Regarding the source area paleogeographical reconstruction, classical geomorphology theory highlights two important source erosion processes, namely, a) the decrease in source altitude (Figure 3A) (Yang L.Z, 1985), and b) the migration of the watershed (Figure 3B; South China Agricultural University, 1980). Based on studies on erosion evolution models focusing on source area residual geomorphology and using the sediment backfilling method (Helland et al., 2016; Shao et al., 2019), the paleogeography of this area was reconstructed.

The methods for sedimentary strata reconstruction paleogeomorphology mainly include geophysical (Bai et al., 2002), imprinting (Xia and Ma, 1999), compaction thickness restoration (Jiang et al., 2009), residual thickness (He et al., 2007), and stratigraphic trend methods (Li et al., 2007). Among them, the imprinting method considers the influence of pre-sedimentary structures and can quantitatively represent paleogeomorphology. However, it cannot effectively restore denudation (Xia and Ma, 1999), whereas the stratigraphic trend method can accurately do this (Li et al., 2007; Figure 4). Therefore, in the continental basin, using the imprinting method combined with the stratigraphic trend method is more accurate for reconstructing sedimentary strata paleogeographical characteristics.

The sediment transport on the Miaoxibei Uplift followed valleys, which can divide into V-shaped, U-shaped, and compound valleys, as well as single fault type troughs (Figure 7B). V1, V2, V3, V4, V5, and V6 were single fault troughs, compound valleys, U-shaped, V-shaped, single fault troughs, and V-shaped, respectively.

Based on drilling, thin-section, core, and seismic data, it was found the sediments under study represent fan deltas and lakes.

Fan deltas are composed of conglomeratic sandstone, fine sandstone, and mudstone, and their logging characteristics are dominated by a toothed bell shape and funnel shape (Figure 8). Wall-core data from the second member of the Shahejie Formation of Well C-1 indicated that it consists of light-yellow, gravel-bearing fine sandstone that is poorly sorted and rounded (Figures 9A–C).

Thin-section data from a depth of 2,707.5 m in the second member of Well C-1 showed that it contained coarse sandstone with fine gravel, which is poorly sorted and rounded (Figure 9D). The composition of clastic particles is 85%, the content of matrix is 5%, and the content of cement is 7%. The particle composition includes quartz, debris, feldspar, and a small amount of mica. The gravel and debris are mainly quartzite, shallow-grained rock, tuff, and granite.

Thin-section data from a depth of 2,096 m in the second member of Well A-1 showed that it is composed of carbonate containing argillaceous silty feldspathic lithic sandstone, with a particle size ranging from 0.02 to 1.2 mm and a high clay content (Figure 9E). The clastic particle composition is 79%, the matrix content is 7%, and the cement content is 9%. The gravel and rock debris are mainly quartzite, shallow grained rock, tuff, and metamorphic granite (Figure 9E). Thin section data show that the composition and mineral association of sandstone in the second member of the Shahejie Formation are well matched with quartz schist and granite in the Miaoxibei Uplift, and the sorting and roundness are bad, indicating that its source is from the Miaoxibei Uplift, the transport distance is relatively close and the accumulation rate is fast.

Given the relatively limited thickness layer of the second member (less than 50 ms and 50 m in the seismic and Well data, respectively), it appeared as a group of low-to-mid-frequency reflection phase axes in the seismic profile. This made it difficult to determine its external geometry and internal reflection structures (Figure 10). Therefore, the fan delta facies could only be distinguished from the shallow lake facies based on their seismic amplitude and continuity, which were characterized by medium to strong amplitude reflections and moderate to poor continuity. The drilling data showed that the mean amplitude seismic attribute of the second member had a strong response to the fan delta sedimentation and could be identified reliably (Figure 11).

Logging data calibration of the seismic attributes (average amplitude value) of the second member showed that fan deltas feature an abnormally high amplitude (red and yellow areas) (Figure 11A). This indicated their planar distribution, which varied considerably, with isolated buds as the main type in the south and strip-like, vertically aligned longitudinal axes trending toward north-northwest in the north. The middle fan developed at a large scale, gradually decreasing toward both sides (Figures 11A, B).

Seismic data indicated that in the southern part of the study area, the second member of the Shahejie Formation was predominantly controlled by faults with large displacement, and formation margin erosion was weak (Figure 10), resulting in relatively well-preserved stratigraphic boundaries. In the north of the study area, the fault displacement was relatively small, and formation margin erosion was relatively strong, resulting in relatively bad-preserved stratigraphic boundaries. Based on the paleogeomorphological reconstruction results, the thickness of the second member of the Shahejie Formation in Bodong Sag is between 50 and 150 m, and it is thinner in the east, thicker in the west, thinner in the south, thicker in the north, and the depositional center is in the north, the erosion in the northern part was relatively strong, and the eroded strata extended farthest toward the east at 4.8 km (Figures 12A, B).

