Dynamic Enhanced Weighted Drainage Catchment Basin Method for Extracting Geochemical Anomalies

Abstract

Geochemical measurements of stream sediments are practical for small-scale mineral exploration. However, traditional grid interpolation methods cause element concentrations to diffuse and smooth out anomalies, particularly in complex terrains, making it challenging to reflect the actual distribution of elements accurately. We applied the Dynamic Enhanced Weighted Drainage Catchment Basin (DE-WDCB) method to enhance the retention and identification of local anomalies by limiting the scope of analysis to specific drainage units. This method reduces interference from varying background values across different watersheds, effectively enhancing geochemical element anomalies and aligning better with geomorphic conditions. The DE-WDCB method was tested in the Duobaoshan–Heihe area, a significant copper polymetallic mineral district in northeastern China. Compared with traditional grid interpolation methods, the DE-WDCB method retained and strengthened low and weak abnormal information of favorable mineralization elements, particularly in the Luotuowaizi area. The method demonstrated a higher spatial coverage rate with mineral points and a more vital ore-indicating ability. Specifically, the DE-WDCB method identified anomalies with a mean accuracy of 63.57% (p < 0.05, 95% CI: 47.64%–79.50%), compared to 50.53% for traditional methods. In conclusion, in regions with a complex topography and watershed differences, the DE-WDCB method effectively reduces local geochemical background interference, accurately identifies low and weak geochemical anomalies, and better reflects the actual distribution of elements. This makes it a significantly advantageous method for geochemical anomaly extraction, delineating higher-confidence exploration targets in the Sandaowan–Luotuowaizi area in the east and the triangular area between Duobaoshan, Yubaoshan, Sankuanggou, and the midstream highlands of the Guanbird River in the west.

1. Introduction

Traditional mineral exploration methods primarily rely on surface mapping to trace mineralized outcrops. However, as mineral exploration in China has advanced, the focus has gradually shifted toward locating concealed, semi-concealed, and difficult-to-identify deposits [1,2,3]. The decreasing availability of ‘high, large, and complete’ anomaly information necessitates new methods for identifying weak mineralization anomalies, which are critical for future exploration efforts [4,5,6].

With the continuous development of geological big data, the re-analysis of exploration geochemical data has become increasingly important. The extraction of geochemical anomalies from stream sediments, a vital geochemical exploration method, plays a crucial role in understanding the mineralization patterns of concealed deposits, delineating mineralization-induced anomalies, optimizing exploration targets, and evaluating resource potential [7,8,9,10]. In traditional geological work, grid-based continuous interpolation methods do not account for local variations in element background values caused by topographical differences within the area. This often results in false high-intensity anomaly zones and the smoothing or masking of weak anomaly information. Consequently, there is a discrepancy with the actual anomaly distribution, leading to the loss of significant weak anomaly information indicative of mineralization [11,12,13,14].

To address the above issues, researchers have proposed the “catchment-controlled area” method [15], which treats catchment basins as control units for stream sediment sampling points. As early as 1987, Bonham-Carter et al. [16] proposed Sample Catchment Basin Analysis (SCB) for predicting lead–zinc deposits in the Cobequid Highlands of Nova Scotia, Canada. This method demonstrated its effectiveness in improving the tracing and extraction capabilities of favorable element anomalies in mountainous and hilly regions. In 2006, Spadoni M. et al. [17] proposed the Extended Sample Catchment Basins (ESCB) method, which refines the catchment basin model by calculating sediment source concentration cut points through interpolation curves. They applied this method to extract anomalies in the southern part of the Mignone Basin in Italy. In 2013, Yousefi M. [18] proposed the Weighted Drainage Catchment Basin (WDCB) method, assigning fuzzy weights to each basin and predicting the mineral prospect of porphyry copper deposits in the Kerman Province of southeastern Iran. In 2019, Farahbakhsh E. [19] proposed the Catchment Basin Score (CBS) method, which describes the suitability of sampling points within catchment basins and improves the anomaly extraction capability of the WDCB method by adjusting watershed weights. This method successfully predicted porphyry copper–gold deposits in the Macquarie River Basin in eastern New South Wales, Australia. Over nearly four decades, methods for identifying and extracting geochemical anomalies based on catchment basins have developed into a systematic workflow that effectively avoids mathematical interference between samples and extracts local anomalies, particularly in mountainous and hilly regions [17,20,21].

The Duobaoshan–Heihe ore cluster lies in Heilongjiang Province, in northeastern China’s Greater Khingan Range. It is one of China’s three significant porphyry copper polymetallic mining areas. Situated at the Xingmeng orogenic belt intersection and the active continental margin of the western Pacific [22], the Duobaoshan area has a rich tectonic evolution history, providing favorable conditions for mineralization. The region hosts a variety of discovered deposits with complex mineralization characteristics, including super-large deposits such as the Duobaoshan porphyry copper deposit, the Tongshan porphyry copper deposit, and the Zhengguang gold deposit. Consequently, the Duobaoshan area has long been a focal point for geological exploration and mineralization theory research in Northeast China [23]. Previous geochemical surveys conducted at scales of 1:50,000 and 1:200,000 in the Duobaoshan ore cluster area have employed traditional grid-based interpolation methods to create contour maps of element anomalies [24,25]. However, these methods are unsuitable for expressing stream sediment measurements’ results, resulting in less reliable geological body reflections and highlighting the need for more appropriate mapping methods.

