Content uploaded by Yan Ping
Author content
All content in this area was uploaded by Yan Ping on Nov 30, 2024
Content may be subject to copyright.
https://doi.org/10.1177/09596836241285782
The Holocene
1 –14
© The Author(s) 2024
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/09596836241285782
journals.sagepub.com/home/hol
Introduction
Aeolian-fluvial processes play a crucial role in shaping landforms
(Belnap et al., 2011; Bullard and Livingstone, 2002). The coexis-
tence of aeolian and fluvial interbedded deposits in arid environ-
ments indicates a shared sediment origin, intertwined spatial
distribution, and interconnected temporal evolution of aeolian and
fluvial sands (Langford and Chan, 1989). River controls the pat-
tern of desert distribution by providing sediment and accommoda-
tion, and in return, desert distribution and aeolian sand activities
restrict river channel development and sediment transportation on
different temporal-spatial scales (Draut, 2012; Yan et al., 2015).
Understanding the response of aeolian-fluvial interaction sedi-
mentary systems to environmental changes in arid and semi-arid
regions is challenging, especially when examining preserved
sedimentary sequences. The aeolian-fluvial interaction arises
from both short-term, localized interactions resulting from com-
petitive sedimentary processes, as well as long-term responses to
non-periodic controlling factors such as climate, sediment supply
from external basins, and basin tectonics (Nanson et al., 1995).
Throughout geological history, wind-dominated and fluvial-dom-
inated periods alternate in response to climate change (Clarke and
Rendell, 1998). At centennial timescales, this alternation is mani-
fested as variations in the strength of aeolian-fluvial interaction
sedimentation and transitions in sedimentary modes with chang-
ing wet and dry conditions (Gao and Tang, 1997). On an annual
scale, wind and fluvial dominance alternate with the seasons.
Spatially, aeolian-fluvial sedimentary landforms exhibit an inter-
laced distribution pattern, ranging from small-scale interactions
between individual sand dunes and seasonal floods to intermedi-
ate-scale interactions between aeolian landforms and fluvial-
lacustrine systems, and even large-scale global aeolian-fluvial
Holocene aeolian-fluvial interactions
patterns and their response to climate
changes in the Paiku Co basin, Southern
Tibetan Plateau
Wenjie Yuan,1,2 Yinghua Zheng,1,3 Ping Yan,1,3 Zhiyong Ding,4
Xiaoxu Wang,1,2 Xiao Zhang,1,2 Yating Gao,1 Jiaqi Chu,1,2
Hua Fan1,2 and Yingna Liu1
Abstract
Aeolian-fluvial processes are crucial in shaping landforms. Despite the advancements in understanding the sedimentary processes of aeolian-fluvial
interactions, the sedimentary patterns of these interactions are complex in spatiotemporal scale, and the records of Holocene small-scale aeolian-fluvial
interaction sedimentary is insufficient, especially in the Tibetan Plateau region. Therefore, this study focuses on analyzing the sedimentary patterns of
aeolian-fluvial interactions and their response to climate change in the Paiku Co basin of the southern Tibetan Plateau (TP). Through the examination of
aeolian-fluvial sequences, Optically Stimulated Luminescence (OSL) dating is utilized to establish a chronological framework. Furthermore, proxy indices
are explored to identify potential sediment models and their reactions to climate change. The findings indicate that fluvial activity predominated during
a relatively warm and wet period around 5.1–4.6 ka, while thicker aeolian sands accumulated extensively during 4.6–4.1 ka as the climate transitioned to
cold and dry conditions. On a millennial scale, the aeolian-fluvial interaction sedimentation is characterized by alternating deposition of clay layers and
medium to fine sand. This sedimentation pattern is predominantly influenced by climate change. Overall, the findings shed light on the complex coupling
between aeolian and fluvial processes in response to past climate changes, which has important implications for understanding the landscape evolution
and environmental changes in this sensitive high-altitude region.
Keywords
aeolian and fluvial interaction, climate change, mid-Holocene, Paiku Co, sediments model
Received 14 June 2024; revised manuscript accepted 4 September 2024
1 State Key Laboratory of Earth Surface Processes and Resource Ecology,
Beijing Normal University, China
2 Engineering Center of Desertification and Blown-Sand Control of
Ministry of Education, Faculty of Geographical Science, Beijing Normal
University, China
3
Zhuhai Branch of State Key Laboratory of Earth Surface Processes and
Resource Ecology, Beijing Normal University, China
4 State Key Laboratory of Geohazard Prevention and Geoenvironment
Protection, Chengdu University of Technology, China
Corresponding author:
Ping Yan, State Key Laboratory of Earth Surface Processes and
Resource Ecology, Beijing Normal University, No. 19 Xinjikouwai
Street, Haidian District, Beijing 100875, China.
