Jet-Induced Rainfall Seasonality and C4 Expansion over East Asia

Jet-Induced Rainfall Seasonality and C4 Expansion over East Asia | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Jet-Induced Rainfall Seasonality and C 4 Expansion over East Asia Jiawei Da, Chijun Sun, Lily Serach, Timothy Gallagher, Huayu Lu, and 8 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8290760/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract The Neogene expansion of C 4 grasslands transformed terrestrial ecosystems with marked influence on mammalian evolution, including hominins. However, the asynchronous C 4 expansion on different continents makes it difficult to identify the environmental drivers, especially for higher latitudes. Here we show that rainfall seasonality governed extratropical Plio-Pleistocene C 4 distributions in East Asia. Rainfall oxygen isotope ratios and clumped isotope soil temperatures exhibit coupled variations on the Chinese Loess Plateau (CLP) from 7 to 2.5 Ma, indicating more spring rain during warmer times when the subtropical westerly jet was further poleward, and more concentrated summer rain under cooler climates. We attribute these changes to meridional shifts of a summer rain band on orbital and longer timescales. Organic carbon isotope records reveal that the most C 4 -rich ecosystems tracked this summer rain band, eventually eclipsing the southern CLP margin during late Pleistocene cooling. Our model refines the East Asian paleomonsoon concept and explains the equatorward migration of extratropical C 4 ecosystems, highlighting the tight coupling between regional rainfall seasonality and vegetation. Earth and environmental sciences/Climate sciences/Palaeoclimate Earth and environmental sciences/Hydrology Earth and environmental sciences/Ecology/Palaeoecology Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Growing evidence suggests that the late Neogene C 4 expansion occurred asynchronously across different continents (1), from as early as ~ 21 million years ago (Ma) in eastern Africa (2) and 8 − 6 Ma on the India subcontinent (3) to 3.5 Ma in Australia (4). Although decreasing atmospheric CO 2 concentrations may have set the proverbial stage (5), this demonstrable global asynchroneity of C 4 expansion requires regional drivers or thresholds, which are particularly poorly understood in mid-latitude regions such as East Asia (Fig. 1). East Asian records based mainly on the δ 13 C of tooth enamel and soil carbonates (δ 13 C c ) from the Chinese Loess Plateau (CLP) and surrounding regions, suggest C 4 expansion during the late Miocene (~ 6 Ma) (13). However, neither of these proxies provide quantitative estimates of regional C 4 abundance. Specifically, tooth enamel may more closely record presence/absence rather than C 4 abundance in an ecosystem (14), and in arid-to-semi-arid ecosystems such as the CLP, soil carbonates record a mixed signal of biomass-δ 13 C, atmospheric CO 2 , and productivity (15). In contrast, soil organic matter – a more direct indicator of biomass δ 13 C – suggests that C 4 expansion across East Asia was spatiotemporally heterogeneous (Fig. 2d-f). Consequently, multiple drivers of C 4 expansion have been proposed, including decreasing atmospheric CO 2 levels (7), increased warm season precipitation (13), intensified fire activity (10), and enhanced long-term aridity (8). Importantly, studies on the CLP reveal hominin occupation as early as 2.1 Ma and a close coupling between hominin migration and regional environmental conditions (18, 19). Understanding the spatiotemporal shifting of C₄-rich biomes in East Asia is thus critical for not only elucidating vegetation–climate feedbacks but also for contextualizing the evolution of hominins, which preferred C 3 woody plants for shade, shelter, and nutrition (20). The regional climate over East Asia is regulated by the East Asian monsoon system (EAM). The EAM is unique owing to the dynamic interaction between the low-level, southerly monsoonal flow and the upper-level westerly jet impinging on the Tibetan Plateau (21) (Fig. 1a). Unlike classic tropical monsoons featuring one prolonged rainfall stage, the EAM is characterized by several quasi-stationary rainfall stages, resulting from the seasonal poleward shift of the westerly jet during boreal summer (Fig. 1d) (22). Mounting evidence from paleoclimate records and climate models underscores the role of this jet transition in controlling the meridional migration of the monsoonal rain band, and consequently rainfall seasonality over East Asia on timescales from millennia to glacial cycles (12, 21). However, the role of the jet transition in rainfall seasonality over longer timescales remains unknown. We hypothesized that the asynchronous C 4 expansion and retreat in East Asia was paced by shifting warm season precipitation in response to the long-term equatorward migration of the westerly jet during Plio-Pleistocene global cooling. To test this hypothesis, we reconstructed δ 18 O values (δ 18 O p ) of Mio-Pliocene rainfall over the CLP as an independent test of long-term, jet-transition-driven changes in East Asian rainfall seasonality. We also evaluated, by considering paleoclimate simulations, whether the concept of jet-transition-modulated changes in rainfall seasonality is plausible and, by comparison with previously published paleoenvironmental records, whether our hypothesized mechanism is broadly explanatory. Coupled T Δ47 - δ 18 O p time series We reconstructed rainfall δ 18 O p using the clumped isotope (Δ 47 ) and triple oxygen isotope compositions of calcite nodules collected from three Red Clay sections (Lantian, Shilou, and Jiaxian) on the CLP (Fig. 1b, see Methods ). The clumped isotope temperature (T Δ47 ) and δ 18 O sw from these three sections on the CLP are coherent across the late Miocene - Pliocene (Fig. 3c-d). The most prominent change occurred across the MPB (5.6-5.0 Ma), when T Δ47 and δ 18 O sw increased from 12°C (SE = 1°C, n = 13) and − 9.6‰ (SE = 0.05‰, n = 105) on average during the late Miocene (7-5.3 Ma), to as high as 31 ± 4°C and − 6.5 ± 0.5‰ at 5.1 Ma. Similar changes with smaller magnitudes are observed at 6.2 Ma and 3.5 Ma. The Δ′ 17 O sw values show high fluctuations ranging between − 95 ± 32 per meg at 5.1 Ma and 27 ± 20 per meg at 3.1 Ma (Fig. 3e), with a mean value of -28 per meg (SD = 24 per meg, n = 22). The Δ′ 17 O sw are significantly lower than those of modern meteoric waters across China (25) but generally overlap with those of Holocene soil waters calculated using a global Δ 47 - Δ′ 17 O dataset of pedogenic carbonates formed in various environments (24) (Fig. 3h). However, there is no coherent trend in Δ′ 17 O sw through time (Fig. 3e) or in Δ′ 17 O sw - δ 18 O sw space (Fig. 3h), indicating that the observed δ 18 O sw variability is not primarily driven by evaporation. To quantitatively assess the effects of evaporation and δ 18 O p on δ 18 O sw , we applied a steady-state soil-water-isotope model to our T Δ47 -δ 18 O sw -Δ′ 17 O sw dataset and obtained posterior δ 18 O p estimates using Bayesian inversion (see Methods ). The modeled δ 18 O p values correlate positively with δ 18 O sw (Fig. 3i), indicating that δ 18 O sw variations are largely explained by changes in δ 18 O p , with higher δ 18 O p under warm climate. Westerly Jet-induced rainfall seasonality In the EAM region, δ 18 O p records are typically interpreted to record rainout, with higher δ 18 O p corresponding to a weaker monsoon circulation and a drier climate associated with less rainfall upstream and/or less summer rainfall locally (26, 27). We propose that instead of annual rainfall amount or the overall EAM strength, our observed δ 18 O p variations record changes in rainfall seasonality closely linked to the westerly jet stream dynamics. Under warmer climates, such as the beginning of the Pliocene, the reduced MTG (Fig. 4h) resulted in the weakening and northward migration of the westerly jet (28), leading to a seasonally earlier onset (spring rather than summer) of rainfall over northern China (21) (Fig. 4j). Moreover, the weakened jet impinging upstream on the Tibetan Plateau caused a weaker stationary Rossby wave downstream, resulting in high pressure anomalies over East Asia and the northward shift of the North Pacific subtropical high (NPSH) (29). Consequently, the southerly wind along the northwestern edge of the NPSH migrated northward, resulting in an earlier onset of spring rainfall. The modern monthly δ 18 O p pattern across the EAM region is characterized by the transition from the higher-δ 18 O spring to lower-δ 18 O summer rain, primarily controlled by distillation upstream (27) (Fig. 1c). Because the seasonal march of the monsoonal rain was already established during the Miocene (30), it is reasonable to assume a quasi-stationary monthly δ 18 O p climatology through time. The modern spring-to-summer δ 18 O p shift exceeds 5‰ (Fig. 1c), comparable to changes in the reconstructed δ 18 O p (Fig. 2g). Thus, the T Δ47 - δ 18 O p coupling supports rainfall seasonality changes as the main driver of the observed δ 18 O sw variability, with more spring rain (high δ 18 O p ) during warmer times when the westerly jet was positioned at a higher latitude. Several lines of evidence support the MTG – jet – rainfall seasonality as the drivers of coupled T Δ47 - δ 18 O p variations. First, our T Δ47 records show large amplitude of variations as high as > 10°C increase across the MPB (Fig. 3c). For comparison, mean annual surface temperature variability on the CLP across the Pleistocene glacial cycles is less than 10°C (31). Given that regional climate variability is much less pronounced during the late Miocene-Pliocene periods (32), it is unlikely that changes in T Δ47 are solely caused by mean annual temperature change. Pedogenic carbonates tend to record seasonal climate signals rather than annual means (33). In arid regions such as the CLP, pedogenic carbonate formation is primarily driven by rainfall events through its control over soil water chemistry (34). Consequently, pedogenic carbonates grow after the rainy season under enhanced evaporation and reduced soil respiration (35). During cold periods, rainfall occurred within summer due to the jet at a lower latitude. Consequently, pedogenic carbonates grew during late summer and early fall, recording low summer-δ 18 O p and fall temperature. During warmer periods, an early onset of spring rain and intensive summer evaporation likely facilitate pedogenic carbonate growth during mid-summer, recording high spring-δ 18 O p and summer temperature. Modern surface temperature difference on the CLP between July and October can be as high as 17°C. Thus, the large T Δ47 variability is likely caused by seasonal shifts in soil carbonate formation superimposed on the long-term changes in annual mean temperature. Moreover, minima in meridional sea surface temperature (SST) gradients (17) at 5 and 3.5 Ma correspond well with maxima in T Δ47 and δ 18 O sw (Fig. 4f-h). The effect of the MTG on jet position and strength is also documented by dust flux and provenance records from the North Pacific Ocean (28, 36), suggesting poleward and weakened westerly winds during the warm Pliocene compared with the Pleistocene glacial periods. The jet transition hypothesis is also supported by model simulations. Specifically, we looked at three experiments, mid-Pliocene 280 ppm (Eoi280) and 400 ppm (Eoi400) and pre-industrial control (E280), from the Community Earth System Model version 2 (CESM 2), following the Pliocene Model Intercomparison Project phase 2 (PMIP2) (37, 38) (See Methods). Both Eoi280 and Eoi400 show a weakening and northward shift of the westerly jet compared to E280 (Fig. S6a-b), which led to a northward shift of the western Pacific subtropical high and an earlier seasonal march of the EAM (Fig. S6c-d). These shifts simulated using CESM 2 are consistent with the PMIP2 multi-model ensemble mean (39). Consequently, northern China (> 30°N) received more spring rain and less summer rain during the warm Pliocene compared to the pre-industrial conditions (Fig. S6e-f). Finally, rainfall seasonality reconciles the discrepancy among various proxy records. For instance, soil magnetic susceptibility is mainly controlled by the abundance of ultrafine ferrimagnetic particles (40) formed during redox potential changes caused by soil wet-dry cycles (41). The wet-dry cycles, and presumably redox potential, are most prominent during the summer when potential evapotranspiration is large and soil respiration is water-limited (42). In this view, the increases in magnetic susceptibility toward the Pliocene-Pleistocene boundary (43) (Fig. S4b) were caused by larger or more frequent wet-dry cycles as monsoon rainfall seasonality shifted from spring to summer. This seasonality shift also explains the terrestrial snail record (Fig. S4e), which shows a dominance of warm-humid-adapted species during the early Pliocene. Because land snails tend to be active at high relative humidity (> 70%) and intermediate temperature (10–27°C) (44), spring rain during the early Pliocene resulted in greater water availability due to lower evapotranspiration, which favored the growth of humid-loving taxa. On the other hand, soil bacteria likely reached peak metabolism during the summer growing season (45, 46), and thus the soil tetraether proxy likely record summer drought during the warm early Pliocene (Fig. S4c). During the Pliocene, rainfall maxima over East Asia occurred at obliquity minima as recorded by magnetic susceptibility from the CLP (a record of wet-dry cycles) and marine proxies related to runoff (e.g., salinity, productivity, lithogenic accumulation) from the South China Sea (47). This relationship is opposite of that expected from the effect of fast physics on monsoonal rainfall. For example, at obliquity minima, northern hemisphere summer insolation is small, which should result in smaller land-sea and cross-equatorial pressure gradients and thus weaker monsoonal circulation. Therefore, in the established framework of the EAM, this observed relationship has been enigmatic (47). However, if we view the EAM from the perspective of the jet transition, there is a dynamic explanation for rainfall maxima at obliquity minima. During the Pliocene, the mean position of the westerly jet was further north at obliquity maxima, which pushed summer rainfall maxima to a higher latitude (north of Jiaxian according to our records), and consequently, resulted in fewer summer wet-dry cycles on the CLP and less surface runoff across southern China. The zone of intense summer rainfall moved southward at obliquity minima, following the mean position of the westerly jet. Westerly Jet-induced C 4 migration The meridional migration of the westerly jet provides a dynamic explanation for the asynchronous C 4 expansion in East Asia. Specifically, the summer jet position likely maintained at high latitude across the late Miocene – Pliocene, resulting in rainier spring in East Asia and consequently, the dominance of C 3 plants as indicated by the overall low organic carbon δ 13 C values (Fig. 2d-f). This is supported by model results, showing the summer jet located at > 40°N under Pliocene boundary conditions (Fig. S6). The long-term increase in MTG gradually pushed the mean jet position equatorward, leading to asynchronous C 4 expansion as suggested by records, from 5.5-4 Ma in the U1430 from the Japan Sea, 3.5–2.5 Ma in the G3 on the North China Plain, to 2 − 1 Ma in the southernmost Lingtai section (Fig. 2d-f). Specifically, because sediments at U1430 were mainly of eolian origin brought by the spring storm outbreaks and the westerlies (8) (Fig. 1a), they should largely originate from regions that experienced dry spring and thus we should expect the source region to track the summer rain band (Fig. 2h). The shifting regimes of sediment source explains (1) increased black carbon δ 13 C at U1430 during 5 − 4 Ma despite the southward migration of the jet and (2) the persistently high δ 13 C from 4.2–1.7 Ma (Fig. 2d). Notably, the mean summer jet position likely overlapped with the dust source regions of U1430 beginning in the late Miocene (Fig. 2h), yet δ 13 C did not increase until the MPB (Fig. 2d). We attribute this discrepancy to the significant surface cooling in East Asia during 8 − 7 Ma, which likely inhibited the growth of C 4 plants (48). The onset and intensification of the NHG during 3.6–2.5 Ma pushed the mean jet position further south (28), which intensified summer rainfall in northern China (23, 32) (Fig. 3f-g) and led to major C 4 expansion in the source regions of G3 sediments (Fig. 2e). With further cooling and high-latitude ice sheet expansion during the Pleistocene, the summer jet and associated rain band were pushed even further south across the CLP resulting in migration of C 4 vegetation from the G3 source area equatorward toward Lingtai (Fig. 2e-f). The Pleistocene C 4 expansion is also witnessed in organic δ 13 C records from other sections located on the southern CLP (49, 50). The long term equatorward C 4 migration is inconsistent with a direct control by decreasing atmospheric CO 2 . The late Pleistocene C 4 retreat from Lingtai can be explained by the slightly poleward shift of the jet since 1.1 Ma, as indicated by the decreased MTG and benthic δ 18 O (Fig. 2b-c). The jet transition also explains some of the small δ 13 C shifts across the Miocene-Pliocene. During the late Miocene cooling, as the jet shifted equatorward gradually, organic carbon δ 13 C increased by ~ 2‰ first in the north at G3, then at Jiaxian, and latest at Lingtai, possibly resulting from progressive, small-scale, equatorward C 4 expansion (Fig. 4c-e). As the MTG decreased significantly with warming across the MPB (Fig. 4h), the jet and consequently summer rainfall maxima migrated poleward, leading to rainier spring and C 4 retreat in northern China as evidenced by δ 13 C decreases at both Lingtai and Jiaxian (Fig. 4e-f). The jet shifted equatorward again during 5 − 4 Ma, resulting in C 4 expansion in G3 source regions and Jiaxian, whereas Lingtai was still dominated by C 3 plants (Fig. 4c-e). A consistent spatiotemporal pattern emerges in which warm season precipitation was the main limiting factor governing the multi-stage southward C 4 progression across East Asia, paced by the latitudinal migration of the westerly jet. Obliquity-paced changes in the meridional insolation gradient Although the late Pliocene-Pleistocene