This study aims to evaluate the effectiveness of the Dynamic Enhanced Weighted Drainage Catchment Basin (DE-WDCB) method in accurately identifying geochemical anomalies in complex terrains, specifically within the Duobaoshan–Heihe ore cluster area. We hypothesize that the DE-WDCB method will outperform traditional grid interpolation methods by reducing background noise and enhancing anomaly detection. This study employs inverse distance weighting (IDW), Kriging, and the DE-WDCB method to extract and analyze geochemical anomalies from 1:200,000 scale stream sediment geochemical data in the Duobaoshan area. The findings are expected to provide new insights into geochemical mineralization environments and offer a scientific basis for future regional copper exploration.

2. Methodology

2.1. Multivariate Statistical Analysis

Multivariate statistical analysis is an effective method for analyzing geochemical data. Standard techniques include correlation analysis, R-type cluster analysis, and principal component analysis (PCA). In this study, we employed these methods to extract element combination characteristics, reduce the dimensionality of the measurement data, and explore the internal structure among multiple variables [26]. We used ArcGIS versiom 10.8 for spatial analysis and SPSS version 28 for statistical computations, including multivariate statistical analysis and trend surface fitting.

This approach not only investigates the co-association patterns of various elements but also characterizes the spatial distribution of primary ore-forming element combinations based on principal component scores. It can indicate the mineralization environment, discern the sources of ore-forming materials, simulate and reconstruct the element migration and evolution processes, and reveal geochemical patterns [27]. These insights provide a scientific basis for mineralization prediction and resource evaluation.

2.2. Dynamic Enhanced Weighted Drainage Catchment Basin Method

The Dynamic Enhanced Weighted Drainage Catchment Basin (DE-WDCB) method, an optimized algorithm based on the Weighted Drainage Catchment Basin (WDCB) method, was then applied [28]. Both methods focus on discretely modeling the geochemistry of the catchment basin where the sampling points are located and assigning these characteristics to their respective basins. Unlike the WDCB method, the DE-WDCB method considers variations in the geochemical background field at different locations. It revises the WDCB method’s approach by employing trend surface analysis theory to fit different geochemical background values for each basin unit and grades anomalies for the basins of different sampling points to facilitate anomaly extraction.

Due to the influence of various activities, such as the transportation, migration, and mixing of surface water and groundwater in upstream basins, the concentration of indicative elements tends to decrease with increasing distance from the anomaly source downstream. Thus, strong anomalies often exhibit sliding displacement along the stream distribution in practical geochemical anomaly extraction. However, the complexity and uncertainty of downstream influences mean that the relationship between indicative element concentrations and distance downstream is nonlinear. Therefore, the DE-WDCB method, building on the theoretical foundation of the WDCB method, uses the “suitability degree allocation” weighting method to fit this nonlinear relationship. It introduces “accumulated flow volume” as a parameter to represent the suitability of each sampling point’s basin. The larger the accumulated flow volume, the more likely the location is downstream and prone to depositing elements transported from upstream. Therefore, it should be assigned a smaller weight in anomaly calculations to reduce false anomalies. Conversely, elements in relatively upstream locations are more likely to be carried downstream by water flow and should be assigned a more significant weight to restore the upstream anomaly quantity. Ultimately, this enhances the ability to identify and extract indicative element anomaly sources and accurately reconstruct the anomaly levels in the catchment basin.

The DE-WDCB method for extracting geochemical anomalies mainly involves the following two parts:

Using trend surface analysis, the DE-WDCB method determines the basin background value and anomaly threshold. The following equation expresses the mathematical formula:

Z_i = T_i + M_i

(1)

where Z_i is the observed value, T_i is the regional trend value, and M_i is the residual component, consisting of local variations and random values. During geochemical anomaly extraction, areas where the residual component exceeds the trend (Z_i-T_i > 0) are significant as they often indicate regions of element enrichment.

The DE-WDCB method assigns anomaly scores to each drainage basin. The mathematical formula is given by:

$$C_x={\sum }_{i=1}^n{\displaystyle \frac{a^{{FA}_i}}{{{\sum }_{i=1}^{n}a}^{{FA}_i}}}·\left(n-i+1\right)\left(M_i{\displaystyle \frac{N_i}{N_t}}\right)·100$$

(2)

where C_x is the DE-WDCB score, n is the total number of anomaly grades for sampling point elements, FA_i is the average value of each type of flow accumulation, M_i is the median concentration of elements classified as the highest anomaly grade (i-grade), N_i is the number of sediment samples in the i-grade, and N_t is the total number of sediment samples in each drainage basin unit. a is an introduced constant with a value between 0 and 1.

The DE-WDCB method can employ various grading rules for anomaly classification. In this study, the cumulative frequency method was used for grading to compare with the anomaly extraction results of the control group. Thresholds for low, medium, and high anomalies were set at the 85th, 95th, and 98th percentiles, respectively, i.e., n = 3.

3. Workflow

Figure 1 illustrates the detailed workflow of this study, outlining the steps from data collection to anomaly extraction and analysis. The workflow is structured as follows:

• Data collection and preprocessing using ArcGIS.
• Statistical analysis and extraction of favorable metallogenic elements using SPSS.
• Establishment of the Digital Watershed Model.
• Application of the DE-WDCB method for anomaly extraction.
• Interpretation of results with a focus on spatial distribution and geochemical patterns.