Email: yping@bnu.edu.cn
1285782HOL0010.1177/09596836241285782The HoloceneYuan et al.
research-article2024
Research Paper
2 The Holocene 00(0)
interaction in different climatic zones (Thomas et al., 1993).
Thus, aeolian-fluvial interaction sedimentation displays distinct
distribution characteristics in various regions and at different time
scales, illustrating a pattern of mutual succession in the temporal
sequence.
At different time scales, most research on aeolian-fluvial inter-
action sedimentation has mainly focused on the Last Glacial
period. From a global perspective, the occurrence of aeolian-flu-
vial interaction sedimentation periods varies across different
regions. For instance, taking into account for different age mod-
els, in the southern margin of the Tarim Desert, southeastern Ara-
bia, the foreland of the Andes, and during Marine Isotope Stage 3
(MIS) (55-41 ka), river activity was dominant, followed by aeo-
lian-fluvial interaction sedimentation from 41 to 26 ka, and a
period of intense aeolian activity from 24 to 12 ka (Mueller et al.,
2023). The aeolian-dominated periods correspond to the Younger
Dryas and Heinrich events (Robins et al., 2022). In western and
central Europe, aeolian-fluvial interaction sedimentation occurred
from 28 to 18 ka, followed by an aeolian-fluvial interaction period
from 18 to 10 ka (Kasse, 2002; Kasse et al., 2007). Besides, the
records of aeolian-fluvial interaction sedimentation in the Mu Us
Desert and Qaidam Basin in China mainly focus on the LGM,
particularly the response of aeolian sand to cold events, such as at
the 20.1–18.4 ka and 14.8–12.6 ka (the Bølling/Allerød intergla-
cial period), and aeolian sand activity mainly occurred during
Heinrich Stadial 2, Heinrich Stadial 1, and the Younger Dryas
(Ding et al., 2023; Yu et al., 2015). However, at the Holocene
scale, especially on a smaller scale of centimeters to meters in the
interlayer sedimentation of aeolian-fluvial mixed sediments, the
sedimentation patterns of aeolian-fluvial interactions and their
response patterns to climate change are still underreported.
The nature and intensity of aeolian-fluvial interaction are pri-
marily influenced by the water flow conditions and changes in
sediment supply and availability (Bullard and McTainsh, 2003).
In general, sediment generation is usually associated with wet
periods, while sediment becomes available for aeolian systems
during dry phases. Several studies have shown that intense aeo-
lian activity periods may not align with the occurrence of glacial
period or cold conditions (Kasse, 2002). So, it should be noted
that sediment supply and storage do not always respond solely to
arid periods (Bullard and Livingstone, 2002). For example,
increased aeolian activity occurred after climate warming in the
ice sheet environments (Van Huissteden et al., 2000). In coastal
environments, aeolian sand transport and dune formation are sig-
nificantly enhanced during cold periods with low sea levels, and
temperature facilitates the occurrence of aeolian-fluvial interac-
tions by altering precipitation patterns and vegetation cover
(Pomar et al., 2018). In a braided river plain landscape, changes in
aeolian sand storage/supply respond to the shift from semi-arid to
arid climates (Simpson et al., 2008). Source bordering dunes are
considered to have formed during periods of strong river activity
in the Cooper Creek basin, Lake Eyre Basin, central Australia
(Cohen et al., 2010). These examples highlight the complexity
and non-linear nature of aeolian-fluvial interactions and their
response to climate changes. This requires further analysis of
additional regional climatic characteristics and geomorphic envi-
ronments to properly understand the response mechanisms of
aeolian-fluvial interactions to climate change.
The TP, known as the “Water Tower of Asia” is the largest
plateau in the world. Its unique geographic location, topography,
and climatic conditions make it an important research area for
studying aeolian-fluvial interaction (Immerzeel et al., 2020). The
Paiku Co region, located in the southern of the TP, exhibits inter-
actions between glaciers, rivers, lakes, and surrounding land-
forms, making it an ideal region for investigating aeolian-fluvial
interactions. The region experiences a cold and arid climate,
characterized by strong winds, limited precipitation, and signifi-
cant temperature variations. These environmental conditions
play a crucial role in shaping the landforms and sediment dynam-
ics in the area.
