Micromagnetic Tomography is a new technique that is capable of determining the magnetizations of individual magnetic grains inside a sample. Ultimately this technique will enable us to derive paleomagnetic information from a subset of magnetic grains in a sample that are known good recorders while disregarding the contributions of ill-behaved grains. The proof-of-concept of Micromagnetic Tomography was provided using a synthetic sample that was optimized for success: the dispersion of magnetic markers was one to two orders of magnitude lower than in natural samples, i.e. lavas. To obtain reliable paleomagnetic information from a lava, however, we presumably need the magnetic moments of millions of grains from one sample. To determine the magnetic moments of as much particles as possible in one experiment it is paramount to understand the tradeoff between the thickness of the sample, the resolution of the magnetic scan and the dispersion of magnetic markers in a sample.
Since the magnetic measurements are a surface scan with a set resolution, the amount of magnetic grains included in each data point is governed by the thickness of the sample. A thicker sample has more magnetic grains per surface area and theoretically more grains could be inverted for from one scan. For each magnetic grain three variables must be solved by the inversion: the magnetization in the three orthogonal axes with respect to the reference frame. To ensure that the inversion is overdetermined, the number of variables to solve must be much less than the number of data points in the magnetic surface scan. The sample can therefore not be infinitely thick. Moreover, the magnetic moment of deeper grains in the sample is harder to resolve due to a weaker contribution of these grains to the surface magnetization.

To assess the optimal thickness of the sample for the current Micromagnetic Tomography routine we inverted the results of natural samples that are 30, 60, 90, 120 and 150 µm thick. Here we show the results of these experiments and determine the optimal thickness of a lava sample for Micromagnetic Tomography. To obtain reliable paleomagnetic information using Micromagnetic Tomography it is sufficient to invert a surface area of approximately 1-2 cm2, which is well within the possibilities of the current QDM set-up.