evolution of the MTG, the westerly jet, and C 4 biomass can be explained by the NHG (51), their variations prior to the NHG require an additional mechanism. Orbital insolation directly impacts the MTG and consequently monsoon rainfall stages (52). The 41-ky obliquity cycle influences meridional insolation gradients by affecting high latitude climate as well as global ice volume (53) (Fig. 3a-b). Importantly, the 1.2-Myr cycle that corresponds to the amplitude modulation of obliquity (Fig. 3a) is thought to play a critical role in glaciation and sea-level changes (54). The MTG based on the compiled sea surface temperature records, is generally in pace with the 1.2-Myr cycle between 7 and 3 Ma (Fig. 4h). Our T Δ47 and δ 18 O sw records show clear 1.2-Myr cycle, with three δ 18 O sw peaks co-occurring roughly with maximum amplitudes of the 41-kyr obliquity cycle (Fig. 3c-d). During obliquity maxima at peak amplitudes, enhanced high-latitude summer insolation led to reduced meridional insolation gradient and MTG (Fig. 4h), thereby weakening the jet and pushing it further north, leading to an earlier onset of spring rain in northern China. This 1.2-Myr rhythm was also registered in other proxy-derived EAM records (55) (Fig. 3f-g), supporting its critical role in modulating the regional water cycle. It should be noted that the 1.2-Myr cycle controls the amplitude of the obliquity cycle (Fig. 3a), meaning both highest obliquity maxima and lowest minima occur at peak amplitudes. Our T Δ47 and δ 18 O sw records preserve only obliquity maxima signals due to overprinting. Specifically, during periods other than obliquity maxima, the jet was at a lower latitude and monsoonal rain was more concentrated during summer, when temperature and evaporation are relatively higher than spring, leading to less soil water penetration and shallower formation of pedogenic carbonates. Whereas during obliquity maxima, early onset of monsoonal rain, cooler temperature and less evaporation during spring allowed soil water to percolate deeper into soils, leaching previously formed shallower carbonates and resetting oxygen isotope compositions. In the northern Shilou and Jiaxian sections, the mean sedimentation rate was ~ 10 m/Ma (43), corresponding to 40 cm of sediment per 41-kyr obliquity cycle. The modern mean annual precipitation (MAP) across the CLP ranges between 250 and 650 ppm, which corresponds to carbonate accumulation depth of ~ 40–70 cm (56). Assuming MAP during the late Miocene-Pliocene was within this range, pedogenic carbonates formed during the previous obliquity maxima should be at ~ 80–110 cm depth at the next maxima, which would require extremely high MAP of more than ~ 700–1000 mm to dissolve them. Thus, these deepest-formed carbonates were selectively preserved, recording seasonal environmental conditions during the obliquity maxima. The MTG, δ 18 O sw , and the 1.2-Myr cycle fell out of phase at 4.1 Ma, when the obliquity amplitude minima preceded the MTG maxima and δ 18 O sw minima, and the following obliquity amplitude maxima led the subsequent MTG minima and δ 18 O sw maxima (Fig. 4f and 4h). These offsets likely reflect the increasing influence of Northern Hemisphere cooling on the MTG, the jet position, and rainfall seasonality during the Plio-Pleistocene transition. The prevalence of ice-sheet-related slow physics over insolation-driven fast physics is supported by damped orbital-scale variability of both the East Asian winter and summer monsoon as well as their out-of-phase relationship starting 4.2 Ma (32). Taken together, our nodule-based isotope records spanning 7-2.5 Ma mark the first empirical evidence supporting the jet transition hypothesis across orbital-to-tectonic timescales. Consequently, the jet transition offers a unified explanation of the multi-phase C 4 evolution across East Asia by modulating rainfall seasonality. This complex history of C 4 migration reflects the dynamic interplay between global climate forcings (i.e., ice sheets, insolation, jet dynamics) and regional factors (i.e., EAM, topography). Notably, paleolithic evidence suggests that hominins occupied the southern CLP since 2.12 Ma (18), favoring C 3 -dominated environments with abundant tree cover (19). The decrease in the number of artefacts found near Lantian after 1 Ma corresponds to major C 4 expansion (19), highlighting the central role of rainfall seasonality in hominin migration at higher latitudes through its influence on regional C 3 /C 4 abundance. Our findings provide a critical paleoclimatic framework for linking vegetation change, atmospheric circulation, and early hominin adaptation in East Asia. Materials and Methods Sampling The CLP currently sits on the edge of the EAM region and is sensitive to monsoonal rainfall changes ( 57 ) (Fig. 1 a). The Miocene-Pliocene deposits on the CLP, known collectively as the Red Clay formation, record continuous paleoenvironmental information from 8 to 2.5 million years ago (Ma) ( 58 ). Bulk paleosol and pedogenic carbonate nodule samples were collected from three Red Clay sections: Lantian, Shilou, and Jiaxian. The 66-m thick Lantian section (34.190°N, 109.234°E, elevation = 630 m) was located on the southern edge of the CLP, whereas the 58-m Jiaxian (38.272°N, 110.090°E, elevation = 1025 m) and 71.4-m thick Shilou sections (36.926°N, 110.928 °E, elevation = 1160 m) were situated in the northern CLP (Fig. 1 ). Modern mean annual precipitation (MAP) decreases from 720 mm in the southern Lantian to 529 mm and 433 mm in the northern Shilou and Jiaxian. Mean annual air temperatures (MAT) are 10.9°C, 10.0°C, and 14.6°C in Jiaxian, Shilou, and Lantian, respectively. Bulk paleosol samples were collected at 10 cm intervals, and at least two nodule samples were collected from each distinct depth. To avoid regolith contamination, soil profiles were trenched 1 m deep before sampling. Age model The magnetic susceptibility data from each section were used to correlate to published age models ( 59 – 61 ), which were established through linear interpolation using geomagnetic reversals as age control. Because the Red Clay formation is an aggregated soil profile, it is not possible to constrain the depth of carbonate nodule formation. We therefore use the depositional ages of the soil layer where the nodules were formed to represent the ages of nodules. Based on the modern MAP across the CLP and the MAP-Bk depth relationship ( 62 ), the depth of pedogenic carbonate formation is ~ 40–70 cm, which corresponds to ~ 20–50 ka offset given the sedimentation rate of the three Red Clay profiles. Sample pretreatment We conducted stable isotope analyses on calcite nodules (n = 970) collected from the three Red Clay sections at Lantian, Shilou, and Jiaxian. At least two individual nodule samples from each stratigraphic layer were used for δ 18 O analyses. From those samples, a total of 32 and 22 nodule samples were selected for clumped isotope and triple oxygen isotope analyses, respectively. Prior to analyses, calcite nodule samples were either ultrasonically rinsed and then oven-dried at 50°C overnight or physically cleaned using a Dremel® tool to remove matrix sediments. Samples were then inspected for spars and, if present, only micrites were drilled to collect powder for isotopic analyses. Stable oxygen isotope analysis Sample splits of approximately 200 µg were loaded into exetainers and He-flushed before analysis. δ 18 O analyses were carried out separately in two different labs. For Shilou and Jiaxian samples, sample powders were acidified with 103% phosphoric acid for 2–12 hours at 50°C. The stable isotopic compositions of released CO 2 were measured using a Thermo Gasbench II coupled to a continuous flow Thermo 253 isotope ratio mass spectrometer (IRMS) at the University of Texas at Austin. The isotopic values are reported in per mil (‰) notation relative to Vienna Pee Dee Belemnite standard (VPDB) and were normalized to that scale using an in-house laboratory standard (UT Marble), NBS-18, and NBS-19. The external reproducibility of replicate analyses of standards averaged ± 0.07‰ for δ 18 O (1σ). For Lantian samples, the measurements were performed on a Finnigan DELTAplusXP IRMS attached to a Thermo GasBench II (75°C reaction with 100% phosphoric acid) at Nanjing University. The isotopic values were adjusted to VPDB scale using NBS-18, NBS-19, and an in-house laboratory standard (TTB-1). The standard error (1σ) of replicate analyses of the standard is ± 0.04‰. Clumped isotope analysis Clumped isotope measurements were carried out at the University of Washington Isolab, using two different instruments. Sampled materials are not subject to solid-state reordering of clumped isotope signals due to the shallow burial depth (< 100 m). The Jiaxian samples (n = 18) were measured on a Thermo MAT 253 IRMS configured to measure m/z 44–49 inclusive, following method described in Burgener et al. (2016). For each replicate analysis, sample aliquots containing 6–8 mg of CaCO 3 were reacted in a common phosphoric acid bath at 90°C for 10 minutes. The CO 2 evolved from the acid digestion was purified cryogenically on an automated stainless steel/nickel vacuum line using a liquid N 2 trap, an ethanol-dry ice mixture (~ -80°C), and a Porapak Q trap (50/80 mesh, 122 cm long, 6.35 cm OD) at -20°C. The CO 2 was transferred through the Porapak using He carrier gas, isolated cryogenically, and flame-sealed into a Pyrex break seal before analysis. CO 2 was introduced into the IRMS using an automated 10-port tube cracker inlet system. Isotope ratios were converted to the VPDB scale using intra and interlaboratory standards calibrated against international standards NBS-18 and NBS-19. These include in-house calcites with dissimilar bulk composition (C64 and C2, which are both reagent grade; and a Porites coral sample) and ETH standards (ETH 1–4). For each four carbonate sample unknowns, a carbonate standard was digested, purified, and analyzed using the same procedure. All sample unknowns were analyzed in triplicate at minimum. The Lantian and Shilou samples (n = 14) were analyzed using a fully automated Nu Instruments NuCarb carbonate device coupled to a Nu Perspective IRMS. Sample powders containing approximately 1.2–1.5 mg calcite were loaded in individual glass vials before reaction with phosphoric acid at 70°C, purification using cryogenic and Porapak traps, and analysis. For each of the five carbonate sample unknowns, a carbonate standard was measured, including eight intra and interlaboratory standards (IAEA-C1, IAEA-C2, MERCK, Coral, ETH 1–4). All sample unknowns were analyzed in triplicate at minimum. Δ 47 values were calculated using established methods ( 63 ), with the exception that we used updated 17 O correction parameters ( 64 ), following recommendations of Daëron et al. (2016) ( 65 ) and Schauer et al. (2016) ( 66 ). Δ 47 values analyzed using the Thermo MAT 253 IRMS were corrected to CDES-90°C acid values using CO 2 equilibrated at 4, 60 and 1000°C; Δ 47 values analyzed using the Nu Perspective were corrected to I-CDES ( 67 ) 90°C acid values using ETH1-4, IAEA-C1, IAEA-C2, and MERCK standards ( 68 ). The I-CDES and CDES scales are mathematically identical and interchangeable ( 68 ). Δ 47 uncertainties for individual samples were given as one standard error (1σ) of replicate analyses, calculated using the long-term standard deviation of carbonate standard analyses, or sample replicates, whichever is larger. Clumped isotope temperature (T Δ47 ) values reported here were calculated from mean Δ 47 values using the Andersen et al. (2021) calibration for samples digested at 90°C (no acid fractionation correction). Triple oxygen isotope analysis Triple oxygen isotope measurements were carried out at the University of New Mexico Center for Stable Isotopes using two different instruments. The Jiaxian samples (n = 8) were measured on a Thermo Fisher MAT 253 + isotope ratio mass spectrometer configured for triple oxygen isotope analyses. Sample powders containing at least 300 µg CaCO 3 were first converted to CO 2 using the phosphoric acid digestion system (25°C reaction with 100% phosphoric acid for 16 h). The resulting gas was purified using a cryogenic trap to remove water and any non-condensable gases. CO 2 was then fluorinated to produce O 2 to be measured for triple oxygen isotopes using the nickel bomb fluorination method modified by Wostbrock et al. (2020) ( 69 ). No replicate analysis was carried out. Lantian and Shilou samples (n = 14) were measured using a TILDAS (Tunable Infrared Laser Direct Absorption Spectroscopy) CO 2 isotope monitor for Δ '17 O - CO 2 designed and manufactured by Aerodyne Research, Inc. (ARI, Billerica, MA, USA), configured for triple oxygen isotope analyses of CO 2 . Detailed descriptions of the TILAS and measurement procedure are provided in Perdue et al. (2022) ( 70 ). Before the analyses, sample aliquots containing 0.5 mg CaCO 3 were loaded in exetainer vials sealed with a rubber septum. The vials were placed in a 70°C heating block and He-flushed with a flow rate of 11 m/s for 2 h. 5 ml of phosphoric acid was then injected into the vial and reacted with the sample powder for 8 h. The evolved CO 2 was cryogenically separated from water and purified before being introduced into the TILDAS analyzer. For analysis in the TILDAS system, CO 2 needs to be thoroughly mixed with CO 2 -free dry air in a two-bellow system, which produces a mixture of 456 ppm CO 2 . The isotope values of each sample measurement were determined relative to the reference gas by interpolating the prior and subsequent reference gas measurements in time. The total time required for 12 reference-sample cycles was 30 min. To monitor long-term analytical error and to place carbonate Δ '17 O values on the VPDB scale, an interlaboratory standard (NBS-19, IAEA 603) was measured after four sample unknowns. The external reproducibility of replicate analyses of standards averaged 5 per meg for Δ '17 O (1σ). All sample unknowns were analyzed in triplicate at minimum. Calculating the δ 18 O and Δ '17 O of soil waters With T Δ47 , δ 18 O c , and Δ ' 17 O c available, δ 18 O sw and Δ ' 17 O sw can be calculated using the temperature-dependent fractionation between calcite and water ( 71 , 72 ). Specifically, we first calculated δ 17 O c using δ 18 O c (‰, VSMOW) and Δ ' 17 O c (per meg, VSMOW) ( 73 ): δ ' 18 O c = 1000 × ln(δ 18 O c / 1000 + 1) δ ' 17 O c = Δ ' 17 O c / 1000 + 0.528 × δ ' 18 O c δ 17 O c = (exp(δ ' 17 O c / 1000) – 1) × 1000 Next, we calculated the temperature-dependent equilibrium isotope fractionation factor ( 18/16 𝛼 calcite−water , 17/16 𝛼 calcite−water ) between calcite and soil water ( 69 , 71 ) using the clumped isotope temperature: 18/16 𝛼 calcite−water = exp((16.1 × 1000 / (273.15 + T Δ47 ) – 24.6) / 1000) 17/16 𝛼 calcite−water = ( 18/16 𝛼 calcite−water ) θ θ = -1.39 / (273.15 + T Δ47 ) + 0.5305 The δ 18 O sw (‰, VSMOW) and Δ ' 17 O sw (per meg, VSMOW) can then be calculated using the following equations: δ 1x O sw = (δ 17 O c + 1000) / 1x/16 𝛼 calcite−water – 1000 δ ' 1x O sw = 1000 × ln(δ 1x O c / 1000 + 1) Δ ' 17 O sw = 1000 × (δ ' 17 O sw − 0.528 × δ ' 18 O sw ) Where the superscript 1x refers to either 18 or 17. Reconstructing rainfall δ 18 O The δ 18 O of pedogenic carbonates (δ 18 O c ) record soil-water δ 18 O (δ 18 O sw ), with the formation temperature constrained by clumped isotopes (T Δ47 , see Methods ). In well-drained soils, δ 18 O sw inherits the δ 18 O of local rainfall (δ 18 O p ), which is controlled by the regional water cycle and large-scale air circulation ( 27 ). However, δ 18 O sw values are also affected by soil processes such as evaporation and mixing ( 74 ). The triple oxygen isotope ( 16 O- 17 O- 18 O) compositions of pedogenic carbonates can help quantify the degree of evaporation experienced by soil waters from which the carbonates precipitated, thereby improving estimates of δ 18 O p ( 24 ). Δ′ 17 O represents the deviation in δ′ 17 O-δ′ 18 O space (δ′=1000×ln(δ/1000 + 1)) from the global meteoric water line (GMWL, Δ′ 17 O = δ′ 17 O-λ × δ′ 18 O, λ = 0.528 ( 73 )). Unlike δ 18 O p , Δ′ 17 O p is not sensitive to processes involving equilibrium fractionation such as condensation and rainout from air masses that control the GMWL ( 73 ). Consequently, Δ′ 17 O sw values are primarily governed by evaporation-induced kinetic fractionation, which has a trajectory in δ′ 17 O-δ′ 18 O space shallower than the GMWL slope ( 24 ). Measurements of modern and ancient terrestrial waters support the use of Δ′ 17 O as a proxy of evaporation and thus regional aridity, with higher evaporation corresponding to lower Δ′ 17 O sw and δ 18 O sw ( 24 , 75 ). Here we reconstructed δ 18 O sw and Δ′ 17 O sw using δ 18 O, T Δ47 , and Δ′ 17 O of calcite nodules (Fig. S2, see Methods ). By leveraging Δ′ 17 O sw , we accounted for the effect of evaporation on δ 18 O sw and reconstructed δ 18 O p . Bayesian inversion All the paleoclimate proxies are affected by multiple environmental parameters ( 76 ). To quantitatively constrain environmental variables based on proxy data, here we applied Bayesian inversion to the steady-state soil water isotope model ( 24 ), in which Markov Chain Monte Carlo methods were used to obtain posteriors of environmental parameters conditioned on the proxy data (δ 18 O c , Δ′ 17 O c , Δ 47 ) ( 77 ). The analyses were performed in R version 4.4.2 ( 78 ) using the rjags ( 79 ) and R2jags package ( 80 ). The prior ranges of all the related environmental inputs, including RH, δ 18 O p , carbonate formation depth, and evaporation rate, were set as uniform distributions of 10% − 90%, -25‰ - -5‰, 0–100 cm, and 10 − 10 – 10 − 9 m/s, respectively, based on either theoretical constraints or modern observations. We ran five parallel chains, each consisting of 5×10 5 iterations, with a burn-in of 1×10 4 . Convergence was assessed by the Gelman and Rubin convergence factor (Rhat) and effective sample sizes reported by rjags ( 81 ). All the posteriors show strong convergence (Rhat 3000, Fig. S5). CESM2 experiments The climate model simulations analyzed in this study are run with the Community Earth System Model version 2 (CESM2) ( 82 ). CESM2 incorporates Community Atmospheric Model version 6 (CAM6), Community Land Model version 5 (CLM5), Community Ice Code version 5 (CICE5), and Parallel Ocean Program version 2. The atmosphere and land components are run at a 0.9° x 1.25° horizontal resolution, while the ocean and sea ice components are run with a