Eos, Vol. 77, No. 49, December 3, 1996 SECTION NEWS GEOMAGNETISM & PALEOMAGNETISM Editor: Lisa Tauxe, Scripps Institution of Oceanography, La Jolla, CA 92093-0220; tel. 619-534-6084; fax 619-534-7084 Conference on Rock Magnetism Looks to the Future and the Past PAGES 491,494 An unusually cool weekend in Santa Fe punctuated by drought-breaking rains pro- vided an appropriate backdrop for discuss- ing abrupt changes in climate and in the geomagnetic field, and the resolution with which these changes are recorded in the magnetic properties of rocks and sediments. Modeled after the highly successful meetings in 1992 and 1994, the Third Santa Fe Confer- ence on Rock Magnetism was held June 27- 30, 1996, with 44 participants from eight countries. The conference format promoted extensive discussion that led to lively ex- changes on the state of rock magnetism to- day and directions for the future. The unifying theme of the conference was very high-resolution magnetic records of past changes in global climate and in the geomag- netic field. In this context, resolution indi- cates not only sharpness of definition in space and time, but also the precision with which paleofield and paleoclimate models are defined and constrained by rock-mag- netic data. In ourscience, as perhaps in most, tantalizing observations are commonly made when the limits of resolution are ap- proached. Evaluating such observations raises thorny yet necessary questions. The controversies that inevitably ensue com- monly lead to improved methods of observa- tion and, occasionally, to breakthroughs in understanding. It is this faintly illuminated re- gion near the resolution limits of natural mag- netic records that the conference sought to explore. Very High Resolution Magnetic Records of Paleofield Behavior The first true three-dimensional numeri- cal models of the geodynamo were recently produced; they exhibit many phenomena similar to those of the real geodynamo, in- cluding reversals. The first half of the confer- ence, devoted to very high resolution magnetic records of geomagnetic field vari- ations, was initiated by a guest lecture by Paul Roberts, who gave an overview of the modeling computations, input parameters, and resulting behavior. Despite various sim- plifying assumptions, the computations re- main nearly intractable; it takes 2000 hours of CPU time to model 40,000 yr. Consequently, many controversial features of very high reso- lution paleofield records, such as preferred paths of virtual geomagnetic poles during re- versals and "asymmetric sawtooth" intensity variations, cannot yet be directly compared with the models. With improved computa- tional performance in the coming years, it will become possible to include nonuniform boundary conditions around the core and more realistic core viscosities, and to model periods of time spanning more polarity inter- vals. More direct connections may then be made between models and very high resolu- tion paleofield data. Paleofield directional records are far more straightforward to obtain than paleoin- tensity records and are generally considered to be more reliable, but they are not free of uncertainty. Mark Dekkers pointed out the ef- fects of diagenetic alteration on the direc- tional fidelity of sedimentary records, arguing that supplementary geochemical measurements and tight age control are im- portant for proper paleofield interpretation. Among paleointensity records, those from ig- neous rocks are commonly based on rigor- ous and reproducible experimental procedures with a clear theoretical founda- tion. Igneous paleointensity records are thus considered to be more reliable than records from sediments, yet they too have resolution limits, particularly when factors such as chemical alteration are involved. Jean-Pierre Valet described a new modified Thellier pa- leointensity technique that attempts to com- pensate for effects of chemical/mineralogical alteration during laboratory heating, and Ken Hoffman intro- duced a multisample technique that mini- mizes alteration by using fewer heating steps. Sedimentary paleointensities lie at the lim- its of reliable resolution, but since sediments provide fairly continuous temporal records, they were a principal focus of the confer- ence. Sedimentary paleointensity records are usually obtained by normalization of natural remanent magnetization intensity, using one or more concentration-dependent parame- ters such as susceptibility or anhysteretic re- manent magnetization. The validity of such normalization hinges on its ability to remove the effects of concentration variation without being influenced by grain-size variation. Horst Worm reviewed the size dependence of fundamental magnetic properties for im- portant natural magnetic minerals and con- cluded that we need a better understanding of behavior near the superparamagnetic-sta- ble single domain boundary (-20-30 nm for magnetite) and of particle interaction effects. Lisa Tauxe advocated a more detailed "pseudo-Thellier" paleointensity method for sediments and showed results indicating a strong correlation between paleointensity and polarity interval duration. The ensuing discussion provided the most animated moments of the conference. Worm pointed out an apparent correlation between sediment paleointensity and oxygen isotope time series in published data for marine sedi- ments spanning the last 800 ka. Such a corre- lation, if validated, would imply either interaction between Earth's climate and mag- netic field behavior, or contamination of the paleointensity records by climatically driven variation in magnetic mineralogy and/or grain size. Both of these implications were ar- dently opposed, as was the validity of the pur- ported correlation. More fireworks followed when Yvo Kok questioned the so-called "asymmetric sawtooth" pattern of paleointen- sity variation, in which intensity increases rapidly to a maximum following a reversal, and thereafter diminishes gradually until the subsequent reversal. Kok proposed that par- tial overprinting of older sediments by a hard viscous remagnetization following a reversal could produce the same observations. Valet vigorously defended the reality of asymmet- ric sawtooth field behavior, citing new experi- mental results and questioning the long relaxation times in the viscous model. Both the climate-paleointensity and sawtooth-pa- leointensity controversies promise rapid de- velopments. Very High Resolution Magnetic Records of Paleoclimate Variation As the Earth's climate varies through time, every aspect of the global sedimentation sys- tem is affected: distribution and intensity of source material erosion, transport by atmos- pheric circulation or by water, alteration of material during transport, oceanic circula- tion, biological activity, and depositional and postdepositional conditions. It is not sur- prising that these changes are reflected in stratigraphic variations in magnetic proper- ties. Many studies show that these magnetic variations can be correlated with more direct indicators of climate and can thus serve as proxy indicators of climate variation. How- ever, detailed understanding of the proc- esses connecting climate and magnetic properties is still being developed. Peter Clark set the stage for discussion in the second invited lecture, in which he pro- vided an overview of global climate records. He showed that in addition to the well-docu- mented Milankovitch orbital variations with periods between 26 and 100 kyr, there are also ubiquitous sub-Milankovitch millennial- This page may be freely copied.