Understanding aeolian-fluvial interactions is important for
deciphering past environmental changes, predicting future land-
scape dynamics, and managing natural resources in dryland
regions, particularly during the Holocene. Further research is
needed to improve our knowledge of these interactions. Based on
the above, this study focuses on the Paiku Co Basin as the research
area to analyze and investigate the aeolian-fluvial interaction
sedimentation characteristics and their response to climate change
in the Paiku Co basin during the Holocene. Furthermore, the
study also provides insights into the sedimentation patterns of
aeolian-fluvial interaction for the area. So, the chronologic frame-
work of the aeolian-fluvial sequence was established using OSL
dating. Multiple environmental proxies were used to explore the
response of aeolian-fluvial interaction sedimentation to climate
change and sedimentation patterns. The results enhance the
understanding of the aeolian-fluvial interaction sedimentation
model during the Holocene on the TP and the formation and land-
forms evolution in arid regions.
Geological setting
The Paiku Co basin (28°25′–29°07′N; 85°20′–86°05′E) is located
in the southern Tibetan valley between the Himalayas and the
northern Kailas Range and is a late Cenozoic fault basin (Regional
Geology of Xizang (Tibet) Autonomous Region (1993)(Figure 1).
This basin is distributed in the N–S direction as a “dumbbell”
Figure 1. Study area. (a) Shows the location of Paiku Co basin on the TP. (b) Shows the schematic diagram of the Paiku Co basin and
sampling site.
Yuan et al. 3
shape, that is, wide in the north and south and narrow in the mid-
dle, approximately 130 km long and 20–45 km wide and is located
in a plateau area of 4000–4500 m altitude (Zhu et al., 2008). The
basin is underlain by Jurassic tuffs, sandstones, low-grade meta-
morphic rocks, granites and the unconformity sedimentation of
early Pleistocene-Holocene lacustrine sedimentary strata, which
are nearly tens of meters thick. Both the westerlies in winter and
the Indian summer monsoon (ISM) in summer dominate the cli-
mate in the Paiku Co basin (Li et al., 2011). The annual average
precipitation is 150–200 mm, and the annual average temperature
is −1.4°C (Hu et al., 2014; Lei et al., 2021). The Daqu, Laqu and
Barixiongqu Rivers are there large rivers serving as the main sup-
ply of the Paiku Co basin and the first two are mainly from glacier
meltwater (Lei et al., 2018).
Material and methods
Sampling
The aeolian-fluvial interaction sequence section Daqu (DQ) (alti-
tude of 4640 m, 85.63741E, 28.76181N) has developed in the
three terraces of the Paiku Co shore. The DQ section is located in
about 3.5 km south of Paiku Co. The DQ sample is acquired on
the second terraces west of the Daqu River, south of the G219
highway (Figure 2a–c). The whole slope inclination is 20°, and
the height is 35 m; it was sampled in the middle of the slope (15 m
from the slope of the bottom). The total sampling profile height is
195 cm. About 54 samples were collected at 1–5 cm intervals for
surface texture of quartz sand and grain size measurement. In the
DQ section, nine OSL samples were further collected to establish
the chronology.
Luminescence dating
OSL sample preparation was conducted under low-intensity red
light at the State Key Laboratory of Earth Surface Processes and
Resource Ecology of Beijing Normal University, Beijing. The
coarse grains (90–125 μm) of the samples were wet-sieved and
then treated with 10% HCl and 30% H2O2 to remove carbonate
and organic material, respectively. The grains were extracted
using heavy liquids with densities of 2.58, 2.62, and 2.75 g/cm3.
The pure quartz and K-feldspar grains were extracted by etching
for 40 min with 40% HF and 10% HF to remove alpha-irradiated
layers. Finally, all samples were treated with 10% HCl for 20 min
to remove fluoride precipitates.
OSL sample measurement was conducted using an automated
Risø-TL/OSL-DA-20 reader in the luminescence laboratory,
Institute of Geology, China Earth Quake Administration. Labora-
tory irradiation was carried out using 90Sr/90Y beta sources, and
the stimulated power was set as the 90% of the maximum stimu-
lated power. The K-feldspar IRSL signal was detected through a
combination of Corning 7–59 and Schott BG-39 filters. A two-
step pIRIR dating protocol (Buylaert et al., 2011; Li et al., 2015;
Thiel et al., 2011) was used to determine the De of the samples
(Table 1).
