Study on the Construction of 3D Geological Model of Quaternary Loose Sedimentary Strata Based on the Global Stratigraphic Discrete Points

: Accurately depicting the spatial structure characteristics of Quaternary loose sedimentary strata is not only of great signiﬁcance for the research of Quaternary geological evolution, but also for the analysis of spatial variation characteristics of the inner hydrogeological and engineering geological attributes of the strata. In this study, an approach for constructing a 3D geological model of Quaternary loose sedimentary strata is proposed based on global stratigraphical discrete points. The approach obtains the discrete control point set of each stratum by using limited borehole data for interpolation and encryption, and the contact relationships and intersection modes of adjacent strata can be determined via the analysis of stratigraphic sequence; ﬁnally, taking these as the professional basis, the construction of the 3D geological model of Quaternary loose sedimentary strata can be carried out. This application can not only accurately describe the three-dimensional spatial distribution characteristics of the Quaternary loose sedimentary strata, it can also be used to perform a layered simulation of the spatial variation characteristics of the inner geological properties of the Quaternary loose sedimentary strata, such as lithology, porosity, and water content, by taking the three-dimensional spatial framework of each stratum as the simulation boundary. Finally, this study takes the citizen center of Xiong’an new area as an example in order to verify the reliability and advancement of the 3D geological modeling scheme.

However, two issues may arise from technical schemes that directly analyze the sedimentary characteristics of Quaternary loose strata based on locally limited data. The first is there is never enough data [25,26], especially for large-scale Quaternary geological research; the second is that Quaternary loose strata, which include deposits from various sedimentary sources, possess complicated spatial structural characteristics [27,28], especially with respect to the origin of continental deposits. Hence, directly studying the sedimentary characteristics of Quaternary loose strata using locally limited data results in a large degree of uncertainty [29].
3D geological modeling is an ideal solution to these issues [30][31][32]. Høyer et al. [33] constructed a 3D geological model using AEM resistivity data; Lau et al. [34] established a 3D geologic model to research the structure of Quaternary deposits; Chen et al. [35] constructed a 3D stochastic model to simulate the characterization of the internal attributes of sedimentary strata; Erharter et al. [36] developed a 3D stochastic model to represent the sediment bodies; and there are also many technical approaches, such as machine learning, that have been applied to optimize the construction of 3D geological models, especially when fractures and fracture networks are involved [37][38][39]. While the research mentioned above has presented excellent approaches for determining the spatial structure or inner attributes of Quaternary loose sedimentary strata, they mainly focused on the 3D visual interpretation of boreholes [40] and geochemical data [41] (mainly seismic data [42,43]), or 3D simulation on a small scale with a strict hypothesis, and there are few reports on the relevance of the construction of 3D geological models of Quaternary loose sedimentary strata based on identifying the spatial distribution in the whole study area by deep mining little pieces of Quaternary geological field data.
In this study, an approach for constructing a 3D geological model of Quaternary loose strata on the basis of global stratigraphic discrete points is proposed. Firstly, the global control point set of each stratum is obtained by establishing the geostatistical model based on the Quaternary geological field data; then, the contact relationships and intersection modes of the adjacent strata are determined, which are applied as the professional basis for the 3D geological modeling via an analysis of stratigraphic sequence; finally a 3D geological model of Quaternary loose sedimentary strata is constructed using the global discrete control points of each stratum.

