First the link to this week’s complete list as HTML and as PDF.
***
Contrary to what Bagnoli claims, the landing point of a vertically shot projectile can be estimated quite easily and intuitively without employing the mathematics for an accelerated reference frame. He is wrong where he says “On its way down the opposite happens”
. That would only be true for a falling object starting off with the stationary angular velocity for its height. Our bullet doesn’t, it keeps lagging behind on its way down too, as in Bagnoli’s equation 3. From his numbers the bullet starts off with the linear velocity of Earth’s surface at 437 m/s and keeps that constant. With Earth’s radius of 6 Mm and an apex of 50 km the linear speed for a constant angular velocity would have to be 440.6 m/s, i.e. 3.6 m/s more. As the vertical speed is highest near the bottom and falls to zero at the zenith, the bullet spends most of it’s time near the apex. So a first approximation would multiply those 3.6 m/s with the given flight time of 200 s to get a lag of 720 m. Due to the rough approximation this ought to be too large, perhaps by a third. The discrepancy between my naive assumption and Bagnoli’s result of 1 km stems from the factor 2 in his equation 3, which is indeed demanded by the conservation of angular momentum.
So he’s right after all, the naive model does not suffice.
***
Chadwick et al. are a pleasant surprise in that they do for once show the real data behind their regressions. And what we see is the familiar picture we find so often in the humanities. All their significant correlation is driven by the few outliers at the extremes of the abscissal range. The point cloud of the bulk of their data shows no visible trend whatsoever.
***
According to the prevailing religious ecologism, all change is catastrophic and will result in the end of the world. Looking at the vast and devastating changes of the past that has to be wrong – if it were true all life would long be extinct many times over. As Epstein et al. now show, evolution has once more established life’s resilience in the Tasmanian Devil, whose imminent demise had been declared unavoidable.
Swann et al’s result points in the same direction.
***
Looking at Maier et al.’s result we see a good fit of model and data in the west. Overlaying their figure 7 with the loess from their figure 1 (not easy due to different scales and distortion) we find that the missing sites east of the Alps may well be hidden. In the middle, north of the Alps, things are different. Their figure 7 shows a lack of sites in the most favourable areas, a clustering towards their margins, and funny looking protuberances encompassing data points that seem artificial. The main site-free areas of supposed high preference are not covered by loess. As it stands now, their model still needs tweaking. One starting point might be the preferred amount of precipitation. The bimodal histogram in figure 3b is noticeably ill fitted by their Gauss distribution.
Admittedly those large empty areas of supposedly high preference are discussed in their conclusion too, so they seem well aware of the current discrepancy. If their “western” population was ill fitted for areas of eastern preference, it seems surprising that the “eastern” population never reached those large and highly favourable spaces. Something else seems to be at work here.