A total of 597 heavy mineral data in 32 wells (Figure 1C) and 7 sedimentary sections (Figure 1C) were used for provenance analysis, among which 23 heavy mineral data were conducted in the Langfang Yuneng Rock and Mineral Processing Co., Ltd. in Langfang, Hebei Province, and the other heavy mineral data were collected from Zhou et al. (2019). The detailed heavy mineral analysis process followed the methods described by Mange and Maurer (2012). Authigenic minerals were excluded from the analysis.

A matrix heatmap was used to represent individual values with different colors in the heavy mineral analysis (Zhu et al., 2017; Zhou et al., 2019). The heatmaps were generated by HemI version 1.0.3.3 software. The blue to red scale in the heatmaps represents low to high percentages of heavy minerals (Deng et al., 2014). Each column presents all the data for one specific heavy mineral from all the samples, and each row shows all the heavy mineral compositions of one specific sample. Different heavy mineral assemblages were generated by hierarchical cluster analysis using the average-linkage clustering algorithm. One pie chart on the heavy mineral assemblage distribution map for one specific stratum represents different heavy mineral assemblages indicating different source areas.

The conglomerate outcrop data include conglomerate components and grain sizes from six sections, including Sikeshu River, Manas, Taxi River, Quergou, Baiyanghe, and Dongwozi section (Figure 1). In each section, ca. 100 clasts were counted in an area of 1 m² to analyze the provenance. The lithology of the conglomerate grains was identified in the field (see Supplementary Figure S4) and checked under an optical microscope with thin sections.

The detrital zircon geochronology of five samples was conducted in this study, including JGMK-026 (Kalazha Formation), HG-007 (Kalazha Formation), JY-2718 (Qingshuihe Formation), CS-1045 (Kalazha Formation), and JGMK-148 (Qigu Formation). We also collected detrital zircon geochronological data of eight published samples to compare with the new data, including sample SKSH-012 (Qingshuihe Formation, SKSH-052 (Qigu Formation, XJ10-016 (Qingshuihe Formation, Yang et al., 2013; Guan et al., unpublished data ¹ ), Tou-08 (Qingshuihe Formation, Fang et al., 2015), Tou-07 (Kalazha Formation, Fang et al., 2016), XJ09-100 (Qigu Formation, Yang et al., 2013), Tou-06 (Qigu Formation, Fang et al., 2015), TS12-110 (Qigu Formation, Tang et al., 2014). Sample SKSH-012 and SKSH-052 are from the Sikeshu River section (Figure 1C); sample JGMK-026 and JGMK-148 is from the Jiangong mine section (Figure 1C); the sample XJ10-016, XJ09-100, and HG-07 are from the Manas section (Figure 1C); sample Tou-08, Tou-07, Tou-06 are from the Toutunhe section (Figure 1C); sample TS12-110 is from Baiyanghe section (Figure 1D); sample CS-1045 is from well CS1 (Figure 1D); sample JY-2718 is from well JY2 (Figure 1D).

After separation by standard heavy liquid and magnetic techniques, 250 detrital zircon grains were randomly purified under a binocular microscope for each sample. The grains from samples JGMK-026 and HG-007 underwent cathodoluminescence (CL) imaging with a Quanta 200 FEG scanning electron microscope and U–Pb isotope analyses with LA–ICP–MS (Agilent 7500c ICP–MS coupled with a 193-nm ArFExcimer laser) at the Key Laboratory of Orogen and Crust Evolution, Peking University; the grains from samples JY-2718, CS-1045, and JGMK-148 underwent CL imaging and U–Pb isotope analyses using LA–ICP–MS (Agilent 7900 ICP–MS coupled with ATL (ATLEX 300) excimer laser and a two-volume S155 ablation cell) at Beijing GeoAnalysis Co., LTD. The analysis procedures proposed by Yuan et al. (2004) were followed. Isotopic ratios and elemental concentrations of zircon grains were calculated using the GLITTER 4.0 program (Jackson et al., 2004). Common lead was corrected using the method of Andersen (2002). The age calculations were performed using ISOPLOT 3 (Ludwig, 2003). The ²⁰⁷Pb/²⁰⁶Pb age was utilized when the age was >1,000 Ma, and the ²⁰⁶Pb/²³⁸U age was utilized when the age was <1,000 Ma. Only ages with concordance degrees >90% were analyzed (Supplementary Figures S2, S3).

