基于二维小波自适应阈值的海洋磁场资料去噪方法

构建海洋三维磁空间背景场模型时，解析延拓是实现不同高度磁场数据相互转换的关键。针对向下延拓数据中的高频信息极易被延拓算子放大，并包含在延拓后的磁场资料中，降低了三维磁空间背景场模型的构建精度，引入二维小波自适应阈值去噪方法，在不同小波分解阶层中设定自适应阈值，在向下延拓前去除磁测数据中的高频噪声，改善向下延拓结果。理论模型试验与实例分析表明，在不同噪声条件下，二维小波自适应阈值法可有效提高数据的信噪比，去噪效果优于传统的快速傅里叶（FFT）平滑法，从而有效改善频域迭代向下延拓结果，为三维磁空间背景场模型构建提供基础。

Denoising Method of Marine Magnetic Field Data: Based on the 2D Adaptive Wavelet Threshold

Marine magnetic survey data needs to be converted to different heights to construct the 3D marine magnetic space background field model. The analytical continuation of the magnetic field is a key technique for this transformation. The high frequency information in survey data is easily amplified by the continuation operator, with the result that downward continuation contains lots of noise, reducing the accuracy of the 3D marine magnetic space background field model. For this problem, the 2D adaptive wavelet threshold denoising method is introduced in this paper. The noise variance is estimated in survey data by this method, setting the adaptive threshold according to different wavelet decomposition levels, and high frequency noise in the magnetic data is removed before downward continuation. Actual data processing shows that under the conditions of different noise, the denoising effect of wavelet adaptive threshold method is superior to FFT smoothing, and the result of downward continuation has a significant improvement, providing the basis for the construction of 3D marine magnetic space background field model.