nominal 1° resolution. The pre-industrial simulation included in this study is part of the Climate Model Intercomparison Project phase 6 (CMIP6) Tier 1 experiment ( piControl.001 ) ( 82 ). The mid-Pliocene simulations ( midPliocene-eoi400 ) analyzed in this study have been reported in Feng et al. (2020) ( 38 ), targeted at the mid-Piacenzian Warm Period (MPWP, 3.205 Ma). The MPWP boundary conditions follow the protocol of the second phase of the PlioMIP (PlioMIP2) ( 37 ), including present-day orbital configuration, continental greening, a reduced Greenland Ice Sheet, a deglaciated western Antarctica, and adjusted topography and coastlines ( 38 , 83 ). Notably, the Bering and Canadian Arctic Archipelago straits are closed, and the Sunda, Sahul, Baltic Shelves, and Hudson Bay are exposed. In addition, to assess the climate changes induced by changes in CO 2 during the Pliocene, we report a new fully coupled mid-Pliocene simulation where the same MPWP protocol is used except that the CO 2 concentration is reduced to the pre-industrial level at 280 ppm ( midPliocene-eoi280). Recalculating SST records To further investigate the evolution of the MTG, we collated high-quality \(\:{\text{U}}_{\text{37}}^{\text{k'}}\) -based SST records around the global ocean ( 17 ). These SST records were divided into three groups based on their latitudes: high-latitude > 50°N regions (ODP 883/884, 907, 982), mid-latitude regions (20–50°N: ODP 1010, 1021, and 1208), and tropics (< 20°N: ODP 722, 846, 850, U1338, 1241) (Fig. S1 ). To account for the seasonality of alkenones and nonlinear response of \(\:{\text{U}}_{\text{37}}^{\text{k'}}\) to temperature, we recalculated the \(\:{\text{U}}_{\text{37}}^{\text{k'}}\) -based SSTs by applying the original \(\:{\text{U}}_{\text{37}}^{\text{k'}}\) data to a Bayesian B-spline regression model, BAYSPLINE ( 84 ). The integrated SSTs of each group were computed by binning the SSTs from each OSP site to 100-kyr windows and averaging the estimates. The SST differences between the tropics and mid-latitude regions and the tropics and high-latitude regions were used to infer the MTG. Declarations Acknowledgments The authors acknowledge the aid of Andrew Schauer and Jeff Cullen in supervising clumped isotope and stable isotope analyses. This work was supported by the United States National Science Foundation (1545859 to D.O.B. and 1156134 to K.H.) and the National Natural Science Foundation of China (42030503 to J.F. and 42021001 to H.L.). R.F. would like to acknowledge NSF-2103055 and NSF-2238875 for supporting the production of simulations used in this study. References Steinthorsdottir M et al (2021) The Miocene: The Future of the Past. Paleoceanography and Paleoclimatology 36, e2020PA004037 Peppe DJ et al (2023) Oldest evidence of abundant C4 grasses and habitat heterogeneity in eastern Africa. Science 380:173–177 Quade J, Cerling TE, Bowman JR (1989) Development of Asian monsoon revealed by marked ecological shift during the latest Miocene in northern Pakistan. Nature 342:163 Andrae JW et al (2018) Initial expansion of C4 vegetation in Australia during the late Pliocene. Geophys Res Lett 45:4831–4840 THE CENOZOIC CO2 PROXY INTEGRATION PROJECT (CENCO2PIP) CONSORTIUM (2023) Toward a Cenozoic history of atmospheric CO2. Science 382:eadi5177 Luo X et al (2024) Mapping the global distribution of C4 vegetation using observations and optimality theory. Nat Commun 15:1219 Lu J et al (2019) The Early Pliocene global expansion of C4 grasslands: A new organic carbon-isotopic dataset from the north China plain. Palaeogeogr Palaeoclimatol Palaeoecol 109454. https://doi.org/10.1016/j.palaeo.2019.109454 Shen X et al (2018) Increased seasonality and aridity drove the C4 plant expansion in Central Asia since the Miocene–Pliocene boundary. Earth Planet Sci Lett 502:74–83 Gallagher TM et al (2021) Regional patterns in Miocene-Pliocene aridity across the Chinese Loess Plateau revealed by high resolution records of paleosol carbonate and occluded organic matter. Paleoceanography and Paleoclimatology n/a. e2021PA004344 Zhou B et al (2014) Late Pliocene–Pleistocene expansion of C4 vegetation in semiarid East Asia linked to increased burning. Geology 42:1067–1070 Schiemann R, Lüthi D, Schär C (2009) Seasonality and Interannual Variability of the Westerly Jet in the Tibetan Plateau Region. https://doi.org/10.1175/2008JCLI2625.1 Zhang H et al (2018) East Asian hydroclimate modulated by the position of the westerlies during Termination I. Science 362:580–583 Passey BH et al (2009) Strengthened East Asian summer monsoons during a period of high-latitude warmth? Isotopic evidence from Mio-Pliocene fossil mammals and soil carbonates from northern China. Earth Planet Sci Lett 277:443–452 Edwards EJ, Osborne CP, Strömberg CA, Smith SA, C4 Grasses Consortium (2010) The origins of C4 grasslands: integrating evolutionary and ecosystem science. Science 328:587–591 Da J, Zhang YG, Li G, Ji J (2020) Aridity-driven decoupling of δ13C between pedogenic carbonate and soil organic matter. Geology. https://doi.org/10.1130/G47241.1 Lisiecki LE, Raymo ME (2005) A Plio-Pleistocene stack of 57 globally distributed benthic δ 18 O records. Paleoceanography 20:1–17 Herbert TD et al (2016) Late Miocene global cooling and the rise of modern ecosystems. Nat Geosci 9 Zhu Z et al (2018) Hominin occupation of the Chinese Loess Plateau since about 2.1 million years ago. Nature 559:608–612 Zhou B et al (2023) Hominin Response to Oscillations in Climate and Local Environments During the Mid-Pleistocene Climate Transition in Northern China. Geophysical Research Letters 50, e2023GL104931 Cerling TE et al (2011) Woody cover and hominin environments in the past 6 million years. Nature 476:51–56 Chiang JCH et al (2015) Role of seasonal transitions and westerly jets in East Asian paleoclimate. Q Sci Rev 108:111–129 Ding Y, Chan JCL (2005) The East Asian summer monsoon: an overview. Meteorol Atmos Phys 89:117–142 Yang S et al (2018) A strengthened East Asian Summer Monsoon during Pliocene warmth: Evidence from ‘red clay’ sediments at Pianguan, northern China. J Asian Earth Sci 155:124–133 Kelson JR et al (2023) Triple oxygen isotope compositions of globally distributed soil carbonates record widespread evaporation of soil waters. Geochim Cosmochim Acta. https://doi.org/10.1016/j.gca.2023.06.034 Tian C, Wang L, Tian F, Zhao S, Jiao W (2019) Spatial and temporal variations of tap water 17O-excess in China. Geochim Cosmochim Acta 260:1–14 Cheng H et al (2009) Ice Age Terminations Science 326:248–252 Chiang JCH, Herman MJ, Yoshimura K, Fung IY (2020) Enriched East Asian oxygen isotope of precipitation indicates reduced summer seasonality in regional climate and westerlies. Proceedings of the National Academy of Sciences 117, 14745 Abell JT, Winckler G, Anderson RF, Herbert TD (2021) Poleward and weakened westerlies during Pliocene warmth. Nature 589:70–75 Son J-H, Seo K-H, Wang B (2020) How Does the Tibetan Plateau Dynamically Affect Downstream Monsoon Precipitation? Geophysical Research Letters 47, e2020GL090543 He L et al (2024) Northward Extension of East Asian Summer Monsoon Since the Miocene Set by the Uplift of Tibetan Plateau. Geophysical Research Letters 51, e2023GL107262 Lu H et al (2022) Decoupled Land and Ocean Temperature Trends in the Early-Middle Pleistocene. Geophys Res Lett 49:e2022GL099520 Sun Y, An Z, Clemens SC, Bloemendal J, Vandenberghe J (2010) Seven million years of wind and precipitation variability on the Chinese Loess Plateau. Earth Planet Sci Lett 297:525–535 Kelson JR et al (2020) A proxy for all seasons? A synthesis of clumped isotope data from Holocene soil carbonates. Q Sci Rev 234:106259 Da J, Li GK, Ji J (2023) Seasonal changes in the formation time of pedogenic carbonates on the Chinese Loess Plateau during Quaternary glacial cycles. Q Sci Rev 305:108008 Breecker DO, Sharp ZD, McFadden LD (2009) Seasonal bias in the formation and stable isotopic composition of pedogenic carbonate in modern soils from central New Mexico, USA. Geol Soc Am Bull 121:630–640 Zhong Y et al (2022) Humidification of Central Asia and equatorward shifts of westerly winds since the late Pliocene. Commun Earth Environ 3:1–9 Haywood A et al (2016) The Pliocene model intercomparison project (PlioMIP) phase 2: scientific objectives and experimental design. Clim Past 12:663–675 Feng R, Otto-Bliesner BL, Brady EC, Rosenbloom N (2019) Increased Climate Response and Earth System Sensitivity From CCSM4 to CESM2 in Mid-Pliocene Simulations. Journal of Advances in Modeling Earth Systems 12, eMS002033 (2020) He L, Zhou T, Chen X, Zuo M, Zou L (2024) Earlier seasonal march of the East Asian summer monsoon in the mid-Pliocene. https://doi.org/10.1175/JCLI-D-23-0709.1 Wang X, Nie J, Stevens T, Zhang H, Xiao W (2022) Resolving conflicting models of late Miocene East Asian summer monsoon intensity recorded in Red Clay deposits on the Chinese Loess Plateau. Earth Sci Rev 104200. https://doi.org/10.1016/j.earscirev.2022.104200 Ahmed IAM, Maher BA (2018) Identification and paleoclimatic significance of magnetite nanoparticles in soils. Proc Natl Acad Sci U S A 115:1736–1741 Fan J, Jones SB, Qi LB, Wang QJ, Huang MB (2012) Effects of precipitation pulses on water and carbon dioxide fluxes in two semiarid ecosystems: measurement and modeling. Environ Earth Sci 67:2315–2324 Ao H et al (2016) Late Miocene–Pliocene Asian monsoon intensification linked to Antarctic ice-sheet growth. Earth Planet Sci Lett 444:75–87 Cowie RH (1984) The life-cycle and productivity of the land snail Theba pisana (Mollusca: Helicidae). J Anim Ecol 311–325 Deng L, Jia G, Jin C, Li S (2016) Warm season bias of branched GDGT temperature estimates causes underestimation of altitudinal lapse rate. Org Geochem 96:11–17 Wang H et al (2024) New calibration of terrestrial brGDGT paleothermometer deconvolves distinct temperature responses of two isomer sets. Earth Planet Sci Lett 626:118497 Clemens SC, Prell WL, Sun Y, Liu Z, Chen G (2008) Southern Hemisphere forcing of Pliocene δ18O and the evolution of Indo-Asian monsoons. Paleoceanography 23 Wen Y et al (2023) CO2-forced Late Miocene cooling and ecosystem reorganizations in East Asia. Proceedings of the National Academy of Sciences 120, e2214655120 An Z et al (2005) Multiple expansions of C4 plant biomass in East Asia since 7 Ma coupled with strengthened monsoon circulation. Geology 33:705–708 Sun J, Lü T, Zhang Z, Wang X, Liu W (2012) Stepwise expansions of C4 biomass and enhanced seasonal precipitation and regional aridity during the Quaternary on the southern Chinese Loess Plateau. Q Sci Rev 34:57–65 Long H et al (2025) The westerlies-monsoon interaction shaped asymmetric lake expansions over the Tibetan Plateau in warming periods. Sci Bull. https://doi.org/10.1016/j.scib.2025.07.023 Kong W, Swenson LM, Chiang JCH (2017) Seasonal Transitions and the Westerly Jet in the Holocene East Asian Summer Monsoon. J Clim 30:3343–3365 Huybers P, Wunsch C (2005) Obliquity pacing of the late Pleistocene glacial terminations. Nature 434:491–494 Li M, Hinnov LA, Huang C, Ogg JG (2018) Sedimentary noise and sea levels linked to land–ocean water exchange and obliquity forcing. Nat Commun 9:1004 Qin J et al (2022) 1.2 Myr Band of Earth-Mars Obliquity Modulation on the Evolution of Cold Late Miocene to Warm Early Pliocene Climate. Journal of Geophysical Research: Solid Earth 127, e2022JB024131 Retallack GJ (2005) Pedogenic carbonate proxies for amount and seasonality of precipitation in paleosols. Geology 33:333–336 An Z et al (2014) Late Cenozoic climate change in monsoon-arid Asia and global changes in. Late Cenozoic Climate Change in Asia. Springer, pp 491–581 Liu X et al (2001) Magnetic properties of the Tertiary red clay from Gansu. Sci China Ser D-Earth Sci 44:635–651 Sun D, Liu D, Chen M, An Z, John S (1997) Magnetostratigraphy and palaeoclimate of Red Clay sequences from Chinese Loess Plateau. Sci China Ser D-Earth Sci 40:337–343 Qiang XK, Li Z-X, Powell CM, Zheng HB (2001) Magnetostratigraphic record of the Late Miocene onset of the East Asian monsoon, and Pliocene uplift of northern Tibet. Earth Planet Sci Lett 187:83–93 Ao H et al (2021) Global warming-induced Asian hydrological climate transition across the Miocene–Pliocene boundary. Nat Commun 12:6935 Retallack GJ (2009) Refining a pedogenic-carbonate CO2 paleobarometer to quantify a middle Miocene greenhouse spike. Palaeogeogr Palaeoclimatol Palaeoecol 281:57–65 Huntington KW et al (2009) Methods and limitations of ‘clumped’ CO2 isotope (∆47) analysis by gas-source isotope ratio mass spectrometry. J Mass Spectrom 44:1318–1329 Brand WA, Assonov SS, Coplen TB (2010) Correction for the 17O interference in δ(13C) measurements when analyzing CO2 with stable isotope mass spectrometry (IUPAC Technical Report). Pure and Applied Chemistry 82, 1719–1733 Daëron M, Blamart D, Peral M, Affek HP (2016) Absolute isotopic abundance ratios and the accuracy of ∆47 measurements. Chem Geol 442:83–96 Schauer AJ, Kelson J, Saenger C, Huntington KW (2016) Choice of 17O correction affects clumped isotope (∆47) values of CO2 measured with mass spectrometry. Rapid Commun Mass Spectrom 30:2607–2616 Dennis KJ, Affek HP, Passey BH, Schrag DP, Eiler JM (2011) Defining an absolute reference frame for ‘clumped’ isotope studies of CO2. Geochim Cosmochim Acta 75:7117–7131 Bernasconi SM et al (2021) InterCarb: A Community Effort to Improve Interlaboratory Standardization of the Carbonate Clumped Isotope Thermometer Using Carbonate Standards. Geochemistry, Geophysics, Geosystems 22, e2020GC009588 Wostbrock JAG, Cano EJ, Sharp ZD (2020) An internally consistent triple oxygen isotope calibration of standards for silicates, carbonates and air relative to VSMOW2 and SLAP2. Chem Geol 533:119432 Perdue N, Sharp Z, Nelson D, Wehr R, Dyroff C (2022) A rapid high-precision analytical method for triple oxygen isotope analysis of CO2 gas using tunable infrared laser direct absorption spectroscopy. Rapid Commun Mass Spectrom 36:e9391 Tremaine DM, Froelich PN, Wang Y (2011) Speleothem calcite farmed in situ: Modern calibration of δ18O and δ13C paleoclimate proxies in a continuously-monitored natural cave system. Geochim Cosmochim Acta 75:4929–4950 Wostbrock JAG et al (2020) Calibration of carbonate-water triple oxygen isotope fractionation: Seeing through diagenesis in ancient carbonates. Geochim Cosmochim Acta 288:369–388 Luz B, Barkan E (2010) Variations of 17O/16O and 18O/16O in meteoric waters. Geochim Cosmochim Acta 74:6276–6286 Fischer-Femal BJ, Bowen GJ (2021) Coupled carbon and oxygen isotope model for pedogenic carbonates. Geochim Cosmochim Acta 294:126–144 Aron PG et al (2021) Triple oxygen isotopes in the water cycle. Chem Geol 565:120026 Bowen GJ, Fischer-Femal B, Reichart G-J, Sluijs A, Lear CH (2020) Joint inversion of proxy system models to reconstruct paleoenvironmental time series from heterogeneous data. Clim Past 16:65–78 Lunn D, Jackson C, Best N, Thomas A, Spiegelhalter D (2013) The BUGS book. A practical introduction to Bayesian analysis. Chapman Hall, London Core Team R (2024) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing Plummer M (2018) rjags: Bayesian graphical models using MCMC (R package version 4–6, 2016). Available at : https://CRAN.Rproject.org/package= rjags Su Y-S, Yajima M, Su MY-S, SystemRequirements J (2015) Package ‘r2jags.’ R package version 0.03-08, URL http://CRAN . R-project . org/package = R2jags Gelman A, Rubin DB (1992) Inference from Iterative Simulation Using Multiple Sequences. Stat Sci 7:457–472 Danabasoglu G et al (2019) The Community Earth System Model Version 2 (CESM2). Journal of Advances in Modeling Earth Systems 12, eMS001916 (2020) Dowsett H et al (2016) The PRISM4 (mid-Piacenzian) paleoenvironmental reconstruction. Clim Past 12:1519–1538 Tierney JE, Tingley MP (2018) BAYSPLINE: A New Calibration for the Alkenone Paleothermometer. Paleoceanography Paleoclimatology 33:281–301 Additional Declarations There is NO Competing Interest. Supplementary Files C4jetsupplementarymaterials.docx supplementary material Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8290760","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":570897854,"identity":"eae2c848-7124-4383-af4e-1fe7184e2b08","order_by":0,"name":"Jiawei Da","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAr0lEQVRIiWNgGAWjYBACxmYQaWAD4fGQoCWNBC1QcJgELcztzM8kPhScz5OfkcD44G0bUQ5jM5OcYXC72OBGArPhXOK0MJjd5jG4nbhBIoFNmpc4Lezfbv8xOJc4f0YC+28itfCY3WYwOJDYcCOBjZlYLeU/ewySiw3OPGyWnHOOCC2G/cc3G/z4Y5cn35588MObMmK0NEDoBKCFDUSoBwJ5BriWUTAKRsEoGAU4AACF6jTx7YB1zgAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0002-4483-4651","institution":"The Pennsylvania State University","correspondingAuthor":true,"prefix":"","firstName":"Jiawei","middleName":"","lastName":"Da","suffix":""},{"id":570897855,"identity":"5e89ee16-89e6-430a-a817-4a015738ad18","order_by":1,"name":"Chijun Sun","email":"","orcid":"https://orcid.org/0000-0002-3668-346X","institution":"University of California Davis","correspondingAuthor":false,"prefix":"","firstName":"Chijun","middleName":"","lastName":"Sun","suffix":""},{"id":570897856,"identity":"5b7ea795-50f8-4e47-84ea-7985125e166c","order_by":2,"name":"Lily Serach","email":"","orcid":"","institution":"The University of Texas at Austin","correspondingAuthor":false,"prefix":"","firstName":"Lily","middleName":"","lastName":"Serach","suffix":""},{"id":570897857,"identity":"08a39328-3163-49a1-ac97-bc4cd1cb44ca","order_by":3,"name":"Timothy Gallagher","email":"","orcid":"https://orcid.org/0000-0002-7271-1210","institution":"Kent State University","correspondingAuthor":false,"prefix":"","firstName":"Timothy","middleName":"","lastName":"Gallagher","suffix":""},{"id":570897858,"identity":"dcaf9c69-96c8-4755-94d8-0b14ae63389a","order_by":4,"name":"Huayu Lu","email":"","orcid":"https://orcid.org/0000-0002-6306-6985","institution":"Nanjing University","correspondingAuthor":false,"prefix":"","firstName":"Huayu","middleName":"","lastName":"Lu","suffix":""},{"id":570897859,"identity":"e759585f-7c98-4e96-9081-88d2616ebee0","order_by":5,"name":"Katharine Huntington","email":"","orcid":"https://orcid.org/0000-0003-0796-8850","institution":"University of Washington","correspondingAuthor":false,"prefix":"","firstName":"Katharine","middleName":"","lastName":"Huntington","suffix":""},{"id":570897860,"identity":"9764ffce-c26e-4dcb-9786-d574ffe89f78","order_by":6,"name":"Ran Feng","email":"","orcid":"https://orcid.org/0000-0002-4433-4745","institution":"University of