Based on field investigations, eleven lithofacies were delineated in the SJB (Table 1; Figure 3), including fine, sandstone, and conglomerate lithofacies. Fine lithofacies consist of Fsm and Fl lithofacies (Table 1). The lithologies include centimeter-to decimeter-thick fine sandstones, siltstones, siltstones, or mudstones (Figures 3A,B); the sandstone lithofacies include Sr, Sc, Sm, Sh, Sp, St, and SGF lithofacies (Table 1). These sandstone lithofacies have different sedimentary structures and were formed via different depositional processes (Table 1; Figures 3C–F). The conglomerate lithofacies is composed of Gmm and Gcm lithofacies (Table 1). The Gmm lithofacies is composed of granule to pebble beds that are tens of meters thick and matrix-supported (Figure 3G), while the Gcm lithofacies comprises decimeter-to meter-thick granule beds and is clast-supported (Figure 3H). The conglomerate components of Gmm lithofacies include both well-rounded gravels and poor-rounded gravels. The.

Seismite structures were observed in layered sheet-flow deposits with Gmm and Sm lithofacies, occurring in the Kalazha Formation in the Manas, Taxi River, Quergou, Jiangong mine, Baiyanghe, and Shuimohe section (Figures 4A,B, D–H, and the Qigu Formation in the Manas section (Figure 4C). Some sandstone veins are distorted and cut into several pieces (Figures 4A–D). Water escape structures are also present (Figures 4E–H). Tube-like sandstone veins connect to different sandstone layers (Figures 4G,H). The seismite structure extends laterally for hundreds of meters. Sandstone or gravelly sandstone veins penetrate the underlying layers by 10 cm to several meters (Figures 4A–F). Seismites appear frequently in the middle part of the Kalazha Formation and constitute tens of high-frequency cycles (Figure 4A).

Eleven lithofacies have been distinguished and grouped into seven facies associations based on field investigations. Sedimentary environments have been interpreted respectively, including alluvial fan, fan delta, aeolian dune, meandering river, alluvial plain, lakeshore, and shallow lake (Table 2).

Meandering river environments prevailed during the deposition of the Qigu Formation in the SJB (Figure 8). The sedimentary environments in which the Kalazha Formation was deposited varied in the SJB. In the Sikeshu River and Manas sections, alluvial fan deposits developed in the Kalazha Formation (Figure 8A). The Manas section possesses the thickest facies association 1 (Figure 8A). In the Qingshuihe, Taxihe, Quergou, Hutubi, Jiangong mine, Dongwoz, Baiyanggou, and Shuimohe sections, and in well JY2 (Figure 1C), both alluvial fan and fan delta environments were identified in the Kalazha Formation (Figure 8A). The Jiangong Mine section exhibits the thickest DF facies association (Figure 8A).

In the Toutunhe, Wangjiazhuang, and Dahongshan sections, and well CS1, eolian dunes in an erg environment were identified in the Kalazha Formation (Figure 8B). A lacustrine environment prevailed during the deposition of the Qingshuihe Formation in the SJB (Figure 8). The lakeshore environment evolved upward into a shallow lake environment from the bottom.