Global Discrete Point Interpolation and Encryption of Quaternary Loose Sedimentary Stratum
The global discrete points of Quaternary loose sedimentary stratum, which represent the spatial distribution of each stratum in the whole study area, are obtained via interpolation and encryption using limited field data based on geostatistical theories such as Kriging, inverse distance weighting, etc. (Figure 1). In essence, global interpolation and encryption cannot add new stratification information, but it can coordinate the impact of Quaternary stratification data on other locations without geological data in a more scientific way, so as to obtain the globally optimal analysis results. research; the second is that Quaternary loose strata, which include deposits from various sedimentary sources, possess complicated spatial structural characteristics [27,28], especially with respect to the origin of continental deposits. Hence, directly studying the sedimentary characteristics of Quaternary loose strata using locally limited data results in a large degree of uncertainty [29]. 3D geological modeling is an ideal solution to these issues [30][31][32]. Høyer et al. [33] constructed a 3D geological model using AEM resistivity data; Lau et al. [34] established a 3D geologic model to research the structure of Quaternary deposits; Chen et al. [35] constructed a 3D stochastic model to simulate the characterization of the internal attributes of sedimentary strata; Erharter et al. [36] developed a 3D stochastic model to represent the sediment bodies; and there are also many technical approaches, such as machine learning, that have been applied to optimize the construction of 3D geological models, especially when fractures and fracture networks are involved [37][38][39]. While the research mentioned above has presented excellent approaches for determining the spatial structure or inner attributes of Quaternary loose sedimentary strata, they mainly focused on the 3D visual interpretation of boreholes [40] and geochemical data [41] (mainly seismic data [42,43]), or 3D simulation on a small scale with a strict hypothesis, and there are few reports on the relevance of the construction of 3D geological models of Quaternary loose sedimentary strata based on identifying the spatial distribution in the whole study area by deep mining little pieces of Quaternary geological field data.
In this study, an approach for constructing a 3D geological model of Quaternary loose strata on the basis of global stratigraphic discrete points is proposed. Firstly, the global control point set of each stratum is obtained by establishing the geostatistical model based on the Quaternary geological field data; then, the contact relationships and intersection modes of the adjacent strata are determined, which are applied as the professional basis for the 3D geological modeling via an analysis of stratigraphic sequence; finally a 3D geological model of Quaternary loose sedimentary strata is constructed using the global discrete control points of each stratum.

Global Discrete Point Interpolation and Encryption of Quaternary Loose Sedimentary Stratum
The global discrete points of Quaternary loose sedimentary stratum, which represent the spatial distribution of each stratum in the whole study area, are obtained via interpolation and encryption using limited field data based on geostatistical theories such as Kriging, inverse distance weighting, etc. (Figure 1). In essence, global interpolation and encryption cannot add new stratification information, but it can coordinate the impact of Quaternary stratification data on other locations without geological data in a more scientific way, so as to obtain the globally optimal analysis results.

Analysis of the Stratigraphic Sequence of Quaternary Loose Sedimentary Strata
The spatial structure characteristics and stratigraphic sequence of Quaternary loose strata are objective and unique, while the geological information acquired from field exploration such as drilling or profiling is only representative of local areas, and only includes fragments of the overall stratigraphic sequence of the study area. Therefore, it is necessary to integrate all of the field exploration data to determine the synthesized stratigraphic sequence of the Quaternary loose sedimentary strata in the whole study area (Figure 2), which can be used as a temporal and spatial framework for stratigraphic spatial structure analysis and stratigraphic sedimentary evolution.
The stratigraphic sequence determines how the contact relationships between strata are to be treated, which controls the spatial structure of Quaternary loose sedimentary strata. The construction of a 3D geological model of Quaternary loose sedimentary strata requires the synthetic stratigraphic sequence of the study area, which can be obtained by integrating the internal sequence information of all geological data. The spatial structure characteristics and stratigraphic sequence of Quaternary loose strata are objective and unique, while the geological information acquired from field exploration such as drilling or profiling is only representative of local areas, and only includes fragments of the overall stratigraphic sequence of the study area. Therefore, it is necessary to integrate all of the field exploration data to determine the synthesized stratigraphic sequence of the Quaternary loose sedimentary strata in the whole study area (Figure 2), which can be used as a temporal and spatial framework for stratigraphic spatial structure analysis and stratigraphic sedimentary evolution.
The stratigraphic sequence determines how the contact relationships between strata are to be treated, which controls the spatial structure of Quaternary loose sedimentary strata. The construction of a 3D geological model of Quaternary loose sedimentary strata requires the synthetic stratigraphic sequence of the study area, which can be obtained by integrating the internal sequence information of all geological data.

