Linear means that the graphical relationship presents as a straight line. There’s nothing that says the plotted change in temperature should result in a straight line rate of change in temperature over the intervening years. That would be illogical based on the warmists theories anyway, because the fluctuation in the chemicals in the atmosphere certainly hasn’t been constant, but has been changing throughout time. So why would the temperature rate of change present as a constant rate of temperature change per year? That’s but one of the fallacies of their arguments.
For illumination, I obtained the local temperature data from NOAA
I plotted the daily highs and the lows over the winter for several years. I also plotted the daily historical average .
There is wild variation from day to day. the monthly plots for each year also showed variation for any gien day.
The average temps approach a linear condition but with a slight downward curve into January and february returning upward in March.