By integrating these advanced statistical techniques and geospatial analysis tools, we aimed to improve the accuracy and reliability of geochemical anomaly detection in complex terrains.

4. Application Examples

4.1. Geological Setting

4.1.1. Regional Geological Setting

The Duobaoshan–Heihe ore cluster area is located in the Greater Khingan Range arc-basin system, on the west side of the significant fault separating the Xing’an block and the Songnen block (Figure 2a), and belongs to the eastern end of the Central Asian–Xingmeng geosyncline fold belt [29,30]. The region has undergone several geological events, including the formation of the Duobaoshan island arc in the Early Paleozoic, the development of the Handaqi intra-arc rift in the Late Paleozoic, and intense Mesozoic basin-and-range tectonic and magmatic activities (Figure 2b) [31]. These long-term tectonic and magmatic activities have resulted in the formation of the Duobaoshan–Kuanhe rhombohedral structure, controlled by NE- and NW-trending significant faults [32], which significantly influence the distribution of strata, magmatic rocks, and mineral deposits within the area. The sedimentary strata in the area primarily include the Ordovician, Silurian, and small amounts of the Upper Proterozoic, Cambrian, Devonian, Carboniferous, Permian, Jurassic, Cretaceous, and Cenozoic.

The ore cluster area mainly hosts several super-large to large Early Paleozoic epithermal low-temperature hydrothermal copper–molybdenum and copper–gold deposits, such as the Duobaoshan copper deposit, the Tongshan copper deposit, and the Zhengguang gold deposit [33]. The periphery of the ore cluster area predominantly develops Mesozoic magmatic–hydrothermal copper–gold, copper–molybdenum, and copper–lead–zinc deposits (or occurrences), such as the Sankuangou copper–molybdenum deposit and the Xiaoduobaoshan copper deposit. The Duobaoshan-type copper polymetallic deposits primarily formed due to the combined effects of early Caledonian tectonic and magmatic activities. During the Yanshanian period, hydrothermal activities were most intense, significantly impacting mineral enrichment and the destruction of ore deposits in the area [34].

Figure 2. (a) Schematic diagram of the Eurasian continent (adapted from [35]). (b) Tectonic location of the Duobaoshan–Heihe metallogenic belt.

Figure 2. (a) Schematic diagram of the Eurasian continent (adapted from [35]). (b) Tectonic location of the Duobaoshan–Heihe metallogenic belt.

4.1.2. Overview of Typical Duobaoshan-Type Copper Deposits in the Study Area

The strata in the study area are well developed, ranging from the Ordovician to the Quaternary. The Upper Silurian Wotu Formation (S₃w) unconformably overlies the Lower Devonian Heitai Formation (D₂h), whereas the other strata maintain conformable contacts (Figure 3). Typical super-large copper deposits in the area, such as Duobaoshan and Tongshan, mainly occur in the Lower Ordovician Tongshan Formation (O₁t) and the Middle Ordovician Duobaoshan Formation (O₂d), consisting of andesite, andesitic tuff, tuffaceous sandstone, quartz sandstone, siltstone, and carbonaceous slate [36].

The area has experienced frequent magmatic activities, with multiple magmatic intrusions during the Jinning, Caledonian, Variscan, Late Indosinian, and Yanshanian periods. Granite intrusions spread widely across the region. The Early Caledonian tectonic layer, comprising Neoarchean to Early Cambrian volcanic and carbonate clastic rocks, forms the foundation of the Duobaoshan mineralization area [37]. The granitic porphyry and granodiorite of the Duobaoshan volcanic arc are the central ore-bearing rock bodies in the area.

The structural features in the area are well developed, with NE- and NW-trending faults jointly controlling the Yanshanian mineralization belts. The NW-trending Duobaoshan–Taxi fault and the Huasuizhai–Kesu Gulan River fault, formed in the Late Paleozoic, control the distribution of strata in the area [31]. Copper mineralization is primarily controlled by faults formed by submarine volcanic rocks of the Duobaoshan Formation and associated NE- and NW-trending structures.

The alteration associated with the porphyry copper deposits and granitic porphyry in the area exhibits distinct zoning characteristics, forming a planar, progressive porphyry–copper alteration zone. This zone extends outward from the central potassic silicification, followed by zones of potassic feldspar porphyry, quartz–sericite, and chlorination, with localized occurrences of monetization, skarn, and hornfels alterations.

4.2. Basic Data Characteristics and Multivariate Statistical Analysis

This study utilizes 1:200,000 scale stream sediment measurement data from the China National Geochemical Mapping Project to statistically analyze 14 elements (Au, As, Ag, Cd, Co, Cr, Cu, Nb, Ni, Pb, Sb, Ti, W, Zn) in the study area.

4.2.1. Basic Characteristics of Single-Element Distribution

After undergoing a single or multiple natural geochemical processes, the element content in the geological body tends to be normally distributed [38]. The frequency histogram of element concentration (Figure 4) illustrates the distribution pattern of each element. Through the iteration method, we confirmed that the data of all 14 elements passed the regular distribution test. Then, we calculated these elements’ mean, standard deviation, anomaly classification threshold, and other related parameters to understand the spatial distribution and enrichment or depletion.

The statistical analysis in

Table 1. Statistics of geochemical parameters of stream sediment elements in the Duobaoshan area.

