دانشگاه صنعتی شاهرود پژوهش های ژئوفیزیک کاربردی 2476-5007 مقالات آماده انتشار 2025 11 25 Residual and regional separation of gravity data using bi-dimensional ensemble empirical mode decomposition (BEEMD) approach Residual and regional separation of gravity data using bi-dimensional ensemble empirical mode decomposition (BEEMD) approach 3668 10.22044/jrag.2025.16397.1373 FA سید ابوالحسن رضوی تهران، میدان آزادی، بلوار معراج، سازمان زمین شناسی و اکتشافات معدنی کشور 0000-0002-7351-1720 علی نجاتی کلاته دانشکده مهندسی معدن، نفت و ژئوفیزیک، دانشگاه صنعتی شاهرود امین روشندل کاهو دانشکده معدن، نفت و ژئوفیزیک، دانشگاه صنعتی شاهرود Journal Article 2025 10 03 Separating residual and regional anomalies from the Bouguer gravity anomaly is challenging in the gravity data processing. The residual anomalies are commonly used for modeling and processing gravity data. Some methods have been introduced to isolate the residual and regional components of gravity data, resulting in different residual anomalies for the same Bouguer gravity anomaly. One of these separation methods uses a bi-dimensional empirical mode decomposition (BEMD) algorithm. Its results are far from the true residual anomaly in the complex geological structure due to the major significant central problem with the BEMD known as mode-mixing. We use the bi-dimensional ensemble empirical mode decomposition (BEEMD) to improve residual and regional anomaly separation. The BEEMD reduces the mode-mixing problem in BEMD by applying the sifting process on an ensemble of white noise-contaminated signals, followed by averaging the results to arrive at the final favorite result. This method is used for synthetic and field data taken over chromite deposits in Camaguey province, Cuba. Using the presented approach on synthetic and field data, compared to conventional BEMD and polynomial fitting methods, shows the efficiency of this method in residual and regional anomaly separation. Furthermore, the estimated residual anomaly by the presented method on real data matches the drilling information. The accurate separation of residual and regional anomalies from Bouguer gravity data constitutes a fundamental challenge in geophysical data processing, with significant implications for mineral exploration and geological interpretation. Residual anomalies serve as critical inputs for modeling and interpreting gravity data in various geological contexts. Numerous computational methodologies have been developed to isolate these components, frequently producing substantially different residual anomalies from the same initial Bouguer gravity dataset, thereby complicating geological interpretation. One prominent approach employs the bi-dimensional empirical mode decomposition (BEMD) algorithm; however, its effectiveness diminishes considerably in geologically complex terrains due to the mode-mixing phenomenon, a fundamental limitation that produces substantial deviations from the true residual anomaly. To address this critical limitation, our research implements an advanced bi-dimensional ensemble empirical mode decomposition (BEEMD) framework to enhance the separation process. The BEEMD technique effectively alleviates mode-mixing deficiencies by systematically executing the sifting process on multiple realizations of signals augmented with controlled white noise, followed by sophisticated ensemble averaging to derive the final optimal decomposition while preserving essential geological features. We comprehensively validated this methodology using both carefully designed synthetic data models and extensive field data acquired from chromite deposits in Cuba's Camaguey province. Rigorous comparative analysis against conventional BEMD and polynomial fitting methods unequivocally demonstrates the superior efficacy and robustness of our proposed approach in achieving precise separation of residual and regional anomalies across diverse geological scenarios. Critically, and of paramount practical importance, the residual anomaly extracted from field data through our innovative method exhibits remarkable consistency with drilling information, thereby validating its practical utility for mineral exploration programs and highlighting its potential for broader application in geophysical interpretation workflows.