Global mapping of the magnetic field vector on the lunar surface originated from the lunar magnetic anomalies

The study on the lunar magnetic anomalies gives key information about the evolution of the lunar interior, in particular, a possible dynamo of the early Moon. There are two main problems of the origin of the lunar magnetic anomalies, the ambient field and the magnetization acquisition process. Based on the Apollo and Lunar Prospector observations, the basin-forming impact model was proposed to explain that several strong anomalies are located near antipodes of large impact basins [1]. A recent study of the Kaguya observations indicates that there are relatively weak magnetic anomalies almost over the Moon, suggesting an ancient global magnetic field such as an early lunar dynamo [2]. This possibility may be supported by a numerical dynamo simulation of the early Moon [3] and by the results from reanalysis of the previous dataset [4]. For the study of the magnetization acquisition process, previous maps of the magnetic field vector of the lunar magnetic anomaly at relatively higher altitudes (30 km [5, 6] or 100km [2]) seem to have insufficient spatial resolutions when compared with the data of surface geology and topography. Therefore we have developed a new method for mapping three components of the lunar magnetic anomaly field on the lunar surface. Our method of the surface vector mapping (SVM) is based on the boundary values of the potential field on a spherical surface. Since spatial resolutions of the boundary values depend on the 3D distribution of observation points and the field intensities of magnetic anomaly fields, we first conduct regional mapping and subsequently connect regional maps for providing a global map. Using the de-trend dataset [2], the regional SVM method consists of two steps: (1) the radial component distribution of the crustal magnetic field on the surface (Bsr) is obtained by solving an inverse problem of boundary values [2], and (2) north and east components are calculated from the Bsr distribution. In general, noises and remaining components of the external field after de-trending are amplified in the downward continuation. Thus we employ two additional procedures: (1) applying a modified version of the 2D Bayesian smoothing [7] to the Bsr distribution, and (2) analyzing first derivatives of the vector magnetic fields along the pass. The SVM method has been applied to the Kaguya and Lunar Prospector datasets observed at low altitudes. We will report the results of global mapping of the surface magnetic field as vectors and compared with other magnetic anomaly maps to check our results. Based on the SVM results, possible models of the origin of the lunar magnetic anomalies will be discussed.