The characteristics of parent rocks in potential provenances, including heavy mineral assemblages, are summarized in Table 3. The heavy mineral assemblages of sedimentary sections and wells are presented in pie charts in the sedimentary comparison profiles (Figure 8). In the Sikeshu River section, an assemblage of epidote + ilmenite in the Qigu Formation evolves into an assemblage of ilmenite + epidote + sphene + magnetite in the Kalazha Formation and epidote + magnetite + ilmenite in the Qingshuihe Formation (Figure 8A); in the Manas section, an assemblage of epidote + magnetite in the Qigu Formation changes into an assemblage of barite + magnetite + zircon in the Kalazha Formation, which changes into an assemblage of epidote + ilmenite + leucoxene (Figure 8A); in the Qingshuihe section, an assemblage of ilmenite + magnetite in the Qigu Formation evolves into an assemblage of spinel + ilmenite + garnet in the Qingshuihe Formation (Figure 8A); in the Jiangong mine section, the primary heavy mineral in the Qigu Formation is epidote, which evolves into an assemblage of ilmenite + leucoxene + garnet in the Kalazha Formation and leucoxene + garnet + apatite in the Qingshuihe Formation (Figure 8A); in the Toutunhe section, the heavy mineral assemblage is epidote + magnetite in the Qigu Formation, which changes into an assemblage of ilmenite + garnet in the Kalazha Formation (Figure 8A); in well CS1, an assemblage of ilmenite + magnetite in the Kalazha Formation evolves into an assemblage containing primary magnetite in the Qingshuihe Formation (Figure 8B); in the Shuimohe section, an assemblage of epidote + magnetite + ilmenite in the Qigu Formation changes into an assemblage of magnetite + zircon + epidote + garnet in the Kalazha Formation, which changes into an assemblage of epidote + magnetite + ilmenite (Figure 8B). The content of epidote increases from the bottom of the Kalazha Formation to the Qingshuihe Formation (Figure 8B); in well JY2, the heavy mineral assemblage is ilmenite + zircon + magnetite in the Qingshuihe Formation (Figure 8B).

The heavy mineral assemblages in the Qigu and Qingshuihe Formations attained from the clustering analysis are shown in Supplementary Table S1. Five and four heavy mineral assemblages were identified in the Qigu and Qingshuihe Formations, respectively (Figure 9). According to these heavy mineral assemblages, three catchments were identified, which were located in the Sikeshu sag, Qigu fault belt, and Fukang fault belt (Figures 1B, 10, 11). The Sikeshu sag and Qigu fault belt are in the northern piedmont of the Tianshan Orogenic Belt and the Fukang fault belt is in the northern piedmont of the Bogda Orogenic Belt. The heavy mineral assemblages in catchments 1 to 3 in the Qigu Formation are magnetite + epidote + ilmenite, epidote + ilmenite + leucoxene, and epidote + leucoxene (Figure 9A); the heavy mineral assemblages in catchments 1 to 3 in the Qingshuihe Formation are epidote + ilmenite + garnet, ilmenite + garnet + epidote, and epidote + ilmenite + zircon + leucoxene + garnet (Figure 9C). In summary, ilmenite and epidote are common heavy minerals in the SJB (Table 4). The Qigu, Kalazha, and Qingshuihe Formations have abundant epidote, magnetite, and garnet, respectively (Table 4).

The conglomerate components in the Kalazha and Qingshuihe Formations are shown in pie charts (Figure 9B) and Table 4. In catchment 1, the primary conglomerate component is mudstone in the Kalazha Formation; in catchment 2, the conglomerate components mainly include sandstone, mudstone, granite, tuff, and andesite in the Kalazha Formation; and in catchment 3, the conglomerate components are primarily composed of siltstone and andesite in the Kalazha Formation (Figure 10B). The primary conglomerate components in the Qingshuihe Formation evolved into siltstone and mudstone (Figure 10B).

The detrital zircon U–Pb ages of five samples (JGMK-148, JGMK-026, HG-007, CS-1045, and JY-2718) and eight collected samples in the late Mesozoic SJB are given in Supplementary Material S2. All the detrital zircons of the five samples are magmatic in origin because of their clear oscillatory zoning (Supplementary Figure S1) and high Th/U ratios (see Supplementary Material S3).