The radioactive element concentrations of uranium (U), tho-
rium (Th), and potassium (K) in the samples were measured on
inductively coupled plasma‒mass spectrometry. Based on the
altitudes, the burial depths of the samples can be used to calculate
the cosmic ray contributions (Prescott and Hutton, 1994). The
water content was unified at 10% ± 5% to calculate the dose rate.
The estimation of the internal dose rate of the K-feldspar sample
was performed under the assumption of the 12.5% ± 0.5% K con-
tent and the 400 ± 100 ppm Rb content (Huntley and Baril, 1997;
Huntley and Hancock, 2001; Zhao and Li, 2005).
Proxy indices analysis
The experiments were conducted at the State Key Laboratory of
Earth Surface Processes and Resource Ecology of Beijing Nor-
mal University, Beijing. The colors were obtained through a
comprehensive comparison of dry and wet samples against the
international standard chart (M50215B). The grain size were
measured using a Malvern Mastersizer 3000 laser diffraction
particle size analyzer with a measurement range of 0.01–
3500 µm. Four grain-size parameters, namely mean grain size
(Mz), the sorting index (σ), skewness (SK1), and kurtosis (KG),
were calculated using the methods described by Folk (1966). The
sediment grain size was classified into seven classes according to
Figure 2. The details of DQ profile geographic locations. (a)
Shows the location of the DQ profile and Daqu river on Google
Earth. (b and c) Show the environmental conditions around the DQ
profile. (d) Shows the lithostratigraphic units and OSL ages of the
DQ profile.
Table 1. A two-step pIRIR dating protocol (Buylaert et al., 2011).
Step pIRIR dating protocol Result
1 Given dose Di
2 Preheat at 200°C for 60 s
3 IRSL measurement at 50°C for 200 s Lx(50)
4 IRSL measurement at 170°C for 200 s Lx(170)
5 Given test dose Dt
6 Preheat at 200°C for the 60 s
7 IRSL measurement at 50°C for 200 s Tx(50)
8 IRSL measurement at 170°C for 200 s Tx(170)
9 IRSL bleaching at 280°C for 100 s
10 Return to step 1
4 The Holocene 00(0)
a widely used international grain size classification system Blott
and Pye (2012). These classes include clay (0–2 µm), silt (2–
63 μm), very fine sand (63–125 μm), fine sand (125–250 μm),
medium sand (250–500 μm), coarse sand (500–1000 μm), and
very coarse sand (1000–2000 μm). This classification system
was employed as the basis for analyzing the grain size distribu-
tion (GSD) characteristics in the DQ profile. For convenience,
grain size can be measured using either φ or diameter (d). The
conversion formulas between the two are as follows: φ = −log2d.
The unit of d is millimeters (mm). The low-frequency magnetic
susceptibility (χlf) and CaCO3 content of the DQ samples were
measured by a Bartington MS2B Meter at 0.47 kHz and the
Eijkelkamp 0853 tester, respectively. For surface texture obser-
vation of quartz sand, quartz grains sized between 0.125 and 0.5
mm were selected as main object. About 20–25 quartz particles
were blindly chosen for observation and photography using an
S-4800 cold field emission.
End-member modeling analysis
From a statistical perspective, grain size data can be separated
into several independent Weibull distributions, with each distribu-
tion representing a distinct depositional process (Sun et al., 2002;
Xiao et al., 2012). End-member modeling analysis (EMMA) of
grain size is effective to clarify the sediment transport processes
and their environmental significance (Dietze et al., 2012; Weltje,
1997). Therefore, to analyze the sediment grain size data from the
Daqu profile, the curve was fitted using the parametric Weibull
distribution method. The AnalySize software, programed by Pat-
erson and Heslop (2015), was utilized in MATLAB for end-mem-
ber (EM) analysis. In line with theoretical principles, a high R2
value and a low θ value are considered optimal for statistical fit-
ting (Weltje, 1997). For data confidence, the minimum required
number of EMs is obtained by these criteria (Paterson and Hes-
lop, 2015). The relative proportions of each EM were then calcu-
lated for each sample.
