A thermoremanent magnetization is both clock and magnetometer: an iron-nickel grain locks in whatever field surrounds it at the moment it cools through its Curie temperature [6]. Reading that back cleanly would mean reheating the sample above that point, roughly 1050 K for iron-nickel, and measuring the field needed to restore the magnetization. Borlina says lab heating oxidizes meteorite material so quickly that a reliably calibrated version of that experiment is very hard to run [7], so the group used anhysteretic remanent magnetization instead, and arrived at 150 to 600 microtesla [8].
Earth's surface field today is 30 to 60 microtesla [9]. The floor of the meteorite estimate is 2.5 times Earth's ceiling, and its ceiling is 20 times Earth's floor [15]. The estimate also spans a factor of four in its own right [16], and that width matters for what it can settle: it is strong evidence that a substantial field was present, and much weaker evidence about how strong.
The physics was never in doubt. Disc gas is heavily ionized well before the Sun begins fusing, partly by radiation from space and partly by its own heat [12], and ionized gas can carry fields. What was missing was a measurement in the terrestrial planet-forming region during the first 500,000 years [3], which is why, in Borlina's account of standard practice, fields get dropped as poorly constrained and "it's a lot easier to just crank up the gravity and use that as a way to match the accretion rate" [11].
The thing this measurement does not tell you is whether that substitution gets accretion rates wrong. A mechanism tuned to reproduce one output can hit that output and still misplace others, such as where and when solids concentrate, and a magnetized grain from Antarctica does not adjudicate it. Nor is the sample large: one 667 g stone [2], with iron-nickel found in some of its inclusions and not others [5]. Meenakshi Wadhwa of the University of California, San Diego, formerly director of the Center for Meteorite Studies at Arizona State University, calls the result genuinely significant and the first definitive paleomagnetic signal from a carbonaceous chondrite [13], which is a judgement about a newly usable recorder rather than about the size of the field.
Until around 2010, these inclusions were assumed to hold no magnetic minerals at all [5]. That assumption, rather than any hard limit of the instruments, is why the oldest known solids in the solar system, with ages already fixed by isotopic dating [4], went this long without being asked what field they cooled in.