The reported crystallization ages of the intermediate-acidic igneous rocks of the North Tianshan, Central Tianshan, and Bogda Orogenic Belts are shown in histograms (Figure 10). All the age data are categorized into seven populations shown in pie charts (Figure 10): Precambrian (2,623–542 Ma), Cambrian–Devonian (541–359 Ma), Mississippian (359–318 Ma), Pennsylvanian (319–299 Ma), Permian (298–251 Ma), Triassic (250–200 Ma), and Jurassic (199–145.5 Ma). Samples 19-SKSH-052, XJ09-100, JGMK-148, HG-007, JY-2718, CS-1045, and TS12-110 have Jurassic age spectrum peaks of ca. 160 Ma (Figure 10); sample JGMK-026 has a Jurassic age spectrum peak of 156 Ma, and samples Tou-06 and Tou-07 have Jurassic age spectrum peaks of 154 Ma (Figure 10). Most of the thirteen samples have Precambrian zircons, except samples JGMK-148, JY-2718, and CS-1045 (Figure 10). Samples 19-SKSH-052, JGMK-148, and Tou-06 have similar age spectra and more Jurassic zircons than sample XJ09-100 (Figure 10).

Although sample JGMK-026 from the Kalazha Formation has more Precambrian ages, the age spectra of samples JGMK-026 and HG-007 from the Kalazha Formation are similar (Figure 10). Sample Tou-07 has more Permian and Triassic zircons than samples HG-007 and JGMK-026 from the Kalazha Formation (Figure 10).

The age spectra of the Qingshuihe sandstones in the catchment 3 (sample JY-2718) differs from the counterparts in the catchment 1 and 2 (sample SKSH-12, XJ10-016, and Tou-08), because it has an apparent Jurassic age spectrum peak (Figure 10). The age spectra of Qingshuihe sandstones in catchments 1 and 2 have strongly varied age distributions and have more Permian to Precambrian zircons (Figure 10). The Qingshuihe sandstone in the catchment 3 (sample JY-2718) has more Carboniferous zircons and fewer Jurassic zircons than the Qigu and Kalazha sandstones (samples TS12-110 and CS-1045) (Figure 10).

The paleocurrent data indicate that the main provenance was to the south of the SJB (Figures 1C,D). The potential provenances of catchments 1 and 2 in the northern piedmont of the Tianshan Orogenic Belt include the North Tianshan and Central Tianshan Orogenic Belt; the potential provenances of catchment 3 in the northern piedmont of the Tianshan Orogenic Belt include the Bogda, North Tianshan, and Central Tianshan Orogenic Belt.

In the SJB, the North Tianshan Orogenic Belt mainly provided sediments to catchments 1 and 2, and the Bogda Orogenic Belt primarily supplied sediments to catchment 3 in the Late Jurassic, which is suggested by proximal fluvial, alluvial, and fan delta environments (Figure 8). In addition, the detrital zircons also provide evidence. The detrital zircon age spectra of the samples in catchments 1 and 2 are compared with the reported crystallization ages of the intermediate-acidic igneous rocks of the North Tianshan and Central Tianshan Orogenic Belt (Figure 10). The distributional ranges of the Kalazha Formation samples in catchments 1 and 2 have a high level of compliance with the North Tianshan Orogenic Belt (Figure 10).

With Precambrian ages and metamorphic rock sources (Table 3 and Figure 10), the Central Tianshan Orogenic Belt constantly supplied sediments to the SJB in the late Mesozoic, which is evidenced by epidotes in most of the heavy mineral assemblages (Table 4; Figures 9A,C–E) and Precambrian ages of most of the samples (Figure 10). More medium-to high-grade metamorphic rocks were denudated in the Early Cretaceous, which is evidenced by the garnets in the heavy mineral assemblages in the Qingshuihe Formation (Table 4 and Figure 10), because the Central Tianshan exhibits medium-to high-grade metamorphic rocks that could have provided garnets. In addition, the distributional ranges of the Qingshuihe Formation samples in the catchments 1 and 2 have a higher level of compliance with the Central Tianshan Orogenic Belt (Figure 10). Therefore, it is inferred that source-to-sink systems expanded in the Early Cretaceous.