Connecticut","correspondingAuthor":false,"prefix":"","firstName":"Ran","middleName":"","lastName":"Feng","suffix":""},{"id":570897861,"identity":"0ca9ba8e-ebf3-4088-aaa1-c11caf836d7d","order_by":7,"name":"Hanzhi Zhang","email":"","orcid":"","institution":"Nanjing University","correspondingAuthor":false,"prefix":"","firstName":"Hanzhi","middleName":"","lastName":"Zhang","suffix":""},{"id":570897862,"identity":"07e68bb3-4e8e-40eb-b033-8aa12a50aa91","order_by":8,"name":"Hanlin Wang","email":"","orcid":"","institution":"Nanjing Institute of Geology and Paleontology, Chinese Academy of Sciences","correspondingAuthor":false,"prefix":"","firstName":"Hanlin","middleName":"","lastName":"Wang","suffix":""},{"id":570897863,"identity":"b9f6a73e-cdfb-491f-ba7f-b861d7564d1d","order_by":9,"name":"Shunchuan Ji","email":"","orcid":"","institution":"Yunnan Normal University","correspondingAuthor":false,"prefix":"","firstName":"Shunchuan","middleName":"","lastName":"Ji","suffix":""},{"id":570897864,"identity":"cb32778e-e7b0-49d9-8592-d62d2bf30825","order_by":10,"name":"Zachary Sharp","email":"","orcid":"","institution":"University of New Mexico","correspondingAuthor":false,"prefix":"","firstName":"Zachary","middleName":"","lastName":"Sharp","suffix":""},{"id":570897865,"identity":"4690d234-b9ed-462a-867d-ba70f1d47dff","order_by":11,"name":"Junfeng Ji","email":"","orcid":"https://orcid.org/0000-0003-3754-731X","institution":"Nanjing University","correspondingAuthor":false,"prefix":"","firstName":"Junfeng","middleName":"","lastName":"Ji","suffix":""},{"id":570897866,"identity":"e181bb73-ac29-4186-a934-503128a0afb1","order_by":12,"name":"Daniel Breecker","email":"","orcid":"https://orcid.org/0000-0003-0200-223X","institution":"University of Texas at Austin","correspondingAuthor":false,"prefix":"","firstName":"Daniel","middleName":"","lastName":"Breecker","suffix":""}],"badges":[],"createdAt":"2025-12-05 21:20:24","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8290760/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8290760/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":99799095,"identity":"a33fef38-f31b-4cac-92f0-70386acdc72f","added_by":"auto","created_at":"2026-01-08 13:49:13","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":2550725,"visible":true,"origin":"","legend":"","description":"","filename":"C4jetmaintext.docx","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/98b2a710e32d64dbbc53d255.docx"},{"id":99799505,"identity":"ced02ed0-f350-4a76-a574-8a29498f2ec8","added_by":"auto","created_at":"2026-01-08 13:49:41","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":12693,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS25105841T.json","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/fcf039e75a73a1187b20d44d.json"},{"id":99763307,"identity":"edf32f0c-ee74-4853-abc4-18216fda540c","added_by":"auto","created_at":"2026-01-08 07:34:55","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1260984,"visible":true,"origin":"","legend":"","description":"","filename":"C4jetsupplementarymaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/9920095b146676598c727b0c.docx"},{"id":99799079,"identity":"621bab0b-e975-4244-9421-a781dce603ec","added_by":"auto","created_at":"2026-01-08 13:49:13","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":166242,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS25105841T0enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/f74147bb6428c86b410d57bc.xml"},{"id":99799298,"identity":"e2d1c45e-7753-468d-bb1c-df84e37cf639","added_by":"auto","created_at":"2026-01-08 13:49:26","extension":"png","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1001927,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/2534bced36cdee5f89a87442.png"},{"id":99763306,"identity":"abb02200-1ab9-4adb-b409-58ab229aa078","added_by":"auto","created_at":"2026-01-08 07:34:54","extension":"png","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":355582,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/78e8929bba633b8ce2f3257c.png"},{"id":99798635,"identity":"ff989652-5be8-476f-995a-8c5859955c57","added_by":"auto","created_at":"2026-01-08 13:48:41","extension":"png","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":542227,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/4abb719457f5667cdce06293.png"},{"id":99763314,"identity":"2355c42f-4a4f-48f2-b9a4-ee3adec081ca","added_by":"auto","created_at":"2026-01-08 07:34:55","extension":"jpeg","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":557669,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/ec8520a86c9ff24119e130b2.jpeg"},{"id":99763310,"identity":"893afcd3-c345-46e1-89de-c278ef3a7181","added_by":"auto","created_at":"2026-01-08 07:34:55","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":134780,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/4866846b0c6867bae3b58448.png"},{"id":99763312,"identity":"9278e5e5-b838-4b53-bb3f-4d04a912da07","added_by":"auto","created_at":"2026-01-08 07:34:55","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":62435,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/56cef129cdd38d804bf2b7cb.png"},{"id":99798712,"identity":"2ae1ea03-e55e-4431-ad53-7b7a668e3d0f","added_by":"auto","created_at":"2026-01-08 13:48:50","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":78143,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/733af546b2e58feeadfdf66f.png"},{"id":99798704,"identity":"cb1abd6e-b34b-42b0-9a83-7963395909f3","added_by":"auto","created_at":"2026-01-08 13:48:49","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":128784,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/5aa32b337127bdbf1d67720e.png"},{"id":99797544,"identity":"1b3f62a5-fe4e-4ec9-b6e1-7cbbde19f9da","added_by":"auto","created_at":"2026-01-08 13:45:59","extension":"xml","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":163624,"visible":true,"origin":"","legend":"","description":"","filename":"NCOMMS25105841T0structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/d6fe00726af3f7b6b2b80b84.xml"},{"id":99763316,"identity":"354570f5-da09-4677-a475-9605f9a2afac","added_by":"auto","created_at":"2026-01-08 07:34:55","extension":"html","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":172215,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/48d0a4412761f8c4243628e8.html"},{"id":99763298,"identity":"98ad22f6-036d-4327-ba22-6f2caf745e74","added_by":"auto","created_at":"2026-01-08 07:34:54","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":567157,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eStudy region maps and precipitation.\u003c/strong\u003e \u003cstrong\u003e(a)\u003c/strong\u003e Modern natural C\u003csub\u003e4\u003c/sub\u003e grass coverage over East Asia (6). Open squares and colored diamonds show locations of organic δ\u003csup\u003e13\u003c/sup\u003eC records mentioned in this study (7–10) and the modern rainfall δ\u003csup\u003e18\u003c/sup\u003eO data in panel c, respectively. Grey, blue, and orange arrows show schematically the spring cold air outbreaks that deliver dust to the Japan Sea, the mean summer position of the subtropical westerly jet (11), and the low-level southerly monsoonal flow. The yellow box shows the Chinese Loess Plateau (CLP). \u003cstrong\u003e(b)\u003c/strong\u003e Shaded relief map of the CLP. Circles and squares show sampling sites of this study and from literature mentioned here, respectively. \u003cstrong\u003e(c)\u003c/strong\u003e The monthly rainfall-δ\u003csup\u003e18\u003c/sup\u003eO of the GNIP (Global Network of Isotopes in Precipitation) sites across the Asian monsoon region. \u0026nbsp;\u003cstrong\u003e(d)\u003c/strong\u003e Hovmöller diagram of the precipitation climatology (unit: mm/day) between 110°E and 120°E (modified from 12). The blue and red horizontal lines mark the seasonal migration of the westerly jet under modern conditions, and its likely position during the warm Pliocene, respectively. The grey shade shows the latitudinal position of the CLP.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/0853585ebd2ad6c4d350be6d.png"},{"id":99763299,"identity":"7b792677-60c7-400f-970e-93a617e9a1aa","added_by":"auto","created_at":"2026-01-08 07:34:54","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":320887,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe long-term evolution of C\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e4\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e biomass over East Asia and global climate.\u003c/strong\u003e \u003cstrong\u003e(a)\u003c/strong\u003e Atmospheric CO\u003csub\u003e2\u003c/sub\u003e estimates based on published proxy records via Bayesian inversion (5). The dark line and shade show median and 68% credible interval, respectively. \u003cstrong\u003e(b)\u003c/strong\u003e LR04 benthic δ\u003csup\u003e18\u003c/sup\u003eO stack (16). \u003cstrong\u003e(c) \u003c/strong\u003eNormalized meridional\u0026nbsp;sea surface temperature\u0026nbsp;gradient\u0026nbsp;(ΔSST\u003csub\u003enorm\u003c/sub\u003e)\u0026nbsp;based on the\u0026nbsp;differences of compiled SST\u0026nbsp;records from tropical and mid-latitude oceans (green curve) and from tropical and high-latitude oceans (17)\u0026nbsp;(brown curve, see\u0026nbsp;\u003cem\u003eMethods\u003c/em\u003e, Fig. S1). \u003cstrong\u003e(d)\u003c/strong\u003e Black carbon δ\u003csup\u003e13\u003c/sup\u003eC from IODP (Integrated Ocean Drilling Program) U1430 in the Japan Sea (8). \u003cstrong\u003e(e)\u003c/strong\u003e Sediment organic matter δ\u003csup\u003e13\u003c/sup\u003eC from the G3 drilling core on the North China Plain (7). \u003cstrong\u003e(f)\u003c/strong\u003e Black carbon δ\u003csup\u003e13\u003c/sup\u003eC from Lingtai section at southern CLP (10). \u003cstrong\u003e(g)\u003c/strong\u003e Reconstructed δ\u003csup\u003e18\u003c/sup\u003eO values of soil water (δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e) from this study color-coded by sampling sites. \u003cstrong\u003e(h)\u003c/strong\u003e Schematic diagram showing the meridional migration of the summer position of the westerly jet stream, drawn to be consistent with MTG changes inferred from benthic δ\u003csup\u003e18\u003c/sup\u003eO and ΔSST\u003csub\u003enorm\u003c/sub\u003e. The oscillations in CLP δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e and the various organic carbon δ\u003csup\u003e13\u003c/sup\u003eC records can be understood as responses to changes in rainfall seasonality controlled by the movement of the westerly-jet tracking summer rain band.\u0026nbsp;\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/31d80a9404b50a81a1787dea.png"},{"id":99763301,"identity":"09ca1ddc-de8f-4ae0-837c-fe9735d1ef6f","added_by":"auto","created_at":"2026-01-08 07:34:54","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":466913,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eReconstructed soil-water isotopic compositions. (a-b)\u003c/strong\u003e Obliquity signal in La2004 astronomical solution and 25°N - 65°N summer insolation gradient (SIG) along with their filtered ∼1.2-Myr amplitude modulation cycles, respectively. \u003cstrong\u003e(c-e)\u003c/strong\u003e Time series of the reconstructed T\u003csub\u003eΔ47\u003c/sub\u003e, δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e, and Δ\u003csup\u003e'17\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e, color-coded by sampling sites. Calcite formation temperatures used to calculate δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e were interpolated from T\u003csub\u003eΔ47\u003c/sub\u003e of individual sections. δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e calculated from paired δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e-T\u003csub\u003eΔ47\u003c/sub\u003e measurements show similar variations across sections (Fig. S3). The Δ′\u003csup\u003e17\u003c/sup\u003eO of soil waters (Δ′\u003csup\u003e17\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e) were calculated from paired T\u003csub\u003eΔ47\u003c/sub\u003e-Δ′\u003csup\u003e17\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e measurements (\u003cem\u003eMethods\u003c/em\u003e). Error bars and the black line in panel c show 1σ errors from replicate analyses and the LOESS line (span = 0.3), respectively. Error bars in panels d and e extend to the 16\u003csup\u003eth\u003c/sup\u003e and 84\u003csup\u003eth\u003c/sup\u003e percentiles of distributions propagated using Monte Carlo random sampling. Note the reversed y-scales for SIG and Δ′\u003csup\u003e17\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e. \u003cstrong\u003e(f)\u003c/strong\u003e Summer monsoon index (SMI) based on soil magnetic susceptibility and carbonate content from Lingtai section\u003csup\u003e14\u003c/sup\u003e. \u003cstrong\u003e(g)\u003c/strong\u003e Free iron to total iron ratio from Pianguan section (23). The red curves in panels f-g show the 1.2-Myr obliquity amplitude modulation cycle. \u003cstrong\u003e(h) \u003c/strong\u003eReconstructed Red Clay soil waters compared with global Holocene soil waters (24) (red crosses) and modern tap waters across China (25) (grey circles). The red arrow highlights the evaporation trend seen in Holocene soil waters.\u003cstrong\u003e (i) \u003c/strong\u003eModeled estimates of δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e using the Bayesian inversion of a steady-state soil water isotope model (24) plotted against δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e (see \u003cem\u003eMethods\u003c/em\u003e). Error bars extend to the 16\u003csup\u003eth\u003c/sup\u003e and 84\u003csup\u003eth\u003c/sup\u003e percentiles of the modeled posterior distributions. Red Clay data in panels f and g are color-coded by T\u003csub\u003eΔ47\u003c/sub\u003e and shape-coded by site.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/a7873d8320c7edbef4a80310.png"},{"id":99763302,"identity":"73bf9cb6-2007-4a1f-af02-befc5f6a8625","added_by":"auto","created_at":"2026-01-08 07:34:54","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":618090,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe late Miocene-Pliocene evolution of the westerly jet, C\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e4\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e biomass, and global climate.\u003c/strong\u003e \u003cstrong\u003e(a)\u003c/strong\u003e LR04 benthic δ\u003csup\u003e18\u003c/sup\u003eO stack (16). \u003cstrong\u003e(b-e) \u003c/strong\u003eOrganic carbon δ\u003csup\u003e13\u003c/sup\u003eC records across East Asia, including black carbon from U1430 (8), sediment organic matter from the G3 drilling core (7), soil organic matter δ\u003csup\u003e13\u003c/sup\u003eC from Jiaxian section (9), and black carbon from Lingtai section (10). Note C\u003csub\u003e3\u003c/sub\u003e data after 3.5 Ma are outside the axis range and therefore not shown here. \u003cstrong\u003e(f-g)\u003c/strong\u003e Reconstructed δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e and T\u003csub\u003eΔ47\u003c/sub\u003e from this study (same color scheme as in Fig. 3). \u003cstrong\u003e(h)\u003c/strong\u003e Meridional\u0026nbsp;sea surface temperature\u0026nbsp;gradient\u0026nbsp;(ΔSST)\u0026nbsp;based on compiled\u0026nbsp;SST\u0026nbsp;records\u0026nbsp;from tropical and high-latitude oceans (see\u0026nbsp;\u003cem\u003eMethods\u003c/em\u003e, Fig. S1). \u003cstrong\u003e(j)\u003c/strong\u003e Atmospheric CO\u003csub\u003e2\u003c/sub\u003e\u0026nbsp;estimates (5).\u0026nbsp;\u003cstrong\u003e(k)\u003c/strong\u003e Schematic diagram showing\u0026nbsp;the meridional migration of the jet position during summer and April with respect to sampling locations of δ\u003csup\u003e13\u003c/sup\u003eC records, drawn to be consistent with MTG changes inferred from ΔSST. The\u0026nbsp;yellow\u0026nbsp;curves in panels f-h\u0026nbsp;show the 1.2-Myr obliquity amplitude modulation cycle.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/13f4f686fc4ba9f733320c3c.png"},{"id":100356428,"identity":"523d0227-6f20-4dde-9404-feb1f0759003","added_by":"auto","created_at":"2026-01-16 07:09:09","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2994952,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/14ceb42c-59a5-4a45-b07f-e2092d458164.pdf"},{"id":99763303,"identity":"309aac34-240d-48ff-b85a-5ea6cac52bcf","added_by":"auto","created_at":"2026-01-08 07:34:54","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":1260984,"visible":true,"origin":"","legend":"supplementary material","description":"","filename":"C4jetsupplementarymaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-8290760/v1/cb4183786e56c2323c308768.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"\u003cp\u003eJet-Induced Rainfall Seasonality and C\u003csub\u003e4\u003c/sub\u003e Expansion over East Asia\u003c/p\u003e","fulltext":[{"header":"Introduction","content":"\u003cp\u003eGrowing evidence suggests that the late Neogene C\u003csub\u003e4\u003c/sub\u003e expansion occurred asynchronously across different continents (1), from as early as ~ 21\u0026nbsp;million years ago (Ma) in eastern Africa (2) and 8 − 6 Ma on the India subcontinent (3) to 3.5 Ma in Australia (4). Although decreasing atmospheric CO\u003csub\u003e2\u003c/sub\u003e concentrations may have set the proverbial stage (5), this demonstrable global asynchroneity of C\u003csub\u003e4\u003c/sub\u003e expansion requires regional drivers or thresholds, which are particularly poorly understood in mid-latitude regions such as East Asia (Fig. 1).\u003c/p\u003e\n\u003cp\u003eEast Asian records based mainly on the δ\u003csup\u003e13\u003c/sup\u003eC of tooth enamel and soil carbonates (δ\u003csup\u003e13\u003c/sup\u003eC\u003csub\u003ec\u003c/sub\u003e) from the Chinese Loess Plateau (CLP) and surrounding regions, suggest C\u003csub\u003e4\u003c/sub\u003e expansion during the late Miocene (~ 6 Ma) (13). However, neither of these proxies provide quantitative estimates of regional C\u003csub\u003e4\u003c/sub\u003e abundance. Specifically, tooth enamel may more closely record presence/absence rather than C\u003csub\u003e4\u003c/sub\u003e abundance in an ecosystem (14), and in arid-to-semi-arid ecosystems such as the CLP, soil carbonates record a mixed signal of biomass-δ\u003csup\u003e13\u003c/sup\u003eC, atmospheric CO\u003csub\u003e2\u003c/sub\u003e, and productivity (15).