The Miaoxibei Uplift top surface geomorphological characteristics indicated that the slopes on the eastern and western sides of the main watershed are asymmetric (Figure 7A). According to modern geomorphological theory, several factors can cause slope asymmetry on either side of watersheds, such as the influence of geological structures and asymmetrical erosion at the base level and differences in rock properties or faults.

The study area structural characteristics suggested that the Miaoxibei Uplift western slope is gentle and belongs to the Bodong Sag gentle slope zone. In contrast, the eastern slope is steeper and belongs to the Miaoxi Sag steep slope zone (Figure 10). This indicates that the slopes on the Miaoxibei Uplift eastern and western sides may have been asymmetrical during the Bodong Sag rift stage.

The characteristics and scale of sedimentary filling (Figures 10, 11) indicate that the Paleogene strata in the Bodong Sag have a greater thickness and wider spatial distribution range compared with those in the Miaoxi Sag, which were thinner with a narrower spatial distribution range. In terms of the material conservation of the source-to-sink system, this suggests that during the Paleogene, erosion on the two sides of the Miaoxibei Uplift was uneven, with the western side experiencing stronger erosion than the eastern side. The main watershed gradually shifted to the east under the influence of uneven and asymmetrical erosion at the base level.

Based on this, the Miaoxibei Uplift erosion process was traced back by analyzing a series of seismic profiles (Figure 10). During the Paleogene, the main watershed position migrated from west to east (Figure 13A). Under the watershed migration pattern constraint, the Miaoxibei Uplift paleogeomorphological reconstruction was completed (Figure 13B). Based on paleogeomorphology reconstruction and under the constraints of geomorphology, the drainage system above the source of the study area was reproduced by means of digital differential erosion simulation and three-dimensional proportional fluid simulation, and the drainage system corresponding to the fan delta in the study area was picked up and combined (Figure 13B). The drainage system in the middle part of the Miaoxibei Uplift was larger in scale, extended longer and had more tributaries. The drainage systems in the south and north are smaller in scale, with shorter distances and fewer tributaries.

Based on paleogeomorphology reconstruction and under the constraints of geomorphology, the drainage system above the source of the study area is reproduced by means of digital differential erosion simulation and three-dimensional proportional fluid simulation, and the main drainage system corresponding to the fan delta in the study area is picked up and combined in the form of lines (Figure 13B). The ancient water system in the central part of the temple northwest bulge is larger in scale, longer in extension distance and more tributaries, while it is smaller in the south and north with shorter in extension distance and fewer tributaries.

Based on the paleogeomorphological reconstruction in the sedimentary area and the sedimentary system distribution, as well as the paleogeomorphological reconstruction in the source area, the source-to-sink system of the second member of the Shahejie Formation in the Miaoxibei area was reconstructed (Figure 13B). The results show that the Miaoxibei Uplift northern and southern sides comprise large areas with high altitudes and a strong sediment supply capacity. The Miaoxibei Uplift main watershed gradually shifted to the east during the Paleogene, and the transport pathways corresponded well to the fan deltas. The study area southern part is dominated by isolated fan deltas, while the northern part is dominated by continuous fan deltas, with a larger scale in the central area and a smaller scale on the northern and southern sides.

According to the reconstruction results, the sources in the southern and northern parts of the study area comprise a large area at high altitude, resulting in large-scale fan delta development (Figure 13B). During the sedimentation period of the second member of the Shahejie Formation, the tectonic activity was mainly thermal subsidence and weak (Figure 2), with a relatively small influence on the source-to-sink system process. The bedrock types in the Miaoxibei Uplift were mainly granite and quartzite, and modern geomorphic research shows that their resistance to weathering is similar (Yu et al., 2013). Therefore, we can rule out the influence of strong tectonic movement and provenance lithology difference on the development scale of fan delta in the study area.

Based on the reconstructed results, the parameters of various source-to-sink systems in the study area are quantitatively calculated by the method of calculus ( Amorosi et al., 2016; Liu E. T. et al., 2020 ) (Table 1), and the correlations among the elements of various source-to-sink systems are analyzed.

