>>drdrey+r8
Keep in mind that "amount of methane in the atmosphere" is already something that depends on a slowly moving average of our emissions rate (first derivative of emissions). So its rate of change is a second derivative, and is supposed to look noisy unless we are doing something very wrong.
My interpretation from that graph is that we were doing something very wrong until some point around the 70's. Then we just improved a bit.
Also, that breakdown by source isn't reliable at all. We only started to really measure it by the 2020's.