\u003c/p\u003e\n\u003cp\u003eIn contrast, soil organic matter – a more direct indicator of biomass δ\u003csup\u003e13\u003c/sup\u003eC – suggests that C\u003csub\u003e4\u003c/sub\u003e expansion across East Asia was spatiotemporally heterogeneous (Fig. 2d-f). Consequently, multiple drivers of C\u003csub\u003e4\u003c/sub\u003e expansion have been proposed, including decreasing atmospheric CO\u003csub\u003e2\u003c/sub\u003e levels (7), increased warm season precipitation (13), intensified fire activity (10), and enhanced long-term aridity (8). Importantly, studies on the CLP reveal hominin occupation as early as 2.1 Ma and a close coupling between hominin migration and regional environmental conditions (18, 19). Understanding the spatiotemporal shifting of C₄-rich biomes in East Asia is thus critical for not only elucidating vegetation–climate feedbacks but also for contextualizing the evolution of hominins, which preferred C\u003csub\u003e3\u003c/sub\u003e woody plants for shade, shelter, and nutrition (20).\u003c/p\u003e\n\u003cp\u003eThe regional climate over East Asia is regulated by the East Asian monsoon system (EAM). The EAM is unique owing to the dynamic interaction between the low-level, southerly monsoonal flow and the upper-level westerly jet impinging on the Tibetan Plateau (21) (Fig. 1a). Unlike classic tropical monsoons featuring one prolonged rainfall stage, the EAM is characterized by several quasi-stationary rainfall stages, resulting from the seasonal poleward shift of the westerly jet during boreal summer (Fig. 1d) (22). Mounting evidence from paleoclimate records and climate models underscores the role of this jet transition in controlling the meridional migration of the monsoonal rain band, and consequently rainfall seasonality over East Asia on timescales from millennia to glacial cycles (12, 21). However, the role of the jet transition in rainfall seasonality over longer timescales remains unknown.\u003c/p\u003e\n\u003cp\u003eWe hypothesized that the asynchronous C\u003csub\u003e4\u003c/sub\u003e expansion and retreat in East Asia was paced by shifting warm season precipitation in response to the long-term equatorward migration of the westerly jet during Plio-Pleistocene global cooling. To test this hypothesis, we reconstructed δ\u003csup\u003e18\u003c/sup\u003eO values (δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e) of Mio-Pliocene rainfall over the CLP as an independent test of long-term, jet-transition-driven changes in East Asian rainfall seasonality. We also evaluated, by considering paleoclimate simulations, whether the concept of jet-transition-modulated changes in rainfall seasonality is plausible and, by comparison with previously published paleoenvironmental records, whether our hypothesized mechanism is broadly explanatory.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCoupled T\u003csub\u003eΔ47\u003c/sub\u003e - δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e time series\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe reconstructed rainfall δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e using the clumped isotope (Δ\u003csub\u003e47\u003c/sub\u003e) and triple oxygen isotope compositions of calcite nodules collected from three Red Clay sections (Lantian, Shilou, and Jiaxian) on the CLP (Fig. 1b, see \u003cem\u003eMethods\u003c/em\u003e). The clumped isotope temperature (T\u003csub\u003eΔ47\u003c/sub\u003e) and δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e from these three sections on the CLP are coherent across the late Miocene - Pliocene (Fig. 3c-d). The most prominent change occurred across the MPB (5.6-5.0 Ma), when T\u003csub\u003eΔ47\u003c/sub\u003e and δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e increased from 12°C (SE = 1°C, n = 13) and − 9.6‰ (SE = 0.05‰, n = 105) on average during the late Miocene (7-5.3 Ma), to as high as 31 ± 4°C and − 6.5 ± 0.5‰ at 5.1 Ma. Similar changes with smaller magnitudes are observed at 6.2 Ma and 3.5 Ma.\u003c/p\u003e\n\u003cp\u003eThe Δ′\u003csup\u003e17\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e values show high fluctuations ranging between − 95 ± 32 per meg at 5.1 Ma and 27 ± 20 per meg at 3.1 Ma (Fig. 3e), with a mean value of -28 per meg (SD = 24 per meg, n = 22). The Δ′\u003csup\u003e17\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e are significantly lower than those of modern meteoric waters across China (25) but generally overlap with those of Holocene soil waters calculated using a global Δ\u003csub\u003e47\u003c/sub\u003e - Δ′\u003csup\u003e17\u003c/sup\u003eO dataset of pedogenic carbonates formed in various environments (24) (Fig. 3h). However, there is no coherent trend in Δ′\u003csup\u003e17\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e through time (Fig. 3e) or in Δ′\u003csup\u003e17\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e - δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e space (Fig. 3h), indicating that the observed δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e variability is not primarily driven by evaporation.\u003c/p\u003e\n\u003cp\u003eTo quantitatively assess the effects of evaporation and δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e on δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e, we applied a steady-state soil-water-isotope model to our T\u003csub\u003eΔ47\u003c/sub\u003e-δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e-Δ′\u003csup\u003e17\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e dataset and obtained posterior δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e estimates using Bayesian inversion (see \u003cem\u003eMethods\u003c/em\u003e). The modeled δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e values correlate positively with δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e (Fig. 3i), indicating that δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e variations are largely explained by changes in δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e, with higher δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e under warm climate.\u003c/p\u003e\n\u003cdiv id=\"Sec3\"\u003e\n \u003ch2\u003eWesterly Jet-induced rainfall seasonality\u003c/h2\u003e\n \u003cp\u003eIn the EAM region, δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e records are typically interpreted to record rainout, with higher δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e corresponding to a weaker monsoon circulation and a drier climate associated with less rainfall upstream and/or less summer rainfall locally (26, 27). We propose that instead of annual rainfall amount or the overall EAM strength, our observed δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e variations record changes in rainfall seasonality closely linked to the westerly jet stream dynamics.\u003c/p\u003e\n \u003cp\u003eUnder warmer climates, such as the beginning of the Pliocene, the reduced MTG (Fig. 4h) resulted in the weakening and northward migration of the westerly jet (28), leading to a seasonally earlier onset (spring rather than summer) of rainfall over northern China (21) (Fig. 4j). Moreover, the weakened jet impinging upstream on the Tibetan Plateau caused a weaker stationary Rossby wave downstream, resulting in high pressure anomalies over East Asia and the northward shift of the North Pacific subtropical high (NPSH) (29). Consequently, the southerly wind along the northwestern edge of the NPSH migrated northward, resulting in an earlier onset of spring rainfall.\u003c/p\u003e\n \u003cp\u003eThe modern monthly δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e pattern across the EAM region is characterized by the transition from the higher-δ\u003csup\u003e18\u003c/sup\u003eO spring to lower-δ\u003csup\u003e18\u003c/sup\u003eO summer rain, primarily controlled by distillation upstream (27) (Fig. 1c). Because the seasonal march of the monsoonal rain was already established during the Miocene (30), it is reasonable to assume a quasi-stationary monthly δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e climatology through time. The modern spring-to-summer δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e shift exceeds 5‰ (Fig. 1c), comparable to changes in the reconstructed δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e (Fig. 2g). Thus, the T\u003csub\u003eΔ47\u003c/sub\u003e - δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e coupling supports rainfall seasonality changes as the main driver of the observed δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e variability, with more spring rain (high δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e) during warmer times when the westerly jet was positioned at a higher latitude.\u003c/p\u003e\n \u003cp\u003eSeveral lines of evidence support the MTG – jet – rainfall seasonality as the drivers of coupled T\u003csub\u003eΔ47\u003c/sub\u003e - δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e variations. First, our T\u003csub\u003eΔ47\u003c/sub\u003e records show large amplitude of variations as high as \u0026gt; 10°C increase across the MPB (Fig. 3c). For comparison, mean annual surface temperature variability on the CLP across the Pleistocene glacial cycles is less than 10°C (31). Given that regional climate variability is much less pronounced during the late Miocene-Pliocene periods (32), it is unlikely that changes in T\u003csub\u003eΔ47\u003c/sub\u003e are solely caused by mean annual temperature change. Pedogenic carbonates tend to record seasonal climate signals rather than annual means (33). In arid regions such as the CLP, pedogenic carbonate formation is primarily driven by rainfall events through its control over soil water chemistry (34). Consequently, pedogenic carbonates grow after the rainy season under enhanced evaporation and reduced soil respiration (35). During cold periods, rainfall occurred within summer due to the jet at a lower latitude. Consequently, pedogenic carbonates grew during late summer and early fall, recording low summer-δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e and fall temperature. During warmer periods, an early onset of spring rain and intensive summer evaporation likely facilitate pedogenic carbonate growth during mid-summer, recording high spring-δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e and summer temperature. Modern surface temperature difference on the CLP between July and October can be as high as 17°C. Thus, the large T\u003csub\u003eΔ47\u003c/sub\u003e variability is likely caused by seasonal shifts in soil carbonate formation superimposed on the long-term changes in annual mean temperature.\u003c/p\u003e\n \u003cp\u003eMoreover, minima in meridional sea surface temperature (SST) gradients (17) at 5 and 3.5 Ma correspond well with maxima in T\u003csub\u003eΔ47\u003c/sub\u003e and δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e (Fig. 4f-h). The effect of the MTG on jet position and strength is also documented by dust flux and provenance records from the North Pacific Ocean (28, 36), suggesting poleward and weakened westerly winds during the warm Pliocene compared with the Pleistocene glacial periods.\u003c/p\u003e\n \u003cp\u003eThe jet transition hypothesis is also supported by model simulations. Specifically, we looked at three experiments, mid-Pliocene 280 ppm (Eoi280) and 400 ppm (Eoi400) and pre-industrial control (E280), from the Community Earth System Model version 2 (CESM 2), following the Pliocene Model Intercomparison Project phase 2 (PMIP2) (37, 38) (See Methods). Both Eoi280 and Eoi400 show a weakening and northward shift of the westerly jet compared to E280 (Fig. S6a-b), which led to a northward shift of the western Pacific subtropical high and an earlier seasonal march of the EAM (Fig. S6c-d). These shifts simulated using CESM 2 are consistent with the PMIP2 multi-model ensemble mean (39). Consequently, northern China (\u0026gt; 30°N) received more spring rain and less summer rain during the warm Pliocene compared to the pre-industrial conditions (Fig. S6e-f).\u003c/p\u003e\n \u003cp\u003eFinally, rainfall seasonality reconciles the discrepancy among various proxy records. For instance, soil magnetic susceptibility is mainly controlled by the abundance of ultrafine ferrimagnetic particles (40) formed during redox potential changes caused by soil wet-dry cycles (41). The wet-dry cycles, and presumably redox potential, are most prominent during the summer when potential evapotranspiration is large and soil respiration is water-limited (42). In this view, the increases in magnetic susceptibility toward the Pliocene-Pleistocene boundary (43) (Fig. S4b) were caused by larger or more frequent wet-dry cycles as monsoon rainfall seasonality shifted from spring to summer. This seasonality shift also explains the terrestrial snail record (Fig. S4e), which shows a dominance of warm-humid-adapted species during the early Pliocene. Because land snails tend to be active at high relative humidity (\u0026gt; 70%) and intermediate temperature (10–27°C) (44), spring rain during the early Pliocene resulted in greater water availability due to lower evapotranspiration, which favored the growth of humid-loving taxa. On the other hand, soil bacteria likely reached peak metabolism during the summer growing season (45, 46), and thus the soil tetraether proxy likely record summer drought during the warm early Pliocene (Fig. S4c).\u003c/p\u003e\n \u003cp\u003eDuring the Pliocene, rainfall maxima over East Asia occurred at obliquity minima as recorded by magnetic susceptibility from the CLP (a record of wet-dry cycles) and marine proxies related to runoff (e.g., salinity, productivity, lithogenic accumulation) from the South China Sea (47). This relationship is opposite of that expected from the effect of fast physics on monsoonal rainfall. For example, at obliquity minima, northern hemisphere summer insolation is small, which should result in smaller land-sea and cross-equatorial pressure gradients and thus weaker monsoonal circulation. Therefore, in the established framework of the EAM, this observed relationship has been enigmatic (47). However, if we view the EAM from the perspective of the jet transition, there is a dynamic explanation for rainfall maxima at obliquity minima. During the Pliocene, the mean position of the westerly jet was further north at obliquity maxima, which pushed summer rainfall maxima to a higher latitude (north of Jiaxian according to our records), and consequently, resulted in fewer summer wet-dry cycles on the CLP and less surface runoff across southern China. The zone of intense summer rainfall moved southward at obliquity minima, following the mean position of the westerly jet.\u003c/p\u003e\n\u003c/div\u003e\n\u003ch3\u003eWesterly Jet-induced C\u003csub\u003e4\u003c/sub\u003e migration\u003c/h3\u003e\n\u003cp\u003eThe meridional migration of the westerly jet provides a dynamic explanation for the asynchronous C\u003csub\u003e4\u003c/sub\u003e expansion in East Asia. Specifically, the summer jet position likely maintained at high latitude across the late Miocene – Pliocene, resulting in rainier spring in East Asia and consequently, the dominance of C\u003csub\u003e3\u003c/sub\u003e plants as indicated by the overall low organic carbon δ\u003csup\u003e13\u003c/sup\u003eC values (Fig. 2d-f). This is supported by model results, showing the summer jet located at \u0026gt; 40°N under Pliocene boundary conditions (Fig. S6).\u003c/p\u003e\n\u003cp\u003eThe long-term increase in MTG gradually pushed the mean jet position equatorward, leading to asynchronous C\u003csub\u003e4\u003c/sub\u003e expansion as suggested by records, from 5.5-4 Ma in the U1430 from the Japan Sea, 3.5–2.5 Ma in the G3 on the North China Plain, to 2 − 1 Ma in the southernmost Lingtai section (Fig. 2d-f). Specifically, because sediments at U1430 were mainly of eolian origin brought by the spring storm outbreaks and the westerlies (8) (Fig. 1a), they should largely originate from regions that experienced dry spring and thus we should expect the source region to track the summer rain band (Fig. 2h). The shifting regimes of sediment source explains (1) increased black carbon δ\u003csup\u003e13\u003c/sup\u003eC at U1430 during 5 − 4 Ma despite the southward migration of the jet and (2) the persistently high δ\u003csup\u003e13\u003c/sup\u003eC from 4.2–1.7 Ma (Fig. 2d). Notably, the mean summer jet position likely overlapped with the dust source regions of U1430 beginning in the late Miocene (Fig. 2h), yet δ\u003csup\u003e13\u003c/sup\u003eC did not increase until the MPB (Fig. 2d). We attribute this discrepancy to the significant surface cooling in East Asia during 8 − 7 Ma, which likely inhibited the growth of C\u003csub\u003e4\u003c/sub\u003e plants (48).\u003c/p\u003e\n\u003cp\u003eThe onset and intensification of the NHG during 3.6–2.5 Ma pushed the mean jet position further south (28), which intensified summer rainfall in northern China (23, 32) (Fig. 3f-g) and led to major C\u003csub\u003e4\u003c/sub\u003e expansion in the source regions of G3 sediments (Fig. 2e). With further cooling and high-latitude ice sheet expansion during the Pleistocene, the summer jet and associated rain band were pushed even further south across the CLP resulting in migration of C\u003csub\u003e4\u003c/sub\u003e vegetation from the G3 source area equatorward toward Lingtai (Fig. 2e-f). The Pleistocene C\u003csub\u003e4\u003c/sub\u003e expansion is also witnessed in organic δ\u003csup\u003e13\u003c/sup\u003eC records from other sections located on the southern CLP (49, 50). The long term equatorward C\u003csub\u003e4\u003c/sub\u003e migration is inconsistent with a direct control by decreasing atmospheric CO\u003csub\u003e2\u003c/sub\u003e. The late Pleistocene C\u003csub\u003e4\u003c/sub\u003e retreat from Lingtai can be explained by the slightly poleward shift of the jet since 1.1 Ma, as indicated by the decreased MTG and benthic δ\u003csup\u003e18\u003c/sup\u003eO (Fig. 2b-c).\u003c/p\u003e\n\u003cp\u003eThe jet transition also explains some of the small δ\u003csup\u003e13\u003c/sup\u003eC shifts across the Miocene-Pliocene. During the late Miocene cooling, as the jet shifted equatorward gradually, organic carbon δ\u003csup\u003e13\u003c/sup\u003eC increased by ~ 2‰ first in the north at G3, then at Jiaxian, and latest at Lingtai, possibly resulting from progressive, small-scale, equatorward C\u003csub\u003e4\u003c/sub\u003e expansion (Fig. 4c-e). As the MTG decreased significantly with warming across the MPB (Fig. 4h), the jet and consequently summer rainfall maxima migrated poleward, leading to rainier spring and C\u003csub\u003e4\u003c/sub\u003e retreat in northern China as evidenced by δ\u003csup\u003e13\u003c/sup\u003eC decreases at both Lingtai and Jiaxian (Fig. 4e-f). The jet shifted equatorward again during 5 − 4 Ma, resulting in C\u003csub\u003e4\u003c/sub\u003e expansion in G3 source regions and Jiaxian, whereas Lingtai was still dominated by C\u003csub\u003e3\u003c/sub\u003e plants (Fig. 4c-e). A consistent spatiotemporal pattern emerges in which warm season precipitation was the main limiting factor governing the multi-stage southward C\u003csub\u003e4\u003c/sub\u003e progression across East Asia, paced by the latitudinal migration of the westerly jet.