Element	Iterative Data Exclusion	Mean	Standard Deviation	Background Value of Greater Khingan Range	Range Coefficient (%)	Kurtosis Coefficient	Skewness Coefficient	Concentration Coefficient	Variation Coefficient	Abnormal Outer Band	Abnormal Middle Band	Abnormal Inner Band
Au	4	1.4	1.9	1.0	126.5	0.7	1.2	1.4	1.4	1.7	2.2	2.2
As	4	10.0	7.5	7.7	104.9	0.8	0.9	1.3	0.7	11.8	13.4	14.6
Ag	14	135.9	210.5	73.7	1093.5	22.0	4.8	1.2	1.5	97.7	104.1	105.7
Cd	8	75.1	72.0	80.5	59.4	6.1	−1.0	0.9	1.0	79.8	82.2	84.8
Co	3	14.3	5.3	9.0	34.9	0.1	1.2	1.6	0.4	15.2	15.5	15.8
Cr	0	55.0	16.0	27.5	40.2	−0.6	0.0	2.0	0.3	59.8	61.9	63.0
Cu	12	22.2	131.7	16.4	78.4	0.0	8.4	1.4	5.9	25.7	29.9	31.0
Nb	0	13.7	2.2	12.5	40.7	−0.7	0.0	1.1	0.2	15.1	15.8	16.0
Ni	2	24.2	6.6	11.5	77.8	−0.4	0.3	2.1	0.3	28.2	30.1	30.6
Pb	5	22.1	17.3	22.3	19.2	−1.9	0.1	1.0	0.8	23.1	23.5	23.7
Sb	0	0.4	0.1	0.5	97.3	−0.1	0.5	0.9	0.3	0.5	0.6	0.6
Ti	4	4854.3	3558.2	2991.3	24.8	1.6	0.1	1.6	0.7	4986.0	5098.1	5161.3
W	2	2.0	3.9	1.5	52.7	3.1	0.4	1.3	1.9	2.1	2.2	2.2
Zn	7	73.5	83.2	58.3	28.6	0.7	1.7	1.3	1.1	76.8	79.5	81.7

Units: This sentence is not intended to explain the asterisk but rather to clarify the units for the element contents in the table. The asterisk can be removed. for Au, Ag, and Cd are 10⁻⁹ (ppb); units for other elements are 10⁻⁶ (ppm); background values for elements in the Greater Khingan Range [39].

reveals that, except Sb and Cd, the average concentrations of other elements in this region exceed the background levels of the Greater Khingan metallogenic belt, indicating significant enrichment of Au, As, Ag, Co, Cr, Cu, Nb, Ni, Pb, Ti, W, and Zn. Notably, Cu, W, Ag, and Au exhibit high variation coefficients (>1.50) and concentration coefficients (>1.3), suggesting local solid differentiation and marking them as key ore-forming elements. Mineralization points for Cu, Ag, and Au have been confirmed in the study area, reinforcing its potential for further exploration.

4.2.2. Characteristics of Element Combinations

This study employs correlation analysis, cluster analysis, and principal component analysis (PCA) to investigate the correlations between elements and identify indicative element combinations for copper polymetallic deposits in the study area.

1.

Cu as the Primary Ore-Forming Element and its Indicative Correlations

To address compositional data’s “closure effect” [40], we standardized the data using a log transformation before calculating Pearson correlation coefficients. As shown in Figure 5, Cu exhibits a significant positive correlation with Co, Cr, and Ni and a significant negative correlation with Ti, all with absolute coefficients greater than 0.5. These findings suggest that Cu is the primary ore-forming element in the study area, while Co, Cr, Ni, and Ti are likely indicative elements for mineral exploration.

2.

Element Grouping Revealing Copper Polymetallic Mineralization Patterns

We performed an R-type clustering analysis on 14 elements and divided them into four groups at a Euclidean distance of 18 (Figure 6). The first and third groups present the combinations closely related to copper polymetallic mineralization.

• Group 1: Cu, Cr, Ni, Co, Ti

This group is closely related to copper polymetallic mineralization. As typical chalcophile elements, Cu, Ni, and Co are strongly associated with intermediate to acidic magmatic–hydrothermal activities, often enriching and precipitating in high-temperature hydrothermal environments. Cr, primarily a lithophile element, is typically found in the outer alteration zones formed during the quartz–sericite stage of mineralization. Ti, mainly present in ilmenite (FeTiO₃) and rutile (TiO₂), is active in high-temperature, high-pressure, and acidic hydrothermal fluids but remains stable and precipitates in lower-temperature environments [41]. It contrasts with the accumulation and deposition environments of Cu, Co, and Cr. This group’s distribution makes it a critical indicator for locating and assessing copper polymetallic deposits.

• Group 3: Cd, Zn, Pb

These elements, also chalcophile in nature, typically coexist with copper polymetallic minerals and are associated with medium to high-temperature hydrothermal mineralization. They often deposit in lower-temperature environments, typically on the peripheries of copper ore bodies, exhibiting distinct primary halo zoning. This distribution pattern indicates that these elements have undergone hydrothermal mineralization under varying temperature conditions, suggesting the area’s multi-staged or complex geological processes.

• Group 4: As, Sb, Au, Ag

This group indicates low-temperature hydrothermal mineralization, notably different from the high-temperature magmatic activities indicated by the first group. We can view these variations in temperature and pressure conditions as markers of multiple geological processes in the Duobaoshan–Heihe area.

