A step toward the practical stratigraphic automatic correlation of well logs using continuous wavelet transform and dynamic time warping technique

Abstract

This paper introduces a high-definition method of stratigraphic well log correlation, which is more favored to be applied in the automatic correlation of well logs. The method employs a continuous wavelet transform (CWT) and a polynomial normalization technique to create correlatable spectral trend logs such as Detrend Error log (DTEL) and Integrand Detrend Error log (INDTEL). In the next step, the dynamic time warping (DTW) algorithm was used for automatic stratigraphic correlation of correlatable spectral trend logs (i.e., INDTEL and DTEL). The image of normalized power of CWT of Gamma Ray log (GR) was used to verify the correlation manually. Clusters from the classification of normalized power of CWT transform of GR log were used for parasequence identification. The results of this novel method exhibit the suitability of using this technique as an automatic reliable well log trend correlation with high accuracy.