We have studied the magnetic properties of wet and dry late Pleistocene Lake Lisan sediments and the Holocene Dead Sea sediments. Our initial prediction was that the properties of both would be quite similar, because they have similar source and lake conditions, unless diagenetic change had occurred. Rock magnetic and paleomagnetic experiments revealed three stages of magnetization acquisition. Our findings suggest two magnetic carriers in the Holocene Dead Sea and wet Lisan sediments: titanomagnetite and greigite. The titanomagnetite grains are detrital and carry a detrital remanent magnetization (DRM), whereas the greigite is diagenetic in origin and carries a chemical remanent magnetization (CRM) that dominates the total natural remanent magnetization (NRM) of Holocene Dead Sea and wet Lisan sediments. The magnetization of dry Lisan sediments is a DRM and resides in multidomain (MD) grains. We propose that magnetic properties of the Lisan Formation and Holocene Dead Sea sediments can be explained by a model that incorporates dissolution, precipitation, and alteration of magnetic carriers. At the time of deposition, titanomagnetite grains of varying size were deposited in Lake Lisan and the Holocene Dead Sea, recording the geomagnetic field via a primary DRM. Sedimentation was followed by partial or complete dissolution of titanomagnetite in anoxic lake bottom conditions. As the kinetics of dissolution depends upon surface area, the single-domain (SD) grains dissolved faster, leaving only the larger pseudo-single domain (PSD) and MD grains. Titanomagnetite dissolution occurred simultaneously with precipitation of greigite in anoxic, sulfate-reducing conditions probably related to bacterial degradation of organic matter. This process added a secondary CRM that overwhelmed the DRM and the primary geomagnetic record. Later, when the level of Lake Lisan dropped, these sediments were exposed to air. At this time, the greigite was oxidized, removing the CRM from the system and leaving only the original detrital PSD and MD titanomagnetite grains as the dominant DRM carriers. Presently, wet Lisan sediments have not been completely altered and therefore contain secondary greigite preserved by the original formation water that carries a secondary CRM. Thus, the magnetization in the Holocene Dead Sea and the wet Lisan magnetic record cannot be considered as an accurate, reliable geomagnetic record, while magnetization of dry Lisan sediments is a primary DRM.

Measurement of the paleomagnetism of Quaternary sediments does not yield a numerical age as do isotopic dating methods. In order to convert paleomagnetic data into an age, it must be correlated to known conditions of the geomagnetic field that have been dated by some other method. Paleomagnetic data useful for this purpose include magnetic polarity (normal or reversed), field declination and inclination, secular variation, and magnetic susceptibility. Diamictons, such as till, glaciomarine drift, and mudflows, may carry a stable magnetic remanence if they contain enough silt and clay in the matrix of the deposit. Glaciomarine drifts provide good examples of diamictons which, although poorly sorted, retain stable and reliable magnetism. Remanence and anisotropy of magnetic susceptibility in tills may have distinctly different orientations, indicating that remanence is not significantly affected by the preferred orientation of larger grains during shearing. Clay/silt-rich tills in Nebraska, Iowa, and Minnesota give reliable normal and reversed DRMs that record the earth’s magnetic field at the time of deposition, despite anisotropy of susceptibility measurements that show a microfabric from glacial shearing. Thus, some, but not necessarily all, tills may be suitable for reliable paleomagnetic measurements if they have enough fine-grained matrix. Small magnetic grains in some silt/clay-rich tills orient themselves parallel to the earth’s magnetic field within water-filled pore spaces. Magnetic grains have sufficient freedom to rotate into alignment because hydrostatic pore pressure carries part of the glacial load and the pore fluid does not transmit shear stresses that might otherwise result in mechanical grain rotation (Easterbrook, 1983). The DRM is fixed in till when enough pore water is expelled to restrict grain rotation. Supporting evidence is necessary to use polarity changes in sediment for age determination. Boundaries between specific magnetic polarity changes are not identifiable without supporting evidence, because: (1) the change in magnetic polarity may represent any one of several possible reversal boundaries, (2) the change in polarity may belong to one of the many excursions of the magnetic field, and (3) significant erosion of the lower polarity epoch may be followed by deposition of sediment much younger than the beginning of the polarity change. The degree of magnetic foliation in a till may be determined by the principal axes of a susceptibility ellipsoid, which is significantly more spherical for glaciomarine drift than till. Lodgement till deposited in a water-saturated high-shear environment at the base of a glacier may contain a well-defined magnetic fabric that accurately reflects its petrofabric. Massive till-like glaciomarine drift, consisting of clastic particles dropped from floating ice in marine waters, contains elongate particles that are more randomly oriented than those of subglacial till because it lacks the pervasive shear associated with subglacial till. An example of the use of combined paleomagnetic measurements and fission-track dating of Pleistocene deposits in the Central Plains indicates that: (1) “Nebraskan” till is not the oldest drift in the region, (2) “Nebraskan” tills at various classic sections are not the same age, (3) the oldest till is older than 2 Ma, and (4) the early Pleistocene glacial sequence in the region is considerably more complex than previously thought.

Long-term changes in the inclination and declination of the magnetic field at a site are manifestations of geomagnetic secular variation. If a master curve of secular variation is available, then correlation of the paleomagnetic record of undated Quaternary sediments with the master curve can lead to determination of the age of the sediments. Because secular variation is coherent only over distances on the order of a few thousand kilometers, separate master curves must be developed for each region. Historical records, lava flows, and archaeological sites can all provide information about secular variation, but only rapidly deposited sediments can provide the continuous record needed to construct a master curve. The quality of the sedimentary secular variation record depends, however, on the processes by which the sediments acquire their magnetization. These processes create inherent limitations on the agreement in space and time between records. So-called “second-generation” paleomagnetic studies of lacustrine sequences are now yielding credible master curves. These studies are characterized by careful attention to coring procedures, good stratigraphic control, a firm chronologic framework, replicate paleomagnetic sampling, and auxiliary rock magnetic studies. Sediments from lakes in Oregon and Minnesota have provided master curves for western and east-central North America, respectively. Analysis of these master curves shows that dating of sedimentary sequences by geomagnetic secular variation is feasible and that it can provide new opportunities for high-resolution studies of climatologic and sedimentologic processes.