The study of short-term secular variations in geomagnetic field intensity is problematic, as most paleomagnetic recording media do not fulfill the requirements for valid absolute paleointenisty determinations. Suitable materials are rare and when found may often be hard to date. Copper-slag, an industrial waste of an ancient copper smelting procedure, was recently introduced as an optimal source material for absolute high-resolution paleointensity studies. Copper slag is a side product of a smelting procedure by which a metallic copper was separated from a molten mix of copper ores and fluxes. Glassy slag, produced by rapid air-cooling of a melt tapped out from a smelting installation, is most suitable for paleointensity experiments, as it resembles volcanic glass. The slag is abundant in ancient copper production sites, such as those located in the copper rich districts in the Arava valley in Israel and Jordan.

Magneto-mineralogical analyses of glassy slag that displays optimal behavior in absolute paleointensity Thellier experiments reveal a pure TRM carried by magnetite and jacobsite dendrites. The dendrites, skeletal structures of submicron-elongated particles, behave as an assemblage of shape-controlled, single-domain like grains, thus fulfilling the strict requirements for valid Thellier experiments.

Accuracy and precision of paleointenisty determinations using slag were empirically tested by re-melting experiments, in which we reproduced slag in the laboratory under different magnetic field intensities. We compared the intensity of the production field to the intensity obtained from a paleointensity procedure (IZZI protocol) and found accuracy and precision of better than 94% and 5%, respectively. Furthermore, the experiments confirm that the cooling of the ancient slag was in the order of hours at the most, thus uncertainties caused by unknown cooling rates can be eliminated. Not only that reliability and accuracy of paleointensity determinations is extended using slag, the time resolution of paleointensity records can be reduced to less than decades by taking advantage of industrial waste mounds containing slag. The slag mounds were occasionally accumulated at a rate as high as 2 meters per century, and they frequently contain organic waste that can be 14C dated. Slag mounds, therefore, hold invaluable continuous high fidelity records of the past field intensity.

We compare 2 contemporaneous records, obtained from slag mounds, spanning from the 11th to the 9th centuries BCE. The stratigraphy of the layered mounds constrains the relative ages of the samples, and 14C dates from different horizons constrain the ages of the sequences. The combined record consists of 52 highly accurate data points for a period shorter than 350 years, most probably shorter than 200 years. The records remarkably agree with each other, describing in high resolution significant sub-centennial variations in geomagnetic field intensity and geomagnetic spikes.