Contact Relationships of Quaternary Loose Sedimentary Strata
As shown in Figure 3, the different types of contact relationships of Quaternary loose sedimentary strata mainly include conformity, disconformity(para-unconformity), unconformity, erosion, etc. For the construction of a 3D geological model of Quaternary strata, since the stratigraphic sequence is unique, the treatment of disconformable contact relationship is similar to the conformity; regarding unconformity contact, its formation is mainly related to differences in the topography, hydrodynamic conditions, and provenance characteristics of the deposition process, which results in local angular unconformity contact relationships between adjacent strata of different depositional origins, and the stratigraphic sequence is that the old stratum is underneath the new; for erosion contact, the deposition process takes place when part of the old stratum is denuded by erosion due to water or wind flow, and then new sediments fill the eroded area, and hence the stratigraphic sequence is that the new stratum lies beneath the old. The contact relationship is one of the key factors for determining the spatial pattern of the strata. For 3D geological modeling based on the global discrete points, it is necessary to use the 4 types of contact relationships to deal with the intersection modes of adjacent strata.

Contact Relationships of Quaternary Loose Sedimentary Strata
As shown in Figure 3, the different types of contact relationships of Quaternary loose sedimentary strata mainly include conformity, disconformity(para-unconformity), unconformity, erosion, etc. For the construction of a 3D geological model of Quaternary strata, since the stratigraphic sequence is unique, the treatment of disconformable contact relationship is similar to the conformity; regarding unconformity contact, its formation is mainly related to differences in the topography, hydrodynamic conditions, and provenance characteristics of the deposition process, which results in local angular unconformity contact relationships between adjacent strata of different depositional origins, and the stratigraphic sequence is that the old stratum is underneath the new; for erosion contact, the deposition process takes place when part of the old stratum is denuded by erosion due to water or wind flow, and then new sediments fill the eroded area, and hence the stratigraphic sequence is that the new stratum lies beneath the old. The contact relationship is one of the key factors for determining the spatial pattern of the strata. For 3D geological modeling based on the global discrete points, it is necessary to use the 4 types of contact relationships to deal with the intersection modes of adjacent strata. Water 2022, 13, x FOR PEER REVIEW 4 of 11 The discrete point set of a stratum obtained by interpolation and encryption is evenly distributed throughout the study area, which results in the control surface generated from it also being globally distributed. Therefore, it is necessary to topologically cut off the redundant part of the stratum according to the intersection mode of the adjacent strata, which is determined by the contact relationship.
As shown in Figure 4, for conformity (disconformity), no special treatment is needed, and the 3D geological model of the stratum will be automatically constructed by tracking the stratigraphic boundary; for unconformity, the underlying old stratum is taken as the basement, and the superfluous parts of the overlying stratum that overlap with the underlying old stratum are cut off, while retaining the other parts that do not overlap with the old; conversely, the erosion contact needs to cut out and delete the parts of the underlying old stratum that overlap with the overlying new one.

Techincal Scheme
The technical scheme ( Figure 5) of the construction of the 3D geological model of Quaternary loose sedimentary strata based on global stratigraphic discrete points obtains a global stratigraphic control point set by applying the geological field data for interpolation and encryption, and generates a control surface for each sedimentary stratum based on this; then, the contact relationships and intersection modes of the strata are determined on the basis of stratigraphic sequence analysis; finally, the construction of the 3D geological model of the Quaternary loose sedimentary strata can be carried out. The discrete point set of a stratum obtained by interpolation and encryption is evenly distributed throughout the study area, which results in the control surface generated from it also being globally distributed. Therefore, it is necessary to topologically cut off the redundant part of the stratum according to the intersection mode of the adjacent strata, which is determined by the contact relationship.
As shown in Figure 4, for conformity (disconformity), no special treatment is needed, and the 3D geological model of the stratum will be automatically constructed by tracking the stratigraphic boundary; for unconformity, the underlying old stratum is taken as the basement, and the superfluous parts of the overlying stratum that overlap with the underlying old stratum are cut off, while retaining the other parts that do not overlap with the old; conversely, the erosion contact needs to cut out and delete the parts of the underlying old stratum that overlap with the overlying new one.  The discrete point set of a stratum obtained by interpolation and encryption is evenly distributed throughout the study area, which results in the control surface generated from it also being globally distributed. Therefore, it is necessary to topologically cut off the redundant part of the stratum according to the intersection mode of the adjacent strata, which is determined by the contact relationship.
As shown in Figure 4, for conformity (disconformity), no special treatment is needed, and the 3D geological model of the stratum will be automatically constructed by tracking the stratigraphic boundary; for unconformity, the underlying old stratum is taken as the basement, and the superfluous parts of the overlying stratum that overlap with the underlying old stratum are cut off, while retaining the other parts that do not overlap with the old; conversely, the erosion contact needs to cut out and delete the parts of the underlying old stratum that overlap with the overlying new one.

