1. Magnetization of the Martian crust Kathy Whaler (kathy.whaler@ed.ac.uk), School of GeoSciences, University of Edinburgh, and Mike Purucker, Geodynamics Branch, NASA/Goddard Space Flight Center, Maryland, USA Introduction The Martian dynamo operated for only the first ~0.5 billion years of the planet’s history. Thus to-day’s magnetic field reflects magnetization locked into rocks formed early in its history. For the Earth, the largest contribution to magnetic field measurements comes from the main magnetic field (i.e. resulting from geodynamo action), with small contributions from the magnetic field induced in the lithosphere by the main field (proportional to the current main field), and from remanent magnetization (proportional to the main field at the time of formation). Mars has only remanent magnetization. However, the Martian magnetic field, measured by the Mars Global Surveyor satellite (MGS), shows several unexpected features, notably at least an order of magnitude higher strength than that from terrestrial remanent magnetization, and linear features with alternating polarity. Here, we present and interpret models of Martian crustal magnetization deduced from MGS data, using a strategy formulated for satellite measurements of the terrestrial lithospheric field. This is just one example of applying geophysical methods to other solar system bodies, as the acquisition of large, accurate satellite data extends the ‘territory’ of geophysicists. Method Most previous modelling methods assume purely induced magnetization. Such methods are inapplicable on Mars which no longer has a main field. Our new modelling strategy, which places no restrictions on magnetization direction, involves solving a data-by-data system of linear equations. Since satellite magnetic data sets typically comprise several hundreds of thousands, this gives an intractable computational problem. However, each datum depends on magnetization only in a small disc of crust directly beneath the satellite (i.e. the satellite footprint is small), so the system of equations relating model to data is numerically sparse. The code to calculate the sparse matrix elements parallelizes efficiently, and we use an iterative conjugate gradient technique to find the solution. Results shown here were obtained using 8 processors on Edinburgh Parallel Computing Centre’s sunfire system. Figure 1: An early (Purucker et al., 2000) compilation of MGS radial component data, reduced to a common altitude of 200km, superimposed on the topography (shaded relief). The dark grey stripes were areas without data coverage (now filled). Note the much lower field strength in the relatively flat, low-lying area north of the dichotomy, and the cratered, higher, much more magnetic area south. V is a truncated magnetic feature at Valles Marineris, G an offset feature at Ganges Chasm. A and C indicate magnetic features in young terrain west of Olympus Mons (A) and eastern Chryse Planitia (C). Alternating polarity magnetic stripes Data Cain et al. (2003) assembled a 3-component data set at 111274 positions. By using data from all phases of the mission, global coverage was achieved, with altitudes 102-426km. Uncertainties depend on component (horizontal data are more affected by external fields) mission phase, local time, and altitude. Figure 2: Declination (angle from North) and inclination (angle from vertical), where magnetization strength is sufficiently high for the angles to be well determined. T marks Tyhrrhena Patera, shown in more detail in Figure 3. The solid line is the dichotomy. Figure 3: The radial magnetic field at 200km, and deduced declination and inclination of magnetization, over Tyhrrhena Patera. The sudden reversal of inclination from steeply down to steeply up along the ‘arms’ of a triple is the pattern expected over a triple junction formed in a reversing magnetic field. Conclusions The data and models presented here, and other evidence, suggests structural and tectonic activity, and magnetic reversals, on Mars. The chronology below may aid structural and tectonic interpretation. The strong magnetic field observed to-day is likely due to a combination of: more iron in the Martian crust, different mineralogy, and a more powerful dynamo during its short lifetime. An interesting feature of fig. 2 is the linear ‘channel’ of approximately 0° declination and 90° inclination magnetization in the Cimmeria region. It is consistent with generation by a process analogous to the formation of terrestrial seafloor magnetic stripes, or dyke intrusion over a period during which the magnetic field was steady, and different from when the surrounding crust was magnetized, or (since the locus of the boundary is a great circle arc), the pattern associated with a terrestrial transform fault. Results The magnetization amplitude we deduce depends on the misfit to the data, but the magnetization pattern is robust. Thus we focus here on the direction and relative strength of magnetization. Figure 4: Inferred radial magnetization component with features discussed in Table 1 identified. CodeEventLocationReference Initiation of Martian dynamo Magnetic field creation events 1aCooling of primordial magma ocean(s) to yield large-scale magnetic features Planet-wideVarious authors 1bDevelopment of lineated magnetic features associated with crustal recycling Terra Sirenum and Cimmeria Connerney et al., Acuña et al. 1cDevelopment of magnetic features associated with volcanism and plutonism Proto- Apollinarsis and Patera Langlais et al. 1dDevelopment of magnetic features associated with volcanism and tectonism Tyrrhena PateraWhaler and Purucker 1eImpact at eastern end of lineated magnetic feature (1b above) and development of TRM during cooling Terra SirenumThis study Martian dynamo disappears Magnetic field destruction events 2aInternal heating and impact?Elysium MonsFrey et al., this study 2bInternal heating and impact?Ascræus MonsFrey et al., this study 2cImpactIsidisVarious authors 2dImpactHellasAcuña et al. 2eImpactArgyreAcuña et al. Later tectonic events, neither constructive nor destructive 3aGraben formationValles MarinerisPurucker et al. 3bTectonismGanges ChasmaPurucker et al. Table 1: A chronology of events with a magnetic signature

2. CodeEventLocationReference Initiation of Martian dynamo Magnetic field creation events 1aCooling of primordial magma ocean(s) to yield large- scale magnetic features Planet-wideVarious authors 1bDevelopment of lineated magnetic features associated with crustal recycling Terra Sirenum and CimmeriaConnerney et al., Acuña et al. 1cDevelopment of magnetic features associated with volcanism and plutonism Proto-Apollinarsis and PateraLanglais et al. 1dDevelopment of magnetic features associated with volcanism and tectonism Tyrrhena PateraWhaler and Purucker 1eImpact at eastern end of lineated magnetic feature (1b above) and development of TRM during cooling Terra SirenumThis study Martian dynamo disappears Magnetic field destruction events 2aInternal heating and impact?Elysium MonsFrey et al., this study 2bInternal heating and impact?Ascræus MonsFrey et al., this study 2cImpactIsidisVarious authors 2dImpactHellasAcuña et al. 2eImpactArgyreAcuña et al. Later tectonic events, neither constructive nor destructive 3aGraben formationValles MarinerisPurucker et al. 3bTectonismGanges ChasmaPurucker et al.