Denudation of the volcanoclastic or volcanic rocks of Late Jurassic ages was suggested by detrital zircon spectrum peaks of the Late Jurassic ages (154–162 Ma) in all samples (Figure 10). Although large-scale volcanic activities were not reported in the Junggar Basin and Tianshan area in the Mesozoic (Zhu et al., 2020), basalts, tuffs, tuffaceous sandstones, monzonitic granite, diorite, and trachyandesite of Jurassic ages provide direct evidence of Jurassic magmatism (Xu et al., 2008; Deng et al., 2015; Liu et al., 2018, 2019; Cao et al., 2020). The Late Jurassic age spectrum peaks (Figure 10) are absent in the age spectra of Lower and Middle Jurassic samples (Yang et al., 2013; Fang et al., 2015) and provide indirect evidence of Jurassic volcanism. Considering that Jurassic-age detrital zircons account for large percentages of the detrital zircons, igneous and volcaniclastic rocks were important source rocks in the Late Jurassic and may be extensively distributed around the Junggar Basin and Tianshan Orogenic Belt. Given that large-scale Late Jurassic altered zeolite and bentonite ores from volcanic ejection were found in the Wucaiwan area in the Eastern Junggar Basin (Cao et al., 2020) and that all the reported Meso-Cenozoic magmatic rocks are distributed near major strike-slip faults (Guo et al., 2010), the eruptive center was likely located in the Wucaiwan area in the Eastern Junggar Basin and along the North Tianshan Fault (Fang et al., 2015; Cao et al., 2020) (Figure 9A).

Sediment recycling happened in the basin margin in the late Mesozoic because red layers are observed in the brownish-green conglomerates in the Kalazha Formation, which are assumed that red Late Jurassic sediments were recycled in the earliest Cretaceous (Figure 5F). Besides, well-rounded gravels were found in the Kalazha Formation (Figures 4A,B; Supplementary Figure S4C).

In conclusion, the North Tianshan mainly provided sediments to catchments 1 and 2, and the Bogda Orogenic Belt primarily supplied sediments to catchment 3 in the Late Jurassic. The Central Tianshan Orogenic Belt constantly supplied sediments to the SJB in the late Mesozoic.

From the Early to Middle Jurassic, with a warm and humid climate (Ashraf et al., 2010; Deng et al., 2015; Eberth et al., 2001; Hendrix et al., 1992; Li, et al., 2014), paludal environments dominated in the Early to Middle Jurassic, and coal seams were developed. Under a post-orogenic weak extensional tectonic setting, the Tianshan Orogenic Belt experienced peneplanation, sediments were transported from the Central Tianshan Orogenic Belt to the SJB (Yang et al., 2013; Fang et al., 2015, 2019). The western part of the Bogda Orogenic Belt exhibits slightly positive relief, which cannot block the sediments from the Tianshan Orogenic Belt (Fang et al., 2019). Sediments were transported from the North Tianshan to the Fukang fault belt (catchment 3), which is evidenced by northward paleocurrents and fluvio-lacustrine environments (Fang et al., 2019; Guan et al., 2021).

In the Late Jurassic, the tectonic setting evolved transpressional (He et al., 2018), and the North Tianshan and Bogda Orogenic Belts began to uplift, which is evidenced by a cooling event shown in apatite fission data (Dumitru, et al., 2001; Guo et al., 2006; Shen et al., 2006, 2008; Tang et al., 2015). The tectonic activity was frequent, as evidenced by the frequent occurrence of the seismites in the Kalazha Formation (Figure 4). With the climate turning arid in the SJB (Fang et al., 2016; Jolivet et al., 2017), the coal-bearing strata changed into red beds in the late Middle Jurassic. The accommodation space decreased, and deltaic sedimentary environments gradually changed into fluvial environments, with sediments from the North Tianshan and Central Tianshan Orogenic Belt. The source areas include the Bogda and Tianshan Orogenic Belts (Figure 9A). In the latest Jurassic, the meandering river deposits turned into alluvial fan and eolian deposits during the deposition of the Kalazha Formation, with sediments mainly from the North Tianshan Orogenic Belt in catchments 1 and 2 and the Bogda Orogenic Belt in catchment 3. The source areas decreased and the source-to-sink systems shrank (Figures 9B, 11). The Manas section was the depositional center of the alluvial fan deposits and the Wangjiagou section (Figure 8A) was the depositional center of the eolian deposits during the latest Jurassic (Figure 8B).

