藤井/郁子

Fujii/Ikuko

Global three dimensional geomagnetic response functions


著者名 Authors 所属機関名 所属機関名(欧文)
1 #藤井/郁子 Fujii/Ikuko 東京大学地震研究所 Earthquake Research Institute, University of Tokyo
2 F2 Schultz/Adam K2 Institute of Theoretical Geohpysics, University of Cambridge

講演予稿:
世界規模の地磁気観測所ネットワークのデータを用いた地磁気変換関数 の解析方法について講演する。3種類の地磁気変換関数が比較検討され ,我々は新たにD-関数を提案する。D-関数は通常用いられるC-関数より も高次元の構造に対して感度が良く,安定性も良い。このD-関数を用い てマントルまでの電気伝導度構造を推定するためには,数日から数年の 周期のデータが必要である。実際に1961年から1975年までのデータを 用いて解析を行ったところ,10年間の連続データがえられるのは,全世 界でわずか20の観測点しかないことがわかった。

abstract:
We report an attempt to compute geomagnetic response functions from the worldwide geomagnetic observatory network. The geomagnetic response is the actual data to be fit in an inversion to obtain the conductivity distribution in the mantle. It is essential to get the accurate response function for the 3D modelling. Three response functions are examined in this study C, D, and zeta responses. The C response has been widely used because of the simplicity of the form and stability in computation when the source field has a dipole shape and the mantle has a 1D structure. We propose the D response, which is as stable as the C in the computation and much more sensitive to multi-dimensional conductivity structures than the C. The zeta response is the most sensitive to the structures, however, it can be unstable in computation if the multi-dimensional structure produces only weak small scale patterns in the geomagnetic field. The period range of interest to investigate the mantle down to about 700 km is from a few days to a few years. This requires good data sections for at least 10 years at as many places at the Earth's surface as possible. Considering uncertainties of the source fields and necessity to resolve weak high order signals by small scale structures, the number and distribution of the observatories are critical as well. We applied the empirical orthogonal functions (EOF) to the geomagnetic field data from 1961 to 1975. It turned out the data at only 20 observatories in the world are continuous and clean. We will talk about the development of the EOF to a gappy data set in order to increase available data. Also, we will discuss substitution of the spherical harmonic expansion to compute the spatial derivatives of the field. As we reported in the spring meeting, the spherical harmonic expansion has severe problems if the data distribution are spatially sparse and the spatial structure is complicated.

キーワード:
地磁気変換関数, マントルの電気伝導度,3次元モデリング

Keywords:
geomagnetic response function, mantle conductivity, 3D modeling