3.

High-Temperature Element Combination Reflecting Copper Mineralization Potential

According to the significance test (

Table 2. KMO and Bartlett’s test of sphericity.

KMO		0.733
Bartlett’s test of sphericity	Approximate chi-square	3579.201
	Degrees of freedom (df)	91
	Significance	0.000

), the KMO value is 0.733, the degree of freedom, df, is much larger than 0, and the significance is less than 0.01, which aligns with the requirement of using principal component analysis.

The principal component analysis extracted four component factors with eigenvalues greater than 1 (

Table 3. The load matrix of the original component.

Element	Component Factors
	Fac 1	Fac 2	Fac 3	Fac 4
Au	−0.320	0.674	0.381	0.153
As	−0.102	0.884	−0.053	−0.078
Ag	−0.334	0.191	−0.269	−0.540
Cd	0.324	−0.110	0.695	0.424
Co	0.785	0.451	0.193	−0.236
Cr	0.879	−0.010	0.120	−0.218
Cu	0.771	0.378	−0.047	0.188
Nb	0.404	−0.286	−0.439	0.561
Ni	0.888	−0.172	0.111	−0.077
Pb	−0.245	−0.696	0.313	−0.099
Sb	−0.181	0.938	0.101	0.103
Ti	0.804	−0.027	0.097	−0.329
W	0.154	0.192	−0.338	0.698
Zn	−0.339	−0.081	0.801	0.152

).

• Fac 1: This factor is dominated by high-temperature elements, including Co, Cr, Cu, Ni, and Ti. It primarily reflects copper mineralization information, indicating the substantial potential for copper exploration in the area.
• Fac 3: This factor is composed of Cd and Zn, elements typically enriched in medium to low-temperature metallogenic zones. It reflects the primary halo zoning phenomena exhibited by wall rock alteration during various stages of copper mineralization.

4.3. Extraction of Stream Sediment Geochemical Anomalies Using DE-WDCB

This section describes the extraction of watershed characteristics in the study area using Aster 30m resolution DEM data to establish the geochemical landscape of catchment basins (Figure 7 and Figure 8). Using the DE-WDCB method, we determined dynamic background values and anomaly thresholds for different sampling points, effectively addressing geochemical background differences between basin landscapes.

4.3.1. Construction of the Catchment Basin Landscape Model

Before constructing the catchment basin model, the original data were preprocessed with sink-filling analysis to restore the correct outlet positions of the basins, ensuring the continuity of the generated water systems and the accuracy of the model (Figure 8a). Next, using the D8 algorithm, the flow direction was set to eight possible directions, represented by 2ⁿ (where n ranges from 0 to 7). ArcGIS calculates the slope relationships between each grid cell and its eight neighboring cells to determine the steepest downhill neighbor as the flow direction (Figure 8b). We refer to the flow accumulation of all downhill cells flowing into a grid cell as the flow accumulation of that cell (Figure 8c). Higher flow accumulation indicates more concentrated flow and a greater likelihood of runoff formation, reflecting the transport trend of weathering products like rocks across the surface.

We used the Strahler classification method to classify the water systems in the area into four levels (Figure 8d). Rivers originating from the headwaters are classified as first-order rivers, which, as shown in the figure, have a broad influence area, indicating minimal pollution from natural and human factors in the survey area. Finally, we set the outlets of the various water system levels as the spill points for the catchment basins. By capturing these spill points, a catchment basin model matching the water system was generated (Figure 8e,f).

4.3.2. Extraction and Mapping of Geochemical Anomalies for Favorable Ore-Forming Elements and Element Combinations

We used dynamic trend surface analysis to calculate trend values for each catchment basin based on the spatial distribution of high indicator element contents and the catchment basin divisions in the study area. This approach allowed us to obtain the distribution of geochemical background fields in the study area.

The results of the background field extraction (Figure 9) show significant differences in the geochemical background fields of Co, Cr, Cu, Ni, Ti, and their combination elements across different basin landscapes in the study area. The study area’s northwest to west side shows a high trend value concentration for favorable ore-forming single elements. In contrast, low trend values are concentrated in the central to eastern side, with the overall distribution trend weakening from northwest to southeast. The western and northeastern edges of the study area exhibit high trend values for element combinations, with small high-value areas showing a weak distribution trend from southwest to northeast.

The DE-WDCB algorithm’s inversion results reveal that single-element and element combination anomalies predominantly occur in the Duobaoshan–Sankuangou and Luotuowaizi–Sandaowan areas (Figure 10). The anomalies form closely connected block-point patterns in these regions, with primary ore-forming Cu and element combinations showing similar distribution characteristics.

In the Duobaoshan–Sankuangou area on the study area’s west side, high-value Cu and element combination anomalies extend in an NWW direction, forming an elliptical distribution over a large area. The anomaly area exposes a significant amount of favorable ore-forming strata, such as the Duobaoshan Formation (O₂d) and the Tongshan Formation (O_1-2t), and includes multiple complex NNW-trending faults. These faults generally align with the extension direction of the Sankuangou–Duobaoshan–Luohe ore-controlling fold belt (Figure 2). The anomaly area covers two super-large copper deposits, Duobaoshan and Tongshan, as well as 11 medium to small copper–molybdenum and copper–lead–zinc deposits and occurrences such as Xiaoduobaoshan and Sankuangou.