\u003c/p\u003e\n\u003ch3\u003eObliquity-paced changes in the meridional insolation gradient\u003c/h3\u003e\n\u003cp\u003eAlthough the late Pliocene-Pleistocene evolution of the MTG, the westerly jet, and C\u003csub\u003e4\u003c/sub\u003e biomass can be explained by the NHG (51), their variations prior to the NHG require an additional mechanism. Orbital insolation directly impacts the MTG and consequently monsoon rainfall stages (52). The 41-ky obliquity cycle influences meridional insolation gradients by affecting high latitude climate as well as global ice volume (53) (Fig. 3a-b). Importantly, the 1.2-Myr cycle that corresponds to the amplitude modulation of obliquity (Fig. 3a) is thought to play a critical role in glaciation and sea-level changes (54). The MTG based on the compiled sea surface temperature records, is generally in pace with the 1.2-Myr cycle between 7 and 3 Ma (Fig. 4h).\u003c/p\u003e\n\u003cp\u003eOur T\u003csub\u003eΔ47\u003c/sub\u003e and δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e records show clear 1.2-Myr cycle, with three δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e peaks co-occurring roughly with maximum amplitudes of the 41-kyr obliquity cycle (Fig. 3c-d). During obliquity maxima at peak amplitudes, enhanced high-latitude summer insolation led to reduced meridional insolation gradient and MTG (Fig. 4h), thereby weakening the jet and pushing it further north, leading to an earlier onset of spring rain in northern China. This 1.2-Myr rhythm was also registered in other proxy-derived EAM records (55) (Fig. 3f-g), supporting its critical role in modulating the regional water cycle.\u003c/p\u003e\n\u003cp\u003eIt should be noted that the 1.2-Myr cycle controls the amplitude of the obliquity cycle (Fig. 3a), meaning both highest obliquity maxima and lowest minima occur at peak amplitudes. Our T\u003csub\u003eΔ47\u003c/sub\u003e and δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e records preserve only obliquity maxima signals due to overprinting. Specifically, during periods other than obliquity maxima, the jet was at a lower latitude and monsoonal rain was more concentrated during summer, when temperature and evaporation are relatively higher than spring, leading to less soil water penetration and shallower formation of pedogenic carbonates. Whereas during obliquity maxima, early onset of monsoonal rain, cooler temperature and less evaporation during spring allowed soil water to percolate deeper into soils, leaching previously formed shallower carbonates and resetting oxygen isotope compositions. In the northern Shilou and Jiaxian sections, the mean sedimentation rate was ~ 10 m/Ma (43), corresponding to 40 cm of sediment per 41-kyr obliquity cycle. The modern mean annual precipitation (MAP) across the CLP ranges between 250 and 650 ppm, which corresponds to carbonate accumulation depth of ~ 40–70 cm (56). Assuming MAP during the late Miocene-Pliocene was within this range, pedogenic carbonates formed during the previous obliquity maxima should be at ~ 80–110 cm depth at the next maxima, which would require extremely high MAP of more than ~ 700–1000 mm to dissolve them. Thus, these deepest-formed carbonates were selectively preserved, recording seasonal environmental conditions during the obliquity maxima.\u003c/p\u003e\n\u003cp\u003eThe MTG, δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e, and the 1.2-Myr cycle fell out of phase at 4.1 Ma, when the obliquity amplitude minima preceded the MTG maxima and δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e minima, and the following obliquity amplitude maxima led the subsequent MTG minima and δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e maxima (Fig. 4f and 4h). These offsets likely reflect the increasing influence of Northern Hemisphere cooling on the MTG, the jet position, and rainfall seasonality during the Plio-Pleistocene transition. The prevalence of ice-sheet-related slow physics over insolation-driven fast physics is supported by damped orbital-scale variability of both the East Asian winter and summer monsoon as well as their out-of-phase relationship starting 4.2 Ma (32).\u003c/p\u003e\n\u003cp\u003eTaken together, our nodule-based isotope records spanning 7-2.5 Ma mark the first empirical evidence supporting the jet transition hypothesis across orbital-to-tectonic timescales. Consequently, the jet transition offers a unified explanation of the multi-phase C\u003csub\u003e4\u003c/sub\u003e evolution across East Asia by modulating rainfall seasonality. This complex history of C\u003csub\u003e4\u003c/sub\u003e migration reflects the dynamic interplay between global climate forcings (i.e., ice sheets, insolation, jet dynamics) and regional factors (i.e., EAM, topography). Notably, paleolithic evidence suggests that hominins occupied the southern CLP since 2.12 Ma (18), favoring C\u003csub\u003e3\u003c/sub\u003e-dominated environments with abundant tree cover (19). The decrease in the number of artefacts found near Lantian after 1 Ma corresponds to major C\u003csub\u003e4\u003c/sub\u003e expansion (19), highlighting the central role of rainfall seasonality in hominin migration at higher latitudes through its influence on regional C\u003csub\u003e3\u003c/sub\u003e/C\u003csub\u003e4\u003c/sub\u003e abundance. Our findings provide a critical paleoclimatic framework for linking vegetation change, atmospheric circulation, and early hominin adaptation in East Asia.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eSampling\u003c/h2\u003e \u003cp\u003eThe CLP currently sits on the edge of the EAM region and is sensitive to monsoonal rainfall changes (\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e) (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003ea). The Miocene-Pliocene deposits on the CLP, known collectively as the Red Clay formation, record continuous paleoenvironmental information from 8 to 2.5\u0026nbsp;million years ago (Ma) (\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e). Bulk paleosol and pedogenic carbonate nodule samples were collected from three Red Clay sections: Lantian, Shilou, and Jiaxian. The 66-m thick Lantian section (34.190\u0026deg;N, 109.234\u0026deg;E, elevation\u0026thinsp;=\u0026thinsp;630 m) was located on the southern edge of the CLP, whereas the 58-m Jiaxian (38.272\u0026deg;N, 110.090\u0026deg;E, elevation\u0026thinsp;=\u0026thinsp;1025 m) and 71.4-m thick Shilou sections (36.926\u0026deg;N, 110.928 \u0026deg;E, elevation\u0026thinsp;=\u0026thinsp;1160 m) were situated in the northern CLP (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Modern mean annual precipitation (MAP) decreases from 720 mm in the southern Lantian to 529 mm and 433 mm in the northern Shilou and Jiaxian. Mean annual air temperatures (MAT) are 10.9\u0026deg;C, 10.0\u0026deg;C, and 14.6\u0026deg;C in Jiaxian, Shilou, and Lantian, respectively. Bulk paleosol samples were collected at 10 cm intervals, and at least two nodule samples were collected from each distinct depth. To avoid regolith contamination, soil profiles were trenched 1 m deep before sampling.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eAge model\u003c/h2\u003e \u003cp\u003eThe magnetic susceptibility data from each section were used to correlate to published age models (\u003cspan additionalcitationids=\"CR60\" citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e), which were established through linear interpolation using geomagnetic reversals as age control. Because the Red Clay formation is an aggregated soil profile, it is not possible to constrain the depth of carbonate nodule formation. We therefore use the depositional ages of the soil layer where the nodules were formed to represent the ages of nodules. Based on the modern MAP across the CLP and the MAP-Bk depth relationship (\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e), the depth of pedogenic carbonate formation is ~\u0026thinsp;40\u0026ndash;70 cm, which corresponds to ~\u0026thinsp;20\u0026ndash;50 ka offset given the sedimentation rate of the three Red Clay profiles.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eSample pretreatment\u003c/h3\u003e\n\u003cp\u003eWe conducted stable isotope analyses on calcite nodules (n\u0026thinsp;=\u0026thinsp;970) collected from the three Red Clay sections at Lantian, Shilou, and Jiaxian. At least two individual nodule samples from each stratigraphic layer were used for δ\u003csup\u003e18\u003c/sup\u003eO analyses. From those samples, a total of 32 and 22 nodule samples were selected for clumped isotope and triple oxygen isotope analyses, respectively. Prior to analyses, calcite nodule samples were either ultrasonically rinsed and then oven-dried at 50\u0026deg;C overnight or physically cleaned using a Dremel\u0026reg; tool to remove matrix sediments. Samples were then inspected for spars and, if present, only micrites were drilled to collect powder for isotopic analyses.\u003c/p\u003e\n\u003ch3\u003eStable oxygen isotope analysis\u003c/h3\u003e\n\u003cp\u003eSample splits of approximately 200 \u0026micro;g were loaded into exetainers and He-flushed before analysis. δ\u003csup\u003e18\u003c/sup\u003eO analyses were carried out separately in two different labs. For Shilou and Jiaxian samples, sample powders were acidified with 103% phosphoric acid for 2\u0026ndash;12 hours at 50\u0026deg;C. The stable isotopic compositions of released CO\u003csub\u003e2\u003c/sub\u003e were measured using a Thermo Gasbench II coupled to a continuous flow Thermo 253 isotope ratio mass spectrometer (IRMS) at the University of Texas at Austin. The isotopic values are reported in per mil (\u0026permil;) notation relative to Vienna Pee Dee Belemnite standard (VPDB) and were normalized to that scale using an in-house laboratory standard (UT Marble), NBS-18, and NBS-19. The external reproducibility of replicate analyses of standards averaged\u0026thinsp;\u0026plusmn;\u0026thinsp;0.07\u0026permil; for δ\u003csup\u003e18\u003c/sup\u003eO (1σ). For Lantian samples, the measurements were performed on a Finnigan DELTAplusXP IRMS attached to a Thermo GasBench II (75\u0026deg;C reaction with 100% phosphoric acid) at Nanjing University. The isotopic values were adjusted to VPDB scale using NBS-18, NBS-19, and an in-house laboratory standard (TTB-1). The standard error (1σ) of replicate analyses of the standard is \u0026plusmn;\u0026thinsp;0.04\u0026permil;.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eClumped isotope analysis\u003c/h2\u003e \u003cp\u003eClumped isotope measurements were carried out at the University of Washington Isolab, using two different instruments. Sampled materials are not subject to solid-state reordering of clumped isotope signals due to the shallow burial depth (\u0026lt;\u0026thinsp;100 m). The Jiaxian samples (n\u0026thinsp;=\u0026thinsp;18) were measured on a Thermo MAT 253 IRMS configured to measure m/z 44\u0026ndash;49 inclusive, following method described in Burgener et al. (2016). For each replicate analysis, sample aliquots containing 6\u0026ndash;8 mg of CaCO\u003csub\u003e3\u003c/sub\u003e were reacted in a common phosphoric acid bath at 90\u0026deg;C for 10 minutes. The CO\u003csub\u003e2\u003c/sub\u003e evolved from the acid digestion was purified cryogenically on an automated stainless steel/nickel vacuum line using a liquid N\u003csub\u003e2\u003c/sub\u003e trap, an ethanol-dry ice mixture (~ -80\u0026deg;C), and a Porapak Q trap (50/80 mesh, 122 cm long, 6.35 cm OD) at -20\u0026deg;C. The CO\u003csub\u003e2\u003c/sub\u003e was transferred through the Porapak using He carrier gas, isolated cryogenically, and flame-sealed into a Pyrex break seal before analysis. CO\u003csub\u003e2\u003c/sub\u003e was introduced into the IRMS using an automated 10-port tube cracker inlet system. Isotope ratios were converted to the VPDB scale using intra and interlaboratory standards calibrated against international standards NBS-18 and NBS-19. These include in-house calcites with dissimilar bulk composition (C64 and C2, which are both reagent grade; and a \u003cem\u003ePorites\u003c/em\u003e coral sample) and ETH standards (ETH 1\u0026ndash;4). For each four carbonate sample unknowns, a carbonate standard was digested, purified, and analyzed using the same procedure. All sample unknowns were analyzed in triplicate at minimum.\u003c/p\u003e \u003cp\u003eThe Lantian and Shilou samples (n\u0026thinsp;=\u0026thinsp;14) were analyzed using a fully automated Nu Instruments NuCarb carbonate device coupled to a Nu Perspective IRMS. Sample powders containing approximately 1.2\u0026ndash;1.5 mg calcite were loaded in individual glass vials before reaction with phosphoric acid at 70\u0026deg;C, purification using cryogenic and Porapak traps, and analysis. For each of the five carbonate sample unknowns, a carbonate standard was measured, including eight intra and interlaboratory standards (IAEA-C1, IAEA-C2, MERCK, Coral, ETH 1\u0026ndash;4). All sample unknowns were analyzed in triplicate at minimum.\u003c/p\u003e \u003cp\u003eΔ\u003csub\u003e47\u003c/sub\u003e values were calculated using established methods (\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e), with the exception that we used updated \u003csup\u003e17\u003c/sup\u003eO correction parameters (\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e), following recommendations of Da\u0026euml;ron et al. (2016) (\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e) and Schauer et al. (2016) (\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e). Δ\u003csub\u003e47\u003c/sub\u003e values analyzed using the Thermo MAT 253 IRMS were corrected to CDES-90\u0026deg;C acid values using CO\u003csub\u003e2\u003c/sub\u003e equilibrated at 4, 60 and 1000\u0026deg;C; Δ\u003csub\u003e47\u003c/sub\u003e values analyzed using the Nu Perspective were corrected to I-CDES (\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e) 90\u0026deg;C acid values using ETH1-4, IAEA-C1, IAEA-C2, and MERCK standards (\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e). The I-CDES and CDES scales are mathematically identical and interchangeable (\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e). Δ\u003csub\u003e47\u003c/sub\u003e uncertainties for individual samples were given as one standard error (1σ) of replicate analyses, calculated using the long-term standard deviation of carbonate standard analyses, or sample replicates, whichever is larger. Clumped isotope temperature (T\u003csub\u003eΔ47\u003c/sub\u003e) values reported here were calculated from mean Δ\u003csub\u003e47\u003c/sub\u003e values using the Andersen et al. (2021) calibration for samples digested at 90\u0026deg;C (no acid fractionation correction).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eTriple oxygen isotope analysis\u003c/h2\u003e \u003cp\u003eTriple oxygen isotope measurements were carried out at the University of New Mexico Center for Stable Isotopes using two different instruments. The Jiaxian samples (n\u0026thinsp;=\u0026thinsp;8) were measured on a Thermo Fisher MAT 253\u003csup\u003e+\u003c/sup\u003e isotope ratio mass spectrometer configured for triple oxygen isotope analyses. Sample powders containing at least 300 \u0026micro;g CaCO\u003csub\u003e3\u003c/sub\u003e were first converted to CO\u003csub\u003e2\u003c/sub\u003e using the phosphoric acid digestion system (25\u0026deg;C reaction with 100% phosphoric acid for 16 h). The resulting gas was purified using a cryogenic trap to remove water and any non-condensable gases. CO\u003csub\u003e2\u003c/sub\u003e was then fluorinated to produce O\u003csub\u003e2\u003c/sub\u003e to be measured for triple oxygen isotopes using the nickel bomb fluorination method modified by Wostbrock et al. (2020) (\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e). No replicate analysis was carried out.\u003c/p\u003e \u003cp\u003eLantian and Shilou samples (n\u0026thinsp;=\u0026thinsp;14) were measured using a TILDAS (Tunable Infrared Laser Direct Absorption Spectroscopy) CO\u003csub\u003e2\u003c/sub\u003e isotope monitor for Δ\u003csup\u003e'17\u003c/sup\u003eO - CO\u003csub\u003e2\u003c/sub\u003e designed and manufactured by Aerodyne Research, Inc. (ARI, Billerica, MA, USA), configured for triple oxygen isotope analyses of CO\u003csub\u003e2\u003c/sub\u003e. Detailed descriptions of the TILAS and measurement procedure are provided in Perdue et al. (2022) (\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e). Before the analyses, sample aliquots containing 0.5 mg CaCO\u003csub\u003e3\u003c/sub\u003e were loaded in exetainer vials sealed with a rubber septum. The vials were placed in a 70\u0026deg;C heating block and He-flushed with a flow rate of 11 m/s for 2 h. 5 ml of phosphoric acid was then injected into the vial and reacted with the sample powder for 8 h. The evolved CO\u003csub\u003e2\u003c/sub\u003e was cryogenically separated from water and purified before being introduced into the TILDAS analyzer. For analysis in the TILDAS system, CO\u003csub\u003e2\u003c/sub\u003e needs to be thoroughly mixed with CO\u003csub\u003e2\u003c/sub\u003e-free dry air in a two-bellow system, which produces a mixture of 456 ppm CO\u003csub\u003e2\u003c/sub\u003e. The isotope values of each sample measurement were determined relative to the reference gas by interpolating the prior and subsequent reference gas measurements in time. The total time required for 12 reference-sample cycles was 30 min. To monitor long-term analytical error and to place carbonate Δ\u003csup\u003e'17\u003c/sup\u003eO values on the VPDB scale, an interlaboratory standard (NBS-19, IAEA 603) was measured after four sample unknowns. The external reproducibility of replicate analyses of standards averaged 5 per meg for Δ\u003csup\u003e'17\u003c/sup\u003eO (1σ). All sample unknowns were analyzed in triplicate at minimum.