In the Early Cretaceous, the lake level rose. Fan deltas replaced the alluvial fans in the SJB during the deposition of the Kalazha Formation in the earliest Cretaceous. Sediments were mainly from the red alluvial fan deposits from the North Tianshan in catchments 1 and 2 and the Bogda Orogenic Belt in catchment 3. When the lake level rose to a high level and became relatively stable, accommodation space increased, and lacustrine environments prevailed. More sediments from the Central Tianshan were transported to the SJB, including sediments from medium-to high-grade metamorphic rocks. The source areas increased and source-to-sink systems expanded (Figures 9C, 11). The depositional center changed from the Manas section to the Jiangong Mine section (Figure 8A).

A sequence boundary is identified between the facies associations 1 and 2 in the Kalazha Formation (Figures 5D,E, 8), because the red facies associations 1 is assumed to be the highstand deposits of the Late Jurassic sequence (Figure 2) and the brownish-green facies association 2 is the lowstand deposit of the overlying sequence. According to Facies association and sedimentary interpretation, The facies association 1 is interpreted as alluvial fan, which is in conformable contact with the underlying Qigu Formation (Figure 6A), and facies association 2 is interpreted as fan delta. The reason why the facies association 1 was deposited before the facies association 2 is that red layers are observed in the brownish-green conglomerates in the Kalazha Formation, which are assumed that red Late Jurassic sediments were recycled in the facies association. Because of the lack of fossils and igneous rocks, the chronological attribution of the Kalazha Formation is hard to determine. The facies association 1 is assigned to the Kimmeridgian according to geochronological, cyclostratigraphic, and magnetostratigraphic studies by Deng et al. (2015). Because the lake transgressed extensively in the Early Cretaceous and the facies association 2 is interpreted as fan delta, the facies association 2 may be the product of lake transgression in the Early Cretaceous. Given that the facies association 2 is in conformable contact and has a gradational transition with the sandstone-dominated Qingshuihe Formation (Figures 7A,B; Eberth et al., 2001), the facies association 2 may be assigned to the Berriasian.

In summary, the sedimentary environments changed from paludal to fluvial and alluvial in the Late Jurassic, and evolved to fan deltaic and lacustrine in the Early Cretaceous. The source-to-sink system shrank in the Late Jurassic and expanded in the Early Cretaceous. Besides, a sequence boundary is identified between the facies associations 1 and 2 in the Kalazha Formation.

The sedimentary and source-to-sink evolution of the intracontinental basin was controlled by tectonics and climate. The shrinkage of the catchment area and the lake in the Late Jurassic was caused by aridification in the Late Jurassic, which is suggested by red beds (Figure 6), eolian dunes (Figures 5G–I), gypsum, desiccation cracks, and the obvious increase in the spore genus Classopollis (Deng et al., 2015; Fang et al., 2016; Jolivet et al., 2017). The sediment flux was also reduced, as evidenced by the reduction in the thickness of the Qigu Formation. The source area also shrank in the Late Jurassic (Figures 9A, 11), as evidenced by the heavy mineral assemblage and detrital zircon U-Pb data (Figures 9A, 10), because the Tianshan and Bogda Orogenic Belts uplifted in the Late Jurassic, which is shown by apatite fission-track data (Dumitru, et al., 2001; Guo et al., 2006; Shen et al., 2006, 2008; Tang et al., 2015). Relatively high relief formed, which may have blocked part of the source area from the Central Tianshan Orogenic Belt. Additionally, the expansion of the catchment area (Figure 9C, Figure 11) in the Early Cretaceous was triggered by the semihumid climate, because the semihumid climate brought an increase in rainfall which led to an extensive lake transgression (Figure 7). The source area expanded in the Early Cretaceous, as suggested by the heavy mineral assemblage and detrital zircon U-Pb data (Figures 9C, 10), which can be attributed to the reduction in the relief of the Tianshan and Bogda Orogenic Belts. More sediments from the Central Tianshan Orogenic Belt were transported to the SJB.