High anomalies cluster in the Sandaowan area on the northeastern side of the study area, forming three dispersed high-concentration centers of Cu elements. The overall distribution area is large, with numerous NNE-trending secondary faults of the Nenjiang-Xinkailing lithospheric fault zone extending within the anomaly area.

4.4. Comprehensive Comparison and Evaluation of Anomaly Extraction Results

The study area has a topography characterized by high central and low eastern and western regions, with the northern ridge of the Xiaoxing’an Mountains cutting through the center, dividing the study area into eastern and western regions (Figure 11). Data and geological analyses were conducted separately for these regions (Figure 12).

This study used inverse distance weighting (IDW) and Kriging methods in the control group to process element content data from geochemical sampling points.

Table 4. Optimal model parameters for control group using IDW and Kriging methods.

Method	Optimal Parameters	Qualitative Parameters
IDW	Optimal Distance Index (OOD)	1 ≤ α ≤ 2
	Search Step Number	12
Kriging	Kernel Function Model	Gaussian Model
	Search Direction	Four Directions
	Search Step Number	12

displays the optimal interpolation model parameters.

The cumulative frequency method was used for grading in the IDW, Kriging, and DE-WDCB methods, with the 85th, 95th, and 98th percentiles as thresholds for low, medium, and high anomalies, respectively, to extract single elements and element combinations related to copper mineralization (

Table 5. Comparison of anomaly areas and mineralization indicators delineated by IDW, Kriging, and DE-WDCB methods.

Methods	Elements	Proportion of Anomalies in Total Area (%)				Number of Mineral Points within Anomalies			Total Mineral Point Coverage (%)
		High Anomaly	Medium Anomaly	Low Anomaly	Total Anomaly	A Region	B Region	Total Region
IDW	Co	0.92	3.03	9.14	13.09	15	0	15	42.8
	Cr	0.95	2.60	14.92	18.48	15	0	15	42.8
	Cu	0.95	2.87	8.68	12.49	16	0	16	45.7
	Ni	0.32	2.65	10.30	13.28	17	0	17	48.5
	Ti	0.72	3.67	7.71	12.10	0	2	2	5.7
	Fac	1.00	2.96	8.02	11.99	17	0	17	48.5
Kriging	Co	1.10	3.87	8.88	13.84	15	0	15	42.8
	Cr	1.45	2.46	9.60	13.51	21	0	21	60.0
	Cu	2.15	3.49	8.59	14.23	17	1	18	51.4
	Ni	1.31	3.70	9.75	14.76	18	0	18	51.4
	Ti	1.64	3.65	8.43	13.73	1	0	1	2.8
	Fac	0.08	3.13	10.48	13.69	20	0	20	57.1
DE-WDCB	Co	0.81	2.92	26.48	30.21	17	2	19	54.3
	Cr	1.09	2.24	27.47	30.81	17	2	19	54.3
	Cu	1.05	2.16	23.29	26.50	18	2	20	57.1
	Ni	0.53	2.21	23.09	25.84	18	2	20	57.1
	Ti	1.16	2.68	22.88	26.72	0	1	1	2.9
	Fac	1.36	2.13	23.99	27.48	19	6	25	71.4

).

4.4.1. Analysis of Traditional Anomaly Extraction Results Based on Grid Interpolation—IDW and Kriging

1.

Favorable Ore-Forming Single Elements

The control group, employing IDW and Kriging methods (

Table 6. Exception level 3 threshold table for IDW and Kriging methods (Unit: ppm).

Thresholds for Abnormal Grading (ppm)	Low Value	Medium Value	High Value
Co	15.2	15.5	15.7
Cr	59.8	61.8	63.0
Cu	25.7	29.8	31.0
Ni	28.1	30.1	30.5
Ti	4985.9	5098.0	5161.2
Fac	1.1	1.6	1.9

), identified Cu, Co, Cr, and Ni anomalies primarily concentrated in the Duobaoshan–Sankuangou area (Region A), with a spatial distribution characterized by higher values in the west and lower values in the east. These anomalies encompassed 14.21% of the study area and overlapped with 48.21% (p < 0.05, 95% CI: 31.61%–64.81%) of known copper mineralization points, with an average Cu anomaly coverage of 48.57% (p < 0.05, 95% CI: 31.98%–65.16%). Significant deposits, such as Duobaoshan, Xiaoduobaoshan, and Tongshan, were mainly associated with medium to low anomaly levels.

Ti anomalies, which negatively correlate with copper mineralization, were predominantly located in Region A’s northern and southern parts. These anomalies corresponded closely with the shale, gneiss, and grey mudstone distribution, exhibiting a distinct negative correlation with Cu, Co, Cr, and Ni anomalies.

Kriging outperformed IDW slightly, providing smoother anomaly boundaries and a closer alignment with NNW-trending faults. The anomaly area identified by Kriging was 0.98 times that of IDW, while its mineral point coverage was 1.14 times higher.

2.

Favorable Ore-Forming Element Combinations

The element combination anomalies of Co-Cr-Cu-Ni-Ti extracted by both methods in the control group show a similar spatial distribution to the Cu anomalies. Anomalies are only identified in the Duobaoshan–Sankuangou area (Region A), mainly exhibiting an elliptical distribution along the NW direction, controlled by the Duobaoshan–Luohe structural belt. The average anomaly area covers 12.83% of the total area, with an average mineral point coverage rate of 52.85% (p < 0.05, 95% CI: 36.23%–69.47%).