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eCalculating the δ\u003csup\u003e18\u003c/sup\u003eO and Δ\u003csup\u003e'17\u003c/sup\u003eO of soil waters\u003c/h2\u003e \u003cp\u003eWith T\u003csub\u003eΔ47\u003c/sub\u003e, δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e, and Δ\u003csup\u003e\u003cem\u003e'\u003c/em\u003e17\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e available, δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e and Δ\u003csup\u003e\u003cem\u003e'\u003c/em\u003e17\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e can be calculated using the temperature-dependent fractionation between calcite and water (\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e, \u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e). Specifically, we first calculated δ\u003csup\u003e17\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e using δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e (\u0026permil;, VSMOW) and Δ\u003csup\u003e\u003cem\u003e'\u003c/em\u003e17\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e (per meg, VSMOW) (\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e):\u003c/p\u003e \u003cp\u003eδ\u003csup\u003e\u003cem\u003e'\u003c/em\u003e18\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e = 1000 \u0026times; ln(δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e / 1000\u0026thinsp;+\u0026thinsp;1)\u003c/p\u003e \u003cp\u003eδ\u003csup\u003e\u003cem\u003e'\u003c/em\u003e17\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;Δ\u003csup\u003e\u003cem\u003e'\u003c/em\u003e17\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e / 1000\u0026thinsp;+\u0026thinsp;0.528\u0026thinsp;\u0026times;\u0026thinsp;δ\u003csup\u003e\u003cem\u003e'\u003c/em\u003e18\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e\u003c/p\u003e \u003cp\u003eδ\u003csup\u003e17\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e = (exp(δ\u003csup\u003e\u003cem\u003e'\u003c/em\u003e17\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e / 1000) \u0026ndash; 1) \u0026times; 1000\u003c/p\u003e \u003cp\u003eNext, we calculated the temperature-dependent equilibrium isotope fractionation factor (\u003csup\u003e18/16\u003c/sup\u003e\u0026#120572;\u003csub\u003ecalcite\u0026minus;water\u003c/sub\u003e, \u003csup\u003e17/16\u003c/sup\u003e\u0026#120572;\u003csub\u003ecalcite\u0026minus;water\u003c/sub\u003e) between calcite and soil water (\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e, \u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e) using the clumped isotope temperature:\u003c/p\u003e \u003cp\u003e \u003csup\u003e18/16\u003c/sup\u003e\u0026#120572;\u003csub\u003ecalcite\u0026minus;water\u003c/sub\u003e = exp((16.1 \u0026times; 1000 / (273.15\u0026thinsp;+\u0026thinsp;T\u003csub\u003eΔ47\u003c/sub\u003e) \u0026ndash; 24.6) / 1000)\u003c/p\u003e \u003cp\u003e \u003csup\u003e17/16\u003c/sup\u003e\u0026#120572;\u003csub\u003ecalcite\u0026minus;water\u003c/sub\u003e = (\u003csup\u003e18/16\u003c/sup\u003e\u0026#120572;\u003csub\u003ecalcite\u0026minus;water\u003c/sub\u003e)\u003csup\u003eθ\u003c/sup\u003e\u003c/p\u003e \u003cp\u003eθ = -1.39 / (273.15\u0026thinsp;+\u0026thinsp;T\u003csub\u003eΔ47\u003c/sub\u003e)\u0026thinsp;+\u0026thinsp;0.5305\u003c/p\u003e \u003cp\u003eThe δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e (\u0026permil;, VSMOW) and Δ\u003csup\u003e\u003cem\u003e'\u003c/em\u003e17\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e (per meg, VSMOW) can then be calculated using the following equations:\u003c/p\u003e \u003cp\u003eδ\u003csup\u003e1x\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e = (δ\u003csup\u003e17\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e + 1000) / \u003csup\u003e1x/16\u003c/sup\u003e\u0026#120572;\u003csub\u003ecalcite\u0026minus;water\u003c/sub\u003e \u0026ndash; 1000\u003c/p\u003e \u003cp\u003eδ\u003csup\u003e\u003cem\u003e'\u003c/em\u003e1x\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e = 1000 \u0026times; ln(δ\u003csup\u003e1x\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e / 1000\u0026thinsp;+\u0026thinsp;1)\u003c/p\u003e \u003cp\u003eΔ\u003csup\u003e\u003cem\u003e'\u003c/em\u003e17\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e = 1000 \u0026times; (δ\u003csup\u003e\u003cem\u003e'\u003c/em\u003e17\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e \u0026minus;\u0026thinsp;0.528\u0026thinsp;\u0026times;\u0026thinsp;δ\u003csup\u003e\u003cem\u003e'\u003c/em\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e)\u003c/p\u003e \u003cp\u003eWhere the superscript 1x refers to either 18 or 17.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eReconstructing rainfall δ\u003csup\u003e18\u003c/sup\u003eO\u003c/h2\u003e \u003cp\u003eThe δ\u003csup\u003e18\u003c/sup\u003eO of pedogenic carbonates (δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e) record soil-water δ\u003csup\u003e18\u003c/sup\u003eO (δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e), with the formation temperature constrained by clumped isotopes (T\u003csub\u003eΔ47\u003c/sub\u003e, see \u003cem\u003eMethods\u003c/em\u003e). In well-drained soils, δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e inherits the δ\u003csup\u003e18\u003c/sup\u003eO of local rainfall (δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e), which is controlled by the regional water cycle and large-scale air circulation (\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e). However, δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e values are also affected by soil processes such as evaporation and mixing (\u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e74\u003c/span\u003e). The triple oxygen isotope (\u003csup\u003e16\u003c/sup\u003eO-\u003csup\u003e17\u003c/sup\u003eO-\u003csup\u003e18\u003c/sup\u003eO) compositions of pedogenic carbonates can help quantify the degree of evaporation experienced by soil waters from which the carbonates precipitated, thereby improving estimates of δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e). Δ\u0026prime;\u003csup\u003e17\u003c/sup\u003eO represents the deviation in δ\u0026prime;\u003csup\u003e17\u003c/sup\u003eO-δ\u0026prime;\u003csup\u003e18\u003c/sup\u003eO space (δ\u0026prime;=1000\u0026times;ln(δ/1000\u0026thinsp;+\u0026thinsp;1)) from the global meteoric water line (GMWL, Δ\u0026prime;\u003csup\u003e17\u003c/sup\u003eO\u0026thinsp;=\u0026thinsp;δ\u0026prime;\u003csup\u003e17\u003c/sup\u003eO-λ\u0026thinsp;\u0026times;\u0026thinsp;δ\u0026prime;\u003csup\u003e18\u003c/sup\u003eO, λ\u0026thinsp;=\u0026thinsp;0.528 (\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e)). Unlike δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e, Δ\u0026prime;\u003csup\u003e17\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e is not sensitive to processes involving equilibrium fractionation such as condensation and rainout from air masses that control the GMWL (\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e). Consequently, Δ\u0026prime;\u003csup\u003e17\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e values are primarily governed by evaporation-induced kinetic fractionation, which has a trajectory in δ\u0026prime;\u003csup\u003e17\u003c/sup\u003eO-δ\u0026prime;\u003csup\u003e18\u003c/sup\u003eO space shallower than the GMWL slope (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e). Measurements of modern and ancient terrestrial waters support the use of Δ\u0026prime;\u003csup\u003e17\u003c/sup\u003eO as a proxy of evaporation and thus regional aridity, with higher evaporation corresponding to lower Δ\u0026prime;\u003csup\u003e17\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e and δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e). Here we reconstructed δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e and Δ\u0026prime;\u003csup\u003e17\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e using δ\u003csup\u003e18\u003c/sup\u003eO, T\u003csub\u003eΔ47\u003c/sub\u003e, and Δ\u0026prime;\u003csup\u003e17\u003c/sup\u003eO of calcite nodules (Fig. S2, see \u003cem\u003eMethods\u003c/em\u003e). By leveraging Δ\u0026prime;\u003csup\u003e17\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e, we accounted for the effect of evaporation on δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003esw\u003c/sub\u003e and reconstructed δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eBayesian inversion\u003c/h2\u003e \u003cp\u003eAll the paleoclimate proxies are affected by multiple environmental parameters (\u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e). To quantitatively constrain environmental variables based on proxy data, here we applied Bayesian inversion to the steady-state soil water isotope model (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e), in which Markov Chain Monte Carlo methods were used to obtain posteriors of environmental parameters conditioned on the proxy data (δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e, Δ\u0026prime;\u003csup\u003e17\u003c/sup\u003eO\u003csub\u003ec\u003c/sub\u003e, Δ\u003csub\u003e47\u003c/sub\u003e) (\u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e77\u003c/span\u003e). The analyses were performed in R version 4.4.2 (\u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e78\u003c/span\u003e) using the rjags (\u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e79\u003c/span\u003e) and R2jags package (\u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e80\u003c/span\u003e). The prior ranges of all the related environmental inputs, including RH, δ\u003csup\u003e18\u003c/sup\u003eO\u003csub\u003ep\u003c/sub\u003e, carbonate formation depth, and evaporation rate, were set as uniform distributions of 10% \u0026minus;\u0026thinsp;90%, -25\u0026permil; - -5\u0026permil;, 0\u0026ndash;100 cm, and 10\u003csup\u003e\u0026minus;\u0026thinsp;10\u003c/sup\u003e \u0026ndash; 10\u003csup\u003e\u0026minus;\u0026thinsp;9\u003c/sup\u003e m/s, respectively, based on either theoretical constraints or modern observations. We ran five parallel chains, each consisting of 5\u0026times;10\u003csup\u003e5\u003c/sup\u003e iterations, with a burn-in of 1\u0026times;10\u003csup\u003e4\u003c/sup\u003e. Convergence was assessed by the Gelman and Rubin convergence factor (Rhat) and effective sample sizes reported by rjags (\u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e81\u003c/span\u003e). All the posteriors show strong convergence (Rhat\u0026thinsp;\u0026lt;\u0026thinsp;1.03, effective sample size\u0026thinsp;\u0026gt;\u0026thinsp;3000, Fig. S5).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eCESM2 experiments\u003c/h2\u003e \u003cp\u003eThe climate model simulations analyzed in this study are run with the Community Earth System Model version 2 (CESM2) (\u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e82\u003c/span\u003e). CESM2 incorporates Community Atmospheric Model version 6 (CAM6), Community Land Model version 5 (CLM5), Community Ice Code version 5 (CICE5), and Parallel Ocean Program version 2. The atmosphere and land components are run at a 0.9\u0026deg; x 1.25\u0026deg; horizontal resolution, while the ocean and sea ice components are run with a nominal 1\u0026deg; resolution. The pre-industrial simulation included in this study is part of the Climate Model Intercomparison Project phase 6 (CMIP6) Tier 1 experiment (\u003cem\u003epiControl.001\u003c/em\u003e) (\u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e82\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe mid-Pliocene simulations (\u003cem\u003emidPliocene-eoi400\u003c/em\u003e) analyzed in this study have been reported in Feng et al. (2020) (\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e), targeted at the mid-Piacenzian Warm Period (MPWP, 3.205 Ma). The MPWP boundary conditions follow the protocol of the second phase of the PlioMIP (PlioMIP2) (\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e), including present-day orbital configuration, continental greening, a reduced Greenland Ice Sheet, a deglaciated western Antarctica, and adjusted topography and coastlines (\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e, \u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e83\u003c/span\u003e). Notably, the Bering and Canadian Arctic Archipelago straits are closed, and the Sunda, Sahul, Baltic Shelves, and Hudson Bay are exposed.\u003c/p\u003e \u003cp\u003eIn addition, to assess the climate changes induced by changes in CO\u003csub\u003e2\u003c/sub\u003e during the Pliocene, we report a new fully coupled mid-Pliocene simulation where the same MPWP protocol is used except that the CO\u003csub\u003e2\u003c/sub\u003e concentration is reduced to the pre-industrial level at 280 ppm (\u003cem\u003emidPliocene-eoi280).\u003c/em\u003e\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eRecalculating SST records\u003c/h2\u003e \u003cp\u003eTo further investigate the evolution of the MTG, we collated high-quality \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{U}}_{\\text{37}}^{\\text{k\u0026#039;}}\\)\u003c/span\u003e\u003c/span\u003e-based SST records around the global ocean (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e). These SST records were divided into three groups based on their latitudes: high-latitude\u0026thinsp;\u0026gt;\u0026thinsp;50\u0026deg;N regions (ODP 883/884, 907, 982), mid-latitude regions (20\u0026ndash;50\u0026deg;N: ODP 1010, 1021, and 1208), and tropics (\u0026lt;\u0026thinsp;20\u0026deg;N: ODP 722, 846, 850, U1338, 1241) (Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). To account for the seasonality of alkenones and nonlinear response of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{U}}_{\\text{37}}^{\\text{k\u0026#039;}}\\)\u003c/span\u003e\u003c/span\u003e to temperature, we recalculated the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{U}}_{\\text{37}}^{\\text{k\u0026#039;}}\\)\u003c/span\u003e\u003c/span\u003e-based SSTs by applying the original \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\text{U}}_{\\text{37}}^{\\text{k\u0026#039;}}\\)\u003c/span\u003e\u003c/span\u003e data to a Bayesian B-spline regression model, BAYSPLINE (\u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e84\u003c/span\u003e). The integrated SSTs of each group were computed by binning the SSTs from each OSP site to 100-kyr windows and averaging the estimates. The SST differences between the tropics and mid-latitude regions and the tropics and high-latitude regions were used to infer the MTG.\u003c/p\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors acknowledge the aid of Andrew Schauer and Jeff Cullen in supervising clumped isotope and stable isotope analyses. \u0026nbsp;This work was supported by the United States National Science Foundation (1545859 to D.O.B. and 1156134 to K.H.) and the National Natural Science Foundation of China (42030503 to J.F. and 42021001 to H.L.). R.F. would like to acknowledge NSF-2103055 and NSF-2238875 for supporting the production of simulations used in this study.\u003c/p\u003e\n"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eSteinthorsdottir M et al (2021) The Miocene: The Future of the Past. \u003cem\u003ePaleoceanography and Paleoclimatology\u003c/em\u003e 36, e2020PA004037\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePeppe DJ et al (2023) Oldest evidence of abundant C4 grasses and habitat heterogeneity in eastern Africa. Science 380:173\u0026ndash;177\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQuade J, Cerling TE, Bowman JR (1989) Development of Asian monsoon revealed by marked ecological shift during the latest Miocene in northern Pakistan. Nature 342:163\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAndrae JW et al (2018) Initial expansion of C4 vegetation in Australia during the late Pliocene. Geophys Res Lett 45:4831\u0026ndash;4840\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTHE CENOZOIC CO2 PROXY INTEGRATION PROJECT (CENCO2PIP) CONSORTIUM (2023) Toward a Cenozoic history of atmospheric CO2. Science 382:eadi5177\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuo X et al (2024) Mapping the global distribution of C4 vegetation using observations and optimality theory. Nat Commun 15:1219\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLu J et al (2019) The Early Pliocene global expansion of C4 grasslands: A new organic carbon-isotopic dataset from the north China plain. Palaeogeogr Palaeoclimatol Palaeoecol 109454. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.palaeo.2019.109454\u003c/span\u003e\u003cspan address=\"10.1016/j.palaeo.2019.109454\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShen X et al (2018) Increased seasonality and aridity drove the C4 plant expansion in Central Asia since the Miocene\u0026ndash;Pliocene boundary. Earth Planet Sci Lett 502:74\u0026ndash;83\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGallagher TM et al (2021) Regional patterns in Miocene-Pliocene aridity across the Chinese Loess Plateau revealed by high resolution records of paleosol carbonate and occluded organic matter. Paleoceanography and Paleoclimatology n/a. e2021PA004344\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou B et al (2014) Late Pliocene\u0026ndash;Pleistocene expansion of C4 vegetation in semiarid East Asia linked to increased burning. Geology 42:1067\u0026ndash;1070\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchiemann R, L\u0026uuml;thi D, Sch\u0026auml;r C (2009) Seasonality and Interannual Variability of the Westerly Jet in the Tibetan Plateau Region. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/2008JCLI2625.1\u003c/span\u003e\u003cspan address=\"10.1175/2008JCLI2625.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang H et al (2018) East Asian hydroclimate modulated by the position of the westerlies during Termination I. Science 362:580\u0026ndash;583\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePassey BH et al (2009) Strengthened East Asian summer monsoons during a period of high-latitude warmth? Isotopic evidence from Mio-Pliocene fossil mammals and soil carbonates from northern China. Earth Planet Sci Lett 277:443\u0026ndash;452\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEdwards EJ, Osborne CP, Str\u0026ouml;mberg CA, Smith SA, C4 Grasses Consortium (2010) The origins of C4 grasslands: integrating evolutionary and ecosystem science. Science 328:587\u0026ndash;591\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDa J, Zhang YG, Li G, Ji J (2020) Aridity-driven decoupling of δ13C between pedogenic carbonate and soil organic matter. Geology. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1130/G47241.1\u003c/span\u003e\u003cspan address=\"10.1130/G47241.