The conglomerates in the Kalazha Formation were attributed to both the tectonics and climate, because the arid environment can facilitate the destruction of the weathering profile and the exportation of coarse sediments during catastrophic rainfalls (Jolivet et al., 2017) and the uplift of the Tianshan and Bogda Orogenic Belts (Tang et al., 2014; Fang et al., 2019) can facilitate the denudation of the source area. During the deposition of conglomerates, tectonic activities were frequent which are recorded by several layers of seismites in the Kalazha Formations (Figure 4).

The coeval alluvial conglomerates with the Kalazha Formation were reported in the Tarim Basin (Jolivet et al., 2017), Turfan Basin (Yuan et al., 2002), Fergana Basin (Jolivet et al., 2017), Yarkand-Fergana Basin (Morin et al., 2020), Beishan (Zhao et al., 2003), and Yingen Basin (Wei et al., 2006) in Central Asia, which suggests the tectonic activity in the Late Mesozoic influenced vast area in Central Asia. The Kalazha Formation was related to the transpression in the Junggar Basin and the uplift of the Tianshan and Bogda Orogenic Belts in the Late Jurassic, which was caused by dextral transpression of the boundary fractural zones in the Late Jurassic (He et al., 2008). This can be related to the far-field effect of the collision of the Lhasa Block along the southern margin of Asia (Hendrix et al., 1992; Li and Peng, 2013), the accretion of the Helmand Block to the southwestern Tianshan Orogenic Belt (Morin et al., 2018), or the rapid closure of the eastern part of the Mongol–Okhotsk Ocean (Yang et al., 2015a; Yang et al., 2015b). Because the collision of the Lhasa Block along the southern margin of Asia mainly occurred during the Early Cretaceous (Kapp et al., 2005; Yang et al., 2009), the collision of the Lhasa Block was unlikely to influence the Junggar Basin and the Mountains. Considering that the tectonic activity was stronger and earlier in the northern Junggar Basin (Cao et al., 2020) and that the Mongol–Okhotsk Ocean was located to the north of the Junggar Basin (Figure 1A), the far-field effect of the rapid closure of the eastern part of the Mongol–Okhotsk Ocean may be attributed to the uplift of the Tianshan and Bogda Orogenic Belts in the late Mesozoic.

The cause of aridification in the Late Jurassic has been debated. Hendrix (2000) assumed that the breakup of Pangea caused the cessation of the monsoon, which brought humid moisture to the Junggar Basin. Li et al. (2014) proposed that the collision of Lhasa and Qiangtang caused high relief and blocked humid moisture from the Tethys Sea. Yi et al. (2019) believed that a true polar wander with a southward displacement of ∼25° displaced the Junggar Basin to a middle-latitude arid zone from a high-latitude humid-temperate zone and triggered aridification, as evidenced by paleomagnetic data. Given that the collision of the Lhasa and Qiangtang blocks mainly occurred during the Early Cretaceous (Kapp et al., 2005; Yang et al., 2009) and Pangea was still considerably intact in the Late Jurassic (Torsvik et al., 2012), the true polar wander hypothesis is more reasonable. The climatic change from arid to semihumid can also be explained by true polar wander because the Junggar Basin moved back to the semihumid zone, as suggested by paleomagnetic data (Yi et al., 2019).

In summary, the sedimentary and source-to-sink evolution of intracontinental basins is controlled by tectonics and climate (Figure 11). The succession in the Late Mesozoic was a response to the aridification in Central Asia and the uplift in the Bogda and Tianshan Orogenic Belt in the late Jurassic. The aridification may be triggered by the true polar wander in the Late Jurassic and the uplift of Bogda and the Tianshan Orogenic Belt may be the far-field effect of the rapid closure of the eastern part of the Mongol–Okhotsk Ocean.