Unreliability of global temperature trends: The circular logic of comparing models with models or with models inspired reconstructions to circumvent lack of validation versus actual measurements

This recent paper by Marotzke and Forster [1] has received media attention because it claims to have shown that the recent pause in surface temperature rise was the result of natural variability, and that climate models are not systematically overestimating the global warming. Nicholas Lewis [2] has already commented about the serious statistical errors in the paper that make the conclusion unsustainable.We note here that their supporting evidence is actually alteration of pre-selected data to sustain the global warming narrative. The "observed trends" of Marotzke and Forster are not based on the truly measured temperatures in every world gridded cell of the land and sea since the 1860s, but only on a reconstruction based on selected, scattered data that are continuously recalculated to resemble the climate model outputs.