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLisiecki LE, Raymo ME (2005) A Plio-Pleistocene stack of 57 globally distributed benthic δ 18 O records. Paleoceanography 20:1\u0026ndash;17\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHerbert TD et al (2016) Late Miocene global cooling and the rise of modern ecosystems. Nat Geosci 9\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhu Z et al (2018) Hominin occupation of the Chinese Loess Plateau since about 2.1 million years ago. Nature 559:608\u0026ndash;612\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou B et al (2023) Hominin Response to Oscillations in Climate and Local Environments During the Mid-Pleistocene Climate Transition in Northern China. \u003cem\u003eGeophysical Research Letters\u003c/em\u003e 50, e2023GL104931\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCerling TE et al (2011) Woody cover and hominin environments in the past 6 million years. Nature 476:51\u0026ndash;56\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChiang JCH et al (2015) Role of seasonal transitions and westerly jets in East Asian paleoclimate. Q Sci Rev 108:111\u0026ndash;129\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDing Y, Chan JCL (2005) The East Asian summer monsoon: an overview. Meteorol Atmos Phys 89:117\u0026ndash;142\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYang S et al (2018) A strengthened East Asian Summer Monsoon during Pliocene warmth: Evidence from \u0026lsquo;red clay\u0026rsquo; sediments at Pianguan, northern China. J Asian Earth Sci 155:124\u0026ndash;133\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKelson JR et al (2023) Triple oxygen isotope compositions of globally distributed soil carbonates record widespread evaporation of soil waters. Geochim Cosmochim Acta. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.gca.2023.06.034\u003c/span\u003e\u003cspan address=\"10.1016/j.gca.2023.06.034\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTian C, Wang L, Tian F, Zhao S, Jiao W (2019) Spatial and temporal variations of tap water 17O-excess in China. Geochim Cosmochim Acta 260:1\u0026ndash;14\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCheng H et al (2009) Ice Age Terminations Science 326:248\u0026ndash;252\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChiang JCH, Herman MJ, Yoshimura K, Fung IY (2020) Enriched East Asian oxygen isotope of precipitation indicates reduced summer seasonality in regional climate and westerlies. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e 117, 14745\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAbell JT, Winckler G, Anderson RF, Herbert TD (2021) Poleward and weakened westerlies during Pliocene warmth. Nature 589:70\u0026ndash;75\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSon J-H, Seo K-H, Wang B (2020) How Does the Tibetan Plateau Dynamically Affect Downstream Monsoon Precipitation? \u003cem\u003eGeophysical Research Letters\u003c/em\u003e 47, e2020GL090543\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHe L et al (2024) Northward Extension of East Asian Summer Monsoon Since the Miocene Set by the Uplift of Tibetan Plateau. \u003cem\u003eGeophysical Research Letters\u003c/em\u003e 51, e2023GL107262\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLu H et al (2022) Decoupled Land and Ocean Temperature Trends in the Early-Middle Pleistocene. Geophys Res Lett 49:e2022GL099520\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun Y, An Z, Clemens SC, Bloemendal J, Vandenberghe J (2010) Seven million years of wind and precipitation variability on the Chinese Loess Plateau. Earth Planet Sci Lett 297:525\u0026ndash;535\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKelson JR et al (2020) A proxy for all seasons? A synthesis of clumped isotope data from Holocene soil carbonates. Q Sci Rev 234:106259\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDa J, Li GK, Ji J (2023) Seasonal changes in the formation time of pedogenic carbonates on the Chinese Loess Plateau during Quaternary glacial cycles. Q Sci Rev 305:108008\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBreecker DO, Sharp ZD, McFadden LD (2009) Seasonal bias in the formation and stable isotopic composition of pedogenic carbonate in modern soils from central New Mexico, USA. Geol Soc Am Bull 121:630\u0026ndash;640\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhong Y et al (2022) Humidification of Central Asia and equatorward shifts of westerly winds since the late Pliocene. Commun Earth Environ 3:1\u0026ndash;9\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHaywood A et al (2016) The Pliocene model intercomparison project (PlioMIP) phase 2: scientific objectives and experimental design. Clim Past 12:663\u0026ndash;675\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFeng R, Otto-Bliesner BL, Brady EC, Rosenbloom N (2019) Increased Climate Response and Earth System Sensitivity From CCSM4 to CESM2 in Mid-Pliocene Simulations. \u003cem\u003eJournal of Advances in Modeling Earth Systems\u003c/em\u003e 12, eMS002033 (2020)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHe L, Zhou T, Chen X, Zuo M, Zou L (2024) Earlier seasonal march of the East Asian summer monsoon in the mid-Pliocene. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1175/JCLI-D-23-0709.1\u003c/span\u003e\u003cspan address=\"10.1175/JCLI-D-23-0709.1\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang X, Nie J, Stevens T, Zhang H, Xiao W (2022) Resolving conflicting models of late Miocene East Asian summer monsoon intensity recorded in Red Clay deposits on the Chinese Loess Plateau. Earth Sci Rev 104200. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.earscirev.2022.104200\u003c/span\u003e\u003cspan address=\"10.1016/j.earscirev.2022.104200\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAhmed IAM, Maher BA (2018) Identification and paleoclimatic significance of magnetite nanoparticles in soils. Proc Natl Acad Sci U S A 115:1736\u0026ndash;1741\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFan J, Jones SB, Qi LB, Wang QJ, Huang MB (2012) Effects of precipitation pulses on water and carbon dioxide fluxes in two semiarid ecosystems: measurement and modeling. Environ Earth Sci 67:2315\u0026ndash;2324\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAo H et al (2016) Late Miocene\u0026ndash;Pliocene Asian monsoon intensification linked to Antarctic ice-sheet growth. Earth Planet Sci Lett 444:75\u0026ndash;87\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCowie RH (1984) The life-cycle and productivity of the land snail Theba pisana (Mollusca: Helicidae). J Anim Ecol 311\u0026ndash;325\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDeng L, Jia G, Jin C, Li S (2016) Warm season bias of branched GDGT temperature estimates causes underestimation of altitudinal lapse rate. Org Geochem 96:11\u0026ndash;17\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang H et al (2024) New calibration of terrestrial brGDGT paleothermometer deconvolves distinct temperature responses of two isomer sets. Earth Planet Sci Lett 626:118497\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eClemens SC, Prell WL, Sun Y, Liu Z, Chen G (2008) Southern Hemisphere forcing of Pliocene δ18O and the evolution of Indo-Asian monsoons. Paleoceanography 23\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWen Y et al (2023) CO2-forced Late Miocene cooling and ecosystem reorganizations in East Asia. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e 120, e2214655120\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAn Z et al (2005) Multiple expansions of C4 plant biomass in East Asia since 7 Ma coupled with strengthened monsoon circulation. Geology 33:705\u0026ndash;708\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun J, L\u0026uuml; T, Zhang Z, Wang X, Liu W (2012) Stepwise expansions of C4 biomass and enhanced seasonal precipitation and regional aridity during the Quaternary on the southern Chinese Loess Plateau. Q Sci Rev 34:57\u0026ndash;65\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLong H et al (2025) The westerlies-monsoon interaction shaped asymmetric lake expansions over the Tibetan Plateau in warming periods. Sci Bull. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.scib.2025.07.023\u003c/span\u003e\u003cspan address=\"10.1016/j.scib.2025.07.023\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKong W, Swenson LM, Chiang JCH (2017) Seasonal Transitions and the Westerly Jet in the Holocene East Asian Summer Monsoon. J Clim 30:3343\u0026ndash;3365\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuybers P, Wunsch C (2005) Obliquity pacing of the late Pleistocene glacial terminations. Nature 434:491\u0026ndash;494\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi M, Hinnov LA, Huang C, Ogg JG (2018) Sedimentary noise and sea levels linked to land\u0026ndash;ocean water exchange and obliquity forcing. Nat Commun 9:1004\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQin J et al (2022) 1.2 Myr Band of Earth-Mars Obliquity Modulation on the Evolution of Cold Late Miocene to Warm Early Pliocene Climate. \u003cem\u003eJournal of Geophysical Research: Solid Earth\u003c/em\u003e 127, e2022JB024131\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRetallack GJ (2005) Pedogenic carbonate proxies for amount and seasonality of precipitation in paleosols. Geology 33:333\u0026ndash;336\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAn Z et al (2014) Late Cenozoic climate change in monsoon-arid Asia and global changes in. Late Cenozoic Climate Change in Asia. Springer, pp 491\u0026ndash;581\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiu X et al (2001) Magnetic properties of the Tertiary red clay from Gansu. Sci China Ser D-Earth Sci 44:635\u0026ndash;651\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun D, Liu D, Chen M, An Z, John S (1997) Magnetostratigraphy and palaeoclimate of Red Clay sequences from Chinese Loess Plateau. Sci China Ser D-Earth Sci 40:337\u0026ndash;343\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eQiang XK, Li Z-X, Powell CM, Zheng HB (2001) Magnetostratigraphic record of the Late Miocene onset of the East Asian monsoon, and Pliocene uplift of northern Tibet. Earth Planet Sci Lett 187:83\u0026ndash;93\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAo H et al (2021) Global warming-induced Asian hydrological climate transition across the Miocene\u0026ndash;Pliocene boundary. Nat Commun 12:6935\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRetallack GJ (2009) Refining a pedogenic-carbonate CO2 paleobarometer to quantify a middle Miocene greenhouse spike. Palaeogeogr Palaeoclimatol Palaeoecol 281:57\u0026ndash;65\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuntington KW et al (2009) Methods and limitations of \u0026lsquo;clumped\u0026rsquo; CO2 isotope (∆47) analysis by gas-source isotope ratio mass spectrometry. J Mass Spectrom 44:1318\u0026ndash;1329\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrand WA, Assonov SS, Coplen TB (2010) Correction for the 17O interference in δ(13C) measurements when analyzing CO2 with stable isotope mass spectrometry (IUPAC Technical Report). \u003cem\u003ePure and Applied Chemistry\u003c/em\u003e 82, 1719\u0026ndash;1733\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDa\u0026euml;ron M, Blamart D, Peral M, Affek HP (2016) Absolute isotopic abundance ratios and the accuracy of ∆47 measurements. Chem Geol 442:83\u0026ndash;96\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchauer AJ, Kelson J, Saenger C, Huntington KW (2016) Choice of 17O correction affects clumped isotope (∆47) values of CO2 measured with mass spectrometry. Rapid Commun Mass Spectrom 30:2607\u0026ndash;2616\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDennis KJ, Affek HP, Passey BH, Schrag DP, Eiler JM (2011) Defining an absolute reference frame for \u0026lsquo;clumped\u0026rsquo; isotope studies of CO2. Geochim Cosmochim Acta 75:7117\u0026ndash;7131\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBernasconi SM et al (2021) InterCarb: A Community Effort to Improve Interlaboratory Standardization of the Carbonate Clumped Isotope Thermometer Using Carbonate Standards. \u003cem\u003eGeochemistry, Geophysics, Geosystems\u003c/em\u003e 22, e2020GC009588\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWostbrock JAG, Cano EJ, Sharp ZD (2020) An internally consistent triple oxygen isotope calibration of standards for silicates, carbonates and air relative to VSMOW2 and SLAP2. Chem Geol 533:119432\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePerdue N, Sharp Z, Nelson D, Wehr R, Dyroff C (2022) A rapid high-precision analytical method for triple oxygen isotope analysis of CO2 gas using tunable infrared laser direct absorption spectroscopy. Rapid Commun Mass Spectrom 36:e9391\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTremaine DM, Froelich PN, Wang Y (2011) Speleothem calcite farmed in situ: Modern calibration of δ18O and δ13C paleoclimate proxies in a continuously-monitored natural cave system. Geochim Cosmochim Acta 75:4929\u0026ndash;4950\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWostbrock JAG et al (2020) Calibration of carbonate-water triple oxygen isotope fractionation: Seeing through diagenesis in ancient carbonates. Geochim Cosmochim Acta 288:369\u0026ndash;388\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuz B, Barkan E (2010) Variations of 17O/16O and 18O/16O in meteoric waters. Geochim Cosmochim Acta 74:6276\u0026ndash;6286\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFischer-Femal BJ, Bowen GJ (2021) Coupled carbon and oxygen isotope model for pedogenic carbonates. Geochim Cosmochim Acta 294:126\u0026ndash;144\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAron PG et al (2021) Triple oxygen isotopes in the water cycle. Chem Geol 565:120026\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBowen GJ, Fischer-Femal B, Reichart G-J, Sluijs A, Lear CH (2020) Joint inversion of proxy system models to reconstruct paleoenvironmental time series from heterogeneous data. Clim Past 16:65\u0026ndash;78\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLunn D, Jackson C, Best N, Thomas A, Spiegelhalter D (2013) The BUGS book. A practical introduction to Bayesian analysis. Chapman Hall, London\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCore Team R (2024) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePlummer M (2018) rjags: Bayesian graphical models using MCMC (R package version 4\u0026ndash;6, 2016). \u003cem\u003eAvailable at\u003c/em\u003e: \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://CRAN.Rproject.org/package= rjags\u003c/span\u003e\u003cspan address=\"https://CRAN.Rproject.org/package= rjags\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSu Y-S, Yajima M, Su MY-S, SystemRequirements J (2015) Package \u0026lsquo;r2jags.\u0026rsquo; \u003cem\u003eR package version 0.03-08, URL\u003c/em\u003e \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://CRAN\u003c/span\u003e\u003cspan address=\"http://CRAN\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. \u003cem\u003eR-project\u003c/em\u003e. \u003cem\u003eorg/package\u0026thinsp;=\u0026thinsp;R2jags\u003c/em\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGelman A, Rubin DB (1992) Inference from Iterative Simulation Using Multiple Sequences. Stat Sci 7:457\u0026ndash;472\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDanabasoglu G et al (2019) The Community Earth System Model Version 2 (CESM2). \u003cem\u003eJournal of Advances in Modeling Earth Systems\u003c/em\u003e 12, eMS001916 (2020)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDowsett H et al (2016) The PRISM4 (mid-Piacenzian) paleoenvironmental reconstruction. Clim Past 12:1519\u0026ndash;1538\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTierney JE, Tingley MP (2018) BAYSPLINE: A New Calibration for the Alkenone Paleothermometer. Paleoceanography Paleoclimatology 33:281\u0026ndash;301\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8290760/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8290760/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe Neogene expansion of C\u003csub\u003e4\u003c/sub\u003e grasslands transformed terrestrial ecosystems with marked influence on mammalian evolution, including hominins. However, the asynchronous C\u003csub\u003e4\u003c/sub\u003e expansion on different continents makes it difficult to identify the environmental drivers, especially for higher latitudes. Here we show that rainfall seasonality governed extratropical Plio-Pleistocene C\u003csub\u003e4\u003c/sub\u003e distributions in East Asia. Rainfall oxygen isotope ratios and clumped isotope soil temperatures exhibit coupled variations on the Chinese Loess Plateau (CLP) from 7 to 2.5 Ma, indicating more spring rain during warmer times when the subtropical westerly jet was further poleward, and more concentrated summer rain under cooler climates. We attribute these changes to meridional shifts of a summer rain band on orbital and longer timescales. Organic carbon isotope records reveal that the most C\u003csub\u003e4\u003c/sub\u003e-rich ecosystems tracked this summer rain band, eventually eclipsing the southern CLP margin during late Pleistocene cooling. Our model refines the East Asian paleomonsoon concept and explains the equatorward migration of extratropical C\u003csub\u003e4\u003c/sub\u003e ecosystems, highlighting the tight coupling between regional rainfall seasonality and vegetation.\u003c/p\u003e","manuscriptTitle":"Jet-Induced Rainfall Seasonality and C4 Expansion over East Asia","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-01-08 07:34:45","doi":"10.21203/rs.3.rs-8290760/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"8b0587e1-e4c6-4b47-9f75-930ccf7f73ec","owner":[],"postedDate":"January 8th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":60793070,"name":"Earth and environmental sciences/Climate sciences/Palaeoclimate"},{"id":60793071,"name":"Earth and environmental sciences/Hydrology"},{"id":60793072,"name":"Earth and environmental sciences/Ecology/Palaeoecology"}],"tags":[],"updatedAt":"2026-05-07T20:15:08+00:00","versionOfRecord":[],"versionCreatedAt":"2026-01-08 07:34:45","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8290760","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8290760","identity":"rs-8290760","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}