Results
Lithology
The detailed sediment characteristics of the DQ profile are
described in Table 2. Based on the differences in sediment char-
acteristics, the DQ profile can be categorized into aeolian depos-
its, fluvial deposits, and aeolian-fluvial mixture sediments
(Figures 2d and 4). These sediment layers are relatively thin,
each with a thickness less than 50 cm. The entire DQ profile of
Munsell color tones ranges from 2.5Y to 10YR, characterized
by sediment alternating between horizontally laminated clay
and medium to fine sand, occasionally containing small pebbles
and iron rust spots. The sediment colors in the DQ profile transi-
tion from light yellowish brown (2.5Y6/3) to light gray
(2.5Y7/2), and from dark yellowish brown (10YR3/4) to yel-
lowish (10YR8/6). It consists primarily of regularly laminated
clay and fine sand, with varying proportions of silt, fine sand,
and clay. These sediments likely represent a mixture of aeolian
and fluvial environments. Several aeolian deposits, mainly rang-
ing in colors from light yellowish brown (2.5Y6/4) to yellow
brown (10YR5/4) to yellow (10YR7/6), consisting of loose fine
sand, occur at depths of 0–5 cm, 58–60 cm, 75–80 cm, 146–
151 cm, and 179–182 cm respectively. River deposits, ranging in
colors from grayish brown (10YR5/2) to very pale brown
(10YR7/4), are mainly characterized by clay enriched with
small pebbles and rust spots. These deposits are recorded at
depths of 85–88 cm, 93–98 cm, 123–125 cm, 144–146 cm, 151–
154 cm, 158–161 cm, and 162–165 cm. They likely represent
flood deposits formed by rapid glacial melting from the Daqu
River (Fouinat et al., 2017; Yanchilina et al., 2019).
OSL chronology
OSL signal of quartz in the DQ section is dim, with photon counts
of around 40–50, close to the background value, and thus does not
meet the conditions for dating (Figure 3a). Therefore, this study
uses K-feldspar to establish a chronology for the DQ section.
A preheat temperature plateau test of the K-feldspar sample
DQ-4 was conducted following the pIR50IR170 dating procedure in
Table 1 (Buylaert et al., 2009; Li et al., 2015). The results indicate
that the pIRIR De value remains constant despite of the varying
preheat and pIRIR temperature. The recycling ratio of all aliquots
fall in 0.9–1.1, and the recuperation for different preheat tempera-
tures are all less than 5% of the natural signal (Figure 3c). Thus,
pIR50IR170 dating protocol is reliable, and the pIRIR signal could
be considered as stable.
After bleaching for 8 h with a solar simulator (Honle UVAC-
UBE 400) lamp, the DQ-4 sample were measured for the dose
recovery plateau and residual dose test (Buylaert et al., 2012; Rei-
mann et al., 2011). A known dose of 24.25 Gy was given to three
bleached aliquots and was measured following the pIR50IR170 dat-
ing protocol in Table 1. As illustrated in Figure 3d, the calculation
results of the measured given dose ratios for the pIRIR170 signal
are between 0.9 and 1.1. The subtracting residual dose for the
pIRIR170 signal is 1.2 ± 0.04 Gy and could be ignored for the DQ
samples. Therefore, the pIR50IR170 dating protocol is suitable for
De determination in this study.
The decay curves indicate that the pIR50IR170 signals are bright
enough for De measurements. The growth curves of the pIRIR170
and IR50 signals were fitted by a single saturation exponential
function (Figure 3b). The average recycling ratio and the natural
signal recuperation of all K-feldspar aliquots are 0.992 ± 0.03 and
less than 2%, respectively (Figure 3e). Together, these results
prove that this pIR50IR170 measurement protocol is reliable. More-
over, all data points of the IR50 De to pIR50IR170 De plot fall within
±10% of the fitted function (Figure 3f). This implies that the
K-feldspar samples from the DQ section had been well bleached
(Buylaert et al., 2013; Thiel et al., 2011). As a result, a central age
model was used to calculate the pIRIR De for each K-feldspar
sample. The age calculation results of the pIR50IR170 of K-feld-
spar samples are shown in Table 3.
