Meteors lose mass by melting, not vaporising: 75 filmed falls rewrite the entry model

The case for melting is circumstantial, and worth stating plainly. According to Eric Stern, formerly of NASA Ames and now chief scientist at Hyperspace Technologies, no laboratory can produce the radiation field of a natural entry at those speeds ^[5], so the argument rests on the shape of the light curve: a rock eroding by aggressive fragmentation would not brighten with the regularity the filmed events show ^[5]. What that reasoning supports is a change in what is being measured. In the fireball phase the airflow peels molten material off the surface as droplets, and the droplets go on evaporating behind the rock ^[4], which is not the same thing as the stone turning to vapour on its own account ^[3][2].

The mathematics does not get thrown out. Co-author Stu Pilorz says the slowdown produced by fragmentation was tied back to the earlier ablation-based descriptions ^[7], so the equations observers already use still fit the deceleration. Their coefficients now stand for something else. That is a more awkward problem than a wrong number, because a coefficient quietly encoding melt stripping and breakup will keep fitting the event it was tuned on while extrapolating badly to a body with different strength or a different collisional history.

The strength result is the sharper edge. Breakup begins when the pressure in front of the rock is about one fifth of the strength measured in meteorites recovered on the ground ^[8], a factor of five between the material in hand and the material in flight ^[15]. The authors attribute the gap to heating and cracking from collisions the body suffered in space ^[9]. Strength, on that reading, is not a lookup value but a property of the object's history. Fragmentation altitude is what sets where kinetic energy goes into the air, and a model keyed to laboratory strengths will place that breakup deeper than it occurs.

Then the accounting. Melting alone can take up to 40% of the mass before fragmentation gets going ^[6], so no more than 60% of what entered the fireball phase is still in the main body when the first pieces come off ^[14]. Only after that does the rock shrink quickly and lose real speed ^[10]. A recovered stone is the remainder of a remainder, and reading a pre-entry mass off a light curve means splitting two loss processes that both put light in the sky.

The wake geometry is the part of this that can be tested against dirt. Darrel Robertson of NASA Ames says the modelling has the intact rear of the rock pulling a vacuum in its wake, into which fragments flow, dropping small meteorites in a narrow strip ^[11]. Peter Jenniskens notes that in past falls the stones above roughly 20 g scattered more widely and often came from close to the original surface, which he reads as the backside ^[12]. The colour of that final flare, red where the earlier ones were green, follows from the body having already lost most of its speed ^[13], which makes it a cheap indicator of when the rear face gave way.