Improving the reliability of airborne and ground-based electromagnetic surveys by 3-D inversion (Its application to landslide investigations)

为了防止山体滑坡等自然灾害的发生，重要的是要进行区域地质和岩土工程调查，确定可能造成破坏的区域。电阻率是一种地球物理性质，可以用来调查地质条件。机载电磁法最适合于在大范围内有效地采集与大地电阻率有关的数据。但存在一个问题，即目前还没有一种实用的解释方法可以在数据前生成二维或三维图像。在本研究中，我们开发了一种考虑地形的三维反演方法，利用合成数据进行精度校验，并将其应用于Kouchi县滑坡地区的实际数据集。我们证实，航空电磁数据反演的结果与地面电阻率调查的结果相当。以水平回路为发射机和接收机的地面电磁方法可分为两类；第一种类型使用发射器和接收器环路之间非常小的间隔，例如1米，而第二种类型使用更长的间隔，例如50米或100米。我们比较了这两种类型对反演的适用性。第一种方法适用于异常定性定位，但不能用于地球电阻率结构成像，因为数据特别是同相数据极易受到环境噪声的影响。对于第二种类型，我们通过合成和实际数据实例表明，反演可以成功地应用于数据中，以恢复二维和三维电阻率结构。

In order to prevent natural disaster on slopes such as landslide, it is important to locate the area that can possibly cause damages by carrying out regional geological and geotechnical surveys. The electrical resistivity is a gephysical property that can be used to investigate geological conditions. The airborne electromagnetic method is the most suited for efficiently collecting the data associated with the electrical resistivity of the earth in a large area., but it has a problem that there is no practical interpretation method that can produce 2-D or 3-D images front the data. In this study, we developed a 3-D inversion method that can take into account topography checked the accuracy using synthetic data, and applied it to a real data set collected over a landslide area in Kouchi Prefecture. We confirmed that the results from inversion of airborne electromagnetic data are comparable to those from a ground-based electrical resistivity survey.Ground-based electromagnetic methods with horizontal loops as the transmitter and receivers can be grouped into two types ; the first type uses a very small separation between the transmitter and receiver loops, say 1 m, and the second type uses much longer separation, say 50m or 100m. We compared the applicability of these two types to inversion. It is concluded for the first type that it is suitable for locating anomalies qualitatively, but cannot be used to image resistivity structures of the earth because the data, particularly in-phase data, are very susceptible to ambient noise. For the second type, we showed using synthetic and real data examples that the inversion can be successfully applied to the data to recover 2-D and 3-D resistivity structures.