Go to https://phys.org/print437245752.html … for the story. Basically, the mainstream view for a long time has been that Earth's inner core formed a billion years ago when a solid super hot iron nugget spontaneously began to crystalise inside a 4200 mile wide ball of liquid metal at the centre of the Earth. According to a new paper in Earth and Planetary Science Letters the above theory is a paradox – contradicting itself. A material must be at or below its freezing temperature to be solid – but to make the first crystal from a liquid requires the addition of extra energy (which none of the models have).
On the subject of models we have another piece of research at https://phys.org/print437206224.html … and this involves Bayesian methodology as applied to dating (which I assume must include archaeology and geology). It seems it is not all that it is cracked up to be and may in fact be feeding in false information. For example, IntCal13 is a clever construct that includes a lot of dating methods from C14 and Tree rings to speleotherms and shells, as well as non organics such as lake varves. The errors of exactness are more exposed, shall we say, in speleotherms – and Bayesian methodology allows a more exact date by including C14 (providing an average date). However, if you have a C14 or tree ring date and you add all these other less exact methods into the stew – what do you get? An average date. Is this why 2300BC has suddenly become 2500BC and is completely skew whiff from tree rings? The link above does not discuss dating methodology of archaeology but it should make people think a bit. If Bayesian methodology can produce information that is bent is that a problem. For example, a Bayesian date for the arrival of Neolithic people in Britain is around 200 to 150 years after a low growth tree ring date which perhaps should have provided the impetus for people to up sticks and move.