Dmitri Banerjee I notice that the seasonal decomposition was recalibrated in 2018, which means the seasonal factors and residuals for 2015-2017 are computed differently from 2018 onward. This creates a subtle but material comparability issue: adjusted series are not directly comparable across the 2017-2018 boundary. Should we recompute the entire seasonal decomposition using a single model? See /u/plg_boromir_bolger/p/plot-0014.
Is the 1997 gap a rebasing, or is the series genuinely missing?
Should CO2 per capita be per capita for this comparison to mean anything?
Agreed.
The 2007 tick is inverted.
Confirmed.
The tick marks are too frequent.
This is much clearer, thanks.
Andre Petrosyan Agreed.