<p>The search for the correct value of the regularization parameter is the crucial step in the concept of the Tikhonov regularization. The present sequential methods are using the specialized curves (L<sub>p</sub> norms or L curves) on which the points with given properties are searched and tested as desired values of the regularization parameter. Here, we extended one of those methods into two dimensions for the regularized downward continuation operator, while there are situations where the noise shows some directional dependence. The two regularization parameters—one for the each direction (<i>x</i> and <i>y</i>), could be the partial answer to these problems. The method is tested on the synthetic and two real data sets from detailed and regional magnetometry to demonstrate its possibilities.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Two Dimensional Regularization Parameter in the Downward Continuation of Potential Fields

  • Roland Karcol,
  • Roman Pašteka

摘要

The search for the correct value of the regularization parameter is the crucial step in the concept of the Tikhonov regularization. The present sequential methods are using the specialized curves (Lp norms or L curves) on which the points with given properties are searched and tested as desired values of the regularization parameter. Here, we extended one of those methods into two dimensions for the regularized downward continuation operator, while there are situations where the noise shows some directional dependence. The two regularization parameters—one for the each direction (x and y), could be the partial answer to these problems. The method is tested on the synthetic and two real data sets from detailed and regional magnetometry to demonstrate its possibilities.