Techincal Scheme
The technical scheme ( Figure 5) of the construction of the 3D geological model of Quaternary loose sedimentary strata based on global stratigraphic discrete points obtains a global stratigraphic control point set by applying the geological field data for interpolation and encryption, and generates a control surface for each sedimentary stratum based on this; then, the contact relationships and intersection modes of the strata are determined on the basis of stratigraphic sequence analysis; finally, the construction of the 3D geological model of the Quaternary loose sedimentary strata can be carried out.

Techincal Scheme
The technical scheme ( Figure 5) of the construction of the 3D geological model of Quaternary loose sedimentary strata based on global stratigraphic discrete points obtains a global stratigraphic control point set by applying the geological field data for interpolation and encryption, and generates a control surface for each sedimentary stratum based on this; then, the contact relationships and intersection modes of the strata are determined on the basis of stratigraphic sequence analysis; finally, the construction of the 3D geological model of the Quaternary loose sedimentary strata can be carried out.

Study Area
The study area is located in the citizen center of Xiong'an new area, with an area of approximately 500 × 600 m 2 ( Figure 6). There are huge and thick Quaternary loose strata in the area, and their origin is very complex, including the alluvial-proluvial deposits of the Taihang mountains located in the northwest of Xiong'an new area, the floodplain deposits and shallow lacustrine of Baiyangdian lake, as well as fluvial deposits, etc. The modeling scope is a depth of 15 m below the surface.

Study Area
The study area is located in the citizen center of Xiong'an new area, with an area of approximately 500 × 600 m 2 ( Figure 6). There are huge and thick Quaternary loose strata in the area, and their origin is very complex, including the alluvial-proluvial deposits of the Taihang mountains located in the northwest of Xiong'an new area, the floodplain deposits and shallow lacustrine of Baiyangdian lake, as well as fluvial deposits, etc. The modeling scope is a depth of 15 m below the surface.

Modeling Scheme
This study determined the sedimentary facies information of 81 boreholes with a depth of 15 m via granulometric analysis and stratigraphic sequence analysis, of which 70 were used as the data to construct a 3D geological structure model of Quaternary loose sedimentary strata, and the other 11 were used as data to verify the model's accuracy. This study mainly used DeepInsight software as the modeling platform, and the specific modeling process was as follows (Figure 7).

Modeling Scheme
This study determined the sedimentary facies information of 81 boreholes with a depth of 15 m via granulometric analysis and stratigraphic sequence analysis, of which 70 were used as the data to construct a 3D geological structure model of Quaternary loose sedimentary strata, and the other 11 were used as data to verify the model's accuracy. This study mainly used DeepInsight software as the modeling platform, and the specific modeling process was as follows (Figure 7).

Results
The study area is located at the junction of the margin of the piedmont alluvial fan of the Taihang mountains and the catchment area of Baiyangdian lake, where the terrain is relatively flat and tectonic activity has been quite weak since the Holocene of the Quaternary; hence, there are four types of sedimentary genesis: floodplain, alluvial-proluvial,

Results
The study area is located at the junction of the margin of the piedmont alluvial fan of the Taihang mountains and the catchment area of Baiyangdian lake, where the terrain is relatively flat and tectonic activity has been quite weak since the Holocene of the Quaternary; hence, there are four types of sedimentary genesis: floodplain, alluvialproluvial, shallow lacustrine and fluvial, from top to bottom. The 3D geological model of the Quaternary loose sedimentary strata of the citizen center is shown in Figure 8.