4.4.2. Analysis of Anomaly Extraction Results Based on the Discrete Interpolation of the Catchment Basin Model—DE-WDCB

1.

Favorable Ore-Forming Single Elements

Compared to the control group results, the DE-WDCB method extracts high to medium anomalies of Cu, Co, Ni, and Ti over a similar proportion of the total study area while significantly increasing the area of low anomalies. The average three-level anomalies cover 28.34% of the study area, 1.99 times higher than the control group, with an average mineral point coverage rate of 55.71% (p < 0.05, 95% CI: 39.25%–72.17%), which is 1.16 times higher than the control group. Specifically, the coverage rate of Cu anomalies over mineral points is 57.14% (p < 0.05, 95% CI: 50.75%–73.53%), an increase of 1.18 times, with typical copper deposits like Duobaoshan, Tongshan, and Xiaoduobaoshan located in high anomaly areas.

The DE-WDCB-extracted Cu, Co, Cr, and Ni anomalies show a general trend of low values in the central area and high values in the eastern and western regions, consistent with the geological body distribution. In the Duobaoshan–Sankuangou area (Region A), an NWW-trending high anomaly zone aligns with the Duobaoshan–Luohe structural belt, indicating strong structural control over copper mineralization. In the Sandaowan–Songshugangzi area (Region B), NEE-trending anomalies around the Luotuowaizi copper–molybdenum deposit and low anomalies along NEE-trending secondary faults suggest robust structural control on the eastern side.

Analyzing Co, Cr, and Ni anomalies reveals a shift in high anomaly concentration in Region A along an NE-SW axis, reflecting the upstream–downstream water system distribution. In the northern part of Region A (Independence Mountain), where the control group identified no anomalies, the DE-WDCB method shows low anomalies within multiple first-order basins. Ti anomalies, negatively correlated with copper mineralization, mainly appear as widespread low anomalies that do not overlap with copper-related mineral points, highlighting their distinct negative correlation.

2.

Favorable Ore-Forming Element Combinations

Compared to the traditional methods, the DE-WDCB method extracts high and medium anomalies of favorable ore-forming element combinations over a similar proportion of the study area but significantly increases the area of low anomalies. The total anomalies cover 27.48% of the study area, with a mineral point coverage rate of 71.42% (p < 0.05, 95% CI: 56.45%–86.39%), 2.29 times the anomaly area and 1.4 times the mineral point coverage rate of the IDW method. The overall distribution of element combination anomalies shows higher values in the east and west and lower values in the center, similar to the distribution of single-element anomalies. Unlike the control group, the DE-WDCB method identifies anomalies in Regions A and B, with significant changes in anomaly locations.

In Region A, high and medium anomaly information is mainly concentrated in the Duobaoshan–Sankuangou area. Compared to the control group, the DE-WDCB extracted medium anomalies cover large copper deposits such as Duobaoshan, Tongshan, Xiaoduobaoshan, and Nangou copper–molybdenum deposits more extensively. It also covers four copper–lead–zinc mineral points on the southwestern side of Duobaoshan, indicating more substantial mineralization potential. The low anomaly concentration centers around the Zhan copper point northeast of Xiaoduobaoshan have shifted along an NE-SW axis, and researchers identified many low anomalies in the first-order basin areas around Independence Mountain.

Region B has fewer independently distributed basins with high and medium anomalies, while basins with low anomalies are more numerous and distinctly zoned. In the Sandaowan–Luotuowaizi–Songshugangzi area, significant low anomalies extend in an NE direction, consistent with the extension direction of secondary faults of the Xinkailing lithospheric fault zone, indicating robust structural control features.

4.4.3. Comparative Analysis and Advantages of the DE-WDCB Method over Traditional Approaches

Traditional methods rely on uniform background reference values, which can effectively extract distinct third-level anomalies in regions with prominent high anomaly clusters, such as Area A. However, the pronounced topographic differences between the eastern and western ends of the Duobaoshan–Heihe study area present significant challenges. Using excessively high background reference values in these methods often leads to the diffusion and smoothing of relatively weak anomaly signals in Area B. This can result in misrepresentation, where areas in Area B that should display anomalies instead appear as “non-anomalous”, thereby hindering an accurate depiction of the actual elemental distribution.

In contrast, the DE-WDCB method addresses these challenges by accounting for geochemical background variations across different basin units. By integrating trend surface analysis with the “suitability allocation” weighted scoring method, the DE-WDCB approach can accurately align background values to basin units according to their specific geomorphic characteristics.

The comparative analysis demonstrates that the DE-WDCB method significantly outperforms traditional approaches. The average anomaly mineral point coverage using the DE-WDCB method was 63.57% (p < 0.05, 95% CI: 47.64%–79.50%), compared to 50.53% (p < 0.05, 95% CI: 33.97%–67.09%) for traditional methods. This improvement underscores the DE-WDCB method’s enhanced ability to preserve and amplify low and weak anomalies, particularly in Area B. Consequently, this approach achieves a more significant spatial overlap between the extracted anomalies and known mineral points, highlighting its superiority in accurately reflecting elemental distribution, especially in regions with complex topography and variable geochemical backgrounds.

