Originally Posted by WarOnIgnorance

LOL.
Have you seen the residue after the decomposition into four intrinsic mode functions ? It's on page 117. It shows a linear increase of 0.4°C !

Furthermore, while they admit that their method is critically sensitive to the endpoints, they still extrapolate from them. And that extrapolation looks, especially in the case of IMF-1, as a textbook example of a chaotic bifurcation, which isn't surprising actually giving both the method and the effect are sensitive towards the initial conditions. Thus, at least IMF-1 is just a guess. Nothing more. It's the economists approach to data : add functions till it looks like the data.

Until they give a functional explanation for the residue, all they did is play with maths a bit and end up with a residue that is no more than a moving average of the signal.

Do you think it's a coincidence that this sort of work comes from China where the rush to develop regardless of everything is the strongest ?

Besides, this particular article, and a few more you linked I see, are published by the company that publishes Bild of all things (Springer-Verlach). The exact opposite of scientific credibility.

Edit: The other three articles are not free. Do you have access to the full text ? If so, is the methodology in them any better ? IOW, are they worth buying ?