Results
The study area is located at the junction of the margin of the piedmont alluvial fan of the Taihang mountains and the catchment area of Baiyangdian lake, where the terrain is relatively flat and tectonic activity has been quite weak since the Holocene of the Quaternary; hence, there are four types of sedimentary genesis: floodplain, alluvial-proluvial, shallow lacustrine and fluvial, from top to bottom. The 3D geological model of the Quaternary loose sedimentary strata of the citizen center is shown in Figure 8.  According to the results of Quaternary geological exploration, the thickness of the floodplain sedimentary stratum in the study area was about 5 m, was relatively uniform, and covered the whole study area; the alluvial-proluvial sedimentary layer was also the main stratum, and the thickness was about 5-8 m, which varied greatly across the whole area, and even exhibited partial loss in the southwest of the study area; the shallow lacustrine layer covered the whole study area, with the thickness varying sharply between 2 and 8 m, controlled by the evolution of Baiyangdian lake; and the thickness of the fluvial deposit layer, which is dominated by the meandering river sediments, was mostly about 3-6 m, with partial loss in the middle and southwest of the study area ( Figure 9). According to the results of Quaternary geological exploration, the thickness of the floodplain sedimentary stratum in the study area was about 5 m, was relatively uniform, and covered the whole study area; the alluvial-proluvial sedimentary layer was also the main stratum, and the thickness was about 5-8 m, which varied greatly across the whole area, and even exhibited partial loss in the southwest of the study area; the shallow lacustrine layer covered the whole study area, with the thickness varying sharply between 2 and 8 m, controlled by the evolution of Baiyangdian lake; and the thickness of the fluvial deposit layer, which is dominated by the meandering river sediments, was mostly about 3-6 m, with partial loss in the middle and southwest of the study area ( Figure 9).  Finally, we verified the accuracy of the model by means of strata comparison between the virtual boreholes extracted from the model at the same location as the verification drillings and the 11 verification drillings themselves, finding that the accuracy of the model was 92.7%. The verification results show that the modeling approach based on the global strata discrete points is capable of accurately constructing a 3D geological model of Quaternary loose sedimentary strata.

Discussion
In addition to accurately constructing the 3D model of Quaternary loose sedimentary strata, the most significant application of the 3D geological structure modeling technology based on the global stratigraphical discrete points was the layered simulation of the spatial variation characteristics of the internal properties of strata, such as lithology, porosity, and water content.
Due to the differences in sedimentary genesis, the spatial variation characteristics of lithology, porosity, water content, and the other internal attributes of sedimentary strata will be different. The layered geological structure model is able to simulate the spatial variation characteristics of the internal properties of each stratum individually by using itself as the simulation boundary. Therefore, compared with the overall simulation, the accuracy of the layered simulation is always much higher. To verify it, in this study, we established a lithology model of the study area using the sequential indicators stochastic simulation method based on the Quaternary loose sedimentary strata model (Figure 10), and the levels of accuracy of the lithology model were 93.5% (floodplain), 95.8% (alluvial-proluvial), 95.2% (shallow lacustrine) and 94.1% (fluvial), which are significantly higher than the 85.3% achieved by the overall stochastic simulation.

Conclusions
The 3D geological modeling of Quaternary loose sedimentary strata based on the global stratigraphic discrete points, which were firstly used to obtain a spatial structure control point set by applying limited field geological data for interpolation and encryption based on Geostatistics theory; then, the contact relationships and intersection modes were determined via stratigraphic sequence analysis; and finally a model of the strata was constructed using these as the geological basis.
Taking the citizen center of Xiong'an new area as the study area, this study constructed a 3D geological model of four deposited genetic types of Quaternary loose sedi-

Conclusions
The 3D geological modeling of Quaternary loose sedimentary strata based on the global stratigraphic discrete points, which were firstly used to obtain a spatial structure control point set by applying limited field geological data for interpolation and encryption based on Geostatistics theory; then, the contact relationships and intersection modes were determined via stratigraphic sequence analysis; and finally a model of the strata was constructed using these as the geological basis.
Taking the citizen center of Xiong'an new area as the study area, this study constructed a 3D geological model of four deposited genetic types of Quaternary loose sedimentary strata, including the floodplain, the piedmont alluvial-proluvial of the Taihang mountains, the shallow lacustrine of Baiyangdian lake, and fluvial, from top to bottom, with a depth of 15 m from the surface by using the 3D geological modeling approach based on global stratigraphic discrete points. Finally, we evaluated the accuracy (92.7%) of the model by extracting virtual boreholes for stratigraphic comparison with the verification boreholes. The 3D structure model of each sedimentary stratum can be used as a constraint framework to establish a simulation model of the spatial variation of the properties of each stratum, respectively, such as lithology, water content, and porosity. Compared with the overall simulation, analysis results with higher precision were obtained.

Conflicts of Interest:
The authors declare that they have no conflict of interest.