The correlation analysis results of the height of the source, area of the source, volume of the source, length of the drainage system, and the cross-sectional area of the transport pathway show that (Figures 14A–D), the coefficient of determination (R ²) of the volume of the source, area of the source, length of the drainage system, and cross-sectional area of the transport channel are all less than 0.5, in which the coefficient of determination of height of the source and the cross-sectional area of the transport channel is relatively high, and the R ² is 0.447. The cross-sectional area of the transport channel has low correlation with the length of the drainage system, area of the source, and the volume of the source.

According to the correlation analysis results of the area of the source, height of the source, volume of the source, length of the drainage system, width of the transport pathway, depth of the transport pathway, width to depth ratio of the transport pathway, cross-sectional area of the transport pathway, and volume of the delta (Figures 14D–K), the volume of the source, width depth ratio of the transport pathway, and the volume of the delta have the strongest correlation, and the coefficient of determination (R ²) are 0.55 and 0.78. The correlation between the area of the source, height of the source, depth of the transport pathway, width of the transport pathway, channel cross-sectional area of transport pathway, and volume of the delta is low, and the correlation is power function or polynomial, and the coefficient of determination (R ²) is between 0.3 and 0.4. The coefficient of determination (R ²) of length of the drainage system and the volume of the delta is less than 0.2, and there is no obvious correlation between them.

In the study area, the volume of fan delta has a strong correlation with the volume of the source and the width depth ratio of the transport channel, but a weak correlation with the height of the source, area of the source, and the length of drainage system (Figure 14).

From the essence of “conservation of matter” in the “source-to-sink” system, the amount of denudation produced in the source directly determines the size of the delta volume (Xie et al., 2014; Helland et al., 2016; Zeng et al., 2019). Therefore, the delta volume should be the total amount of debris, which depends on the area of denudation, bedrock lithology, geomorphological characteristics, tectonic activity, and weathering intensity in the basin, independent of the transport distance and mode. This well explains the phenomenon in this study that the volume of delta is not obviously related to the transport distance, but is related to the volume of the source.

It is worth mentioning that the width depth ratio of the transport pathway in the study area has a strong correlation with the volume of fan delta, but too large or too small width depth ratio of transport pathway is not conducive to the large-scale development of delta. This phenomenon can be well explained in the theory of erosion cycle proposed by Davis in the middle of 18th century. The width depth ratio of the early development stage of the transport pathway is small, and the sediment flux is also small. In the middle stage, the ratio of width depth ratio was moderate and the deposition flux was maximum. In the late stage, the width depth ratio is larger and the deposition flux was smaller. A similar phenomenon has been reported in the case study of the “source-to-sink” system in the Bozhong Sag, where the width depth ratio of the transport pathway that is too large or too small is not always beneficial for delta enrichment (Zhou et al., 2017).

In areas where there has been relatively little exploration, traditional sedimentary facies research methods are predominantly based on well-log data to mark various geophysical features (Liu et al., 2016; Zhu et al., 2016; Ge et al., 2020; Zhao et al., 2021). The source-to-sink technique works effectively when well-log data is limited (Martinsen et al., 2005; Liu et al., 2019; Lin et al., 2022; Sambrook Smith et al., 2010; Sømme et al., 2009 Sømme and Jackson, 2013; Xu and Du, 2017; Xu et al., 2021; Zhou et al., 2017; Zhu et al., 2014; Zhu et al., 2017). Considering the Miaoxibei area as an example, the sedimentary facies of the second member of the Shahejie Formation have been determined using well-log data and seismic attributes (Figure 11), it was found that transport pathways and fan deltas corresponded more effectively in the study area (Figure 14B). Transport pathways, such as valleys and fault troughs on top of the bedrock, are inheritable and their positions do not change considerably (Yang J. C, 1985; Li, 1999). With a lack of drilling data, this indicates that in the middle and deep layers of the third and fourth members of the Shahejie and Kongdian Formations, the transport channel may also strongly correspond with the fan deltas. This approach can be used to replace drilling data for calibrating geophysical data identifying the distribution of sand-rich sedimentary facies.

The source rock mechanical properties, permeability, mineral composition, and chemical characteristics are likely to directly affect the source supply and reservoir storage performance (Zhu et al., 2017). The carbonate rock weathering products are often transported to the sedimentary basin in the form of dissolution (Yang J. c, 1985; Pei et al., 2008). There are relatively few clastic particles and weaker supply capacity, which often cannot form large-scale sedimentary bodies. After undergoing diagenesis, the calcareous cementation becomes strong, with relatively dense reservoir, which is not conducive to high-quality reservoir development (Li et al., 2005; Feng et al., 2009). The granite rock erosion mainly consists of surface, gully, and collapse erosion, with a strong supply capacity and a high sand fraction (Ruan and Zhou, 1995), providing the conditions needed for high-quality reservoir formation.