5. Discussion

5.1. Interpretation of Results

Our experiments revealed that the traditional IDW and Kriging methods yield results comparable to the DE-WDCB method in identifying anomalies in the western part of the study area (Region A). All three methods successfully detected large-scale high-to-moderate anomalies of single elements and element combinations in the Duobaoshan–Sankuangou region (Figure 13). These anomalies trend in an NNW direction, consistent with the extension of the Duobaoshan–Luohe fault zone, and align well with the geological formations such as the Upper Carboniferous granite porphyry, Lower Ordovician Tongshan Formation (O₁t) andesite, and Middle Ordovician Duobaoshan Formation (O₂d) granite porphyry. This finding validated the effectiveness of these methods in environments with high local element concentrations.

However, in the eastern part of the study area (Region B)(Figure 14), traditional methods like IDW and Kriging identified only minor, patchy low to weak Cu anomalies, which did not spatially coincide with known mineral deposits. This discrepancy indicates the susceptibility of traditional grid interpolation methods to local background value differences in areas with lower element concentrations, thereby limiting their effectiveness in uniform anomaly extraction.

Conversely, the DE-WDCB method demonstrated superior performance in Region B by effectively reducing the interference of complex topography through trend surface analysis. This method minimized the smoothing effect of large non-anomalous areas, successfully identifying low to weak anomalies in the Sandaowan–Luotuowaizi-Songshugangzi region. These NE-trending anomalies, aligned with secondary faults in the Xinkailing lithosphere fault zone, exhibited strong spatial congruence with known copper–molybdenum deposits, underscoring the DE-WDCB method’s enhanced sensitivity and accuracy in detecting low to weak anomalies.

5.2. Comparison with Previous Studies

The geological understanding of the Duobaoshan area obtained from this study is consistent with the results of Liu B.S. et al. (2020) [30]. At the same time, the results of this study also confirmed the research of Zhu R.W. (2022) [28] on the ability of the DE-WDCB method to extract weak anomalies. They also emphasized the advantages of basin-based anomaly detection methods. The DE-WDCB method can retain and enhance low to weak anomalies, especially in areas with complex terrain, which confirms its superiority over traditional grid-based methods. Unlike traditional methods that tend to smooth and distort the distribution of anomalies, the DE-WDCB method can describe more accurately and in detail element anomalies, especially in challenging geological environments.

5.3. Study Limitations

Despite the promising results, the DE-WDCB method has certain limitations. Its effectiveness depends on the division of watershed units and the fitting of background values for basin units. While advantageous in regions with complex topography and significant watershed differences, the method’s benefits may diminish in areas with flat terrain or minimal watershed variation. In such cases, the results might not differ significantly from those obtained through traditional methods. Additionally, the success of the DE-WDCB method relies heavily on the density and uniformity of sampling points. Sparse or unevenly distributed sampling can lead to excessive smoothing or deviations from the actual geological background, potentially obscuring critical low and weak anomaly information.

5.4. Future Research Directions

Future research should focus on refining the DE-WDCB method to address its limitations. Specifically, improving the method’s applicability in flat terrain and areas with minimal watershed variation is crucial. Enhancing the trend surface analysis component to perform reliably even with sparse or unevenly distributed sampling points could further increase the method’s accuracy. Integrating more comprehensive geological data and exploring advanced statistical techniques could contribute to more reliable anomaly detection. Expanding the application of the DE-WDCB method to various geological settings and comparing its performance with emerging anomaly detection techniques will further validate and refine its use in mineral exploration.

6. Conclusions

• The DE-WDCB method significantly enhances the detection of geochemical anomalies in the complex terrain of the Duobaoshan–Heihe area. Our study identifies two distinct geochemical element groups indicative of different mineralization environments: one associated with medium-acid, high-temperature hydrothermal activity (Cu, Cr, Ni, Co, Ti) and the other with low-temperature hydrothermal processes (As, Sb, Au, Ag). Among these, Cu exhibits strong localized enrichment and spatial differentiation, making it a key indicator for copper polymetallic deposits.
• Traditional methods like IDW and Kriging effectively detected high to moderate anomalies in the western Duobaoshan–Sankuangou area but failed to identify anomalies in the eastern regions. Traditional methods’ average anomaly mineral point coverage was 50.53% (p < 0.05, 95% CI: 33.97%–67.09%). In contrast, the DE-WDCB method, by accounting for topographic complexities and applying trend surface analysis, successfully detected and amplified low to weak anomalies, particularly in the Luotuowaizi area. The DE-WDCB method showed an average anomaly mineral point coverage of 63.57% (p < 0.05, 95% CI: 47.64%–79.50%). This method demonstrated superior spatial coverage and a stronger correlation with known mineralization points, underscoring its effectiveness in regions with complex topographic features.
• Future exploration efforts should prioritize the DE-WDCB method, especially in areas like the triangular region between Duobaoshan, Yubaoshan, and Sankuangou (Region A) and the Sandaowan–Luotuowaizi area (Region B). In Region B, the exploration should focus on avoiding the Ti anomaly-rich Songshugangzi area and instead on the higher-confidence zones identified by the DE-WDCB method. Moreover, we believe that future research should focus on expanding the application of the DE-WDCB method to diverse geological settings, particularly in areas with varied geomorphological and geochemical characteristics, to validate its effectiveness further. Integrating advanced geospatial data, such as remote sensing images, could enhance anomaly detection accuracy. Optimizing sampling strategies by exploring the impact of different sampling densities and distributions will also improve geochemical anomaly extraction.