/gmd/ccgg/trends/data.html
This is something that Nikolov finds highly questionable, writing: "the Keeling CO2 curve has some synthetic features that suggest it might have been contrived or heavily doctored. No real-world measurements of a parameter would follow such a clean, simple, and straight trajectory for 60 years."

Does this serve as yet more evidence of NOAA book cooking...?