The variations of grain size, CaCO3 content, and χlf
The grain size within the DQ profile is related to the lithology
changes. The range of Mz is from 53.7 to 260.48 μm, with an aver-
age of 154.52 μm. The highest Mz occurred in aeolian sand is
216.94 μm, and the flood layers of Mz is lowest with 80.73 μm. In
the DQ profile, the grain size distributions suggested the presence
of clay, silt, and sand (Figures 4 and 5b). The dominant component
is sand (48.36%–95.46%, average is 77.88%), while the amount of
the clay was the least, ranging from 0% to 5.58%, with an average
of 2.05%. In the aeolian sand layer, the σ ranging from 0.7 to 0.9,
and the grain size distribution was bimodal, predominantly con-
sisting of sand with an average of 93.58%. For mixed layers, the σ
ranging from 1.2 to 2.0, and the Mz was 166.82 μm. Due to varying
proportions of clay and silt, GSDs exhibited bimodal-multimodal
distribution characteristics, but sand still dominated in this layer.
In flood layers, the σ greater than 2, and clay and silt content
accounted for the highest proportions throughout the profile, aver-
aging 3.82% and 34.84% respectively, with GSDs displaying a
multi-peak distribution (Figures 4, 5b, and c). The probability
cumulative curves for the aeolian-fluvial sediment sequence in the
Paiku Co basin exhibit a three-stage pattern. The dominant salta-
tion components occur at approximately 50%–80% (Figure 4). For
the flood sediment, the inflection points of the peristalsis, salta-
tion, and suspension, components are at 0.8 φ and 2.4 φ (189 and
574 μm), respectively. Specifically, the suspension components
are smaller than 189 μm, the saltation components range from 189
Yuan et al. 5
to 574 μm, with the suspension components, accounting for 60%
and the saltation components accounting for 36%. In the corre-
sponding aeolian sediment, the dominant saltation components
range from 1.2 φ to 3.2 φ (109–435 μm), accounting for 80%,
while the suspension components are smaller than 5%. The mixed
layers exhibit evidence of both wind and water transportation, sug-
gesting simultaneous deposition by these two agents. The domi-
nant component is saltation, accounting for 80% and 60%, with
saltation particle sizes ranging from 1.2 to 3 φ, 1.2 to 3.5 φ (125–
435 μm, 88–435 μm).
The χlf values in the DQ profile were relatively low, ranging
from 2.7 to 8.2 × 10−8 m3/kg, averaged at 5.3 × 10−8 m3/kg (Figure
5d). The lowest χlf value (4.1-4.5 × 10−8 m3/kg) occurred in aeo-
lian sediment, with an average of 4.2 × 10−8 m3/kg, while in the
flood sediment (χlf is 4.8 × 10−8 m3/kg) was slightly higher than
that in aeolian sediment and the average χlf value for mixed layers
was 4.8 × 10−8 m3/kg.
The CaCO3 content in the DQ profile is also generally low,
ranging from 5.2% to 11.2%, with an average content of 7.4%
(Figure 5e). Specifically, the content in aeolian sand layers is con-
centrated in the range of 5.2%–6.2%, while the CaCO3 content in
flood sediments falls within the range of 6.6%–8.1%. The CaCO3
content in mixed layers ranges from 6.1% to 11.2%, slightly
higher than that in aeolian sand layers.
The texture characteristic of quartz surface
This study involved photographing and observing the quartz
grains of aeolian layers, mixed layers, and flood layers in each
layer of the DQ profile, followed by summarizing the morpho-
logical characteristics of quartz grains in these layers. Figure 6
illustrates the typical quartz morphologies found in aeolian lay-
ers, mixed layers, and flood layers.
The surface texture analysis of quartz sand measurements
indicates that grains in the flood layers predominantly exhibit an
irregular outline, with rounded vertices showing mechanically
impacted V-shapes and triangular depressions. Additionally, they
display conchoidal fractures and slight scratches, suggesting
Table 2. The detailed sediment characteristics of DQ profile.
Sedimentary unit Depth (cm) Sedimentary characteristics
Aeolian deposit 0–5 Light yellow (2.5Y7/6), fine sand mixed with modern sand visible, loose, visible red stripes.
Mixed layer 5–15 Light yellowish brown (2.5Y6/3), fine sand, loose, mixed with brown humus.
15–22 Pale brown (10YR6/3), medium-fine sand, tighter than the upper layer, with small lenticles visible and humus.
22–25 Dark yellowish brown (10YR3/4), medium-fine sand, tight, with rust-like oxidized layer, greater clay content.
25–31 Yellowish brown (10YR5/6), fine sand, loose, with lenticles of clay (long is 20 cm and thick is 3 cm) interspersed
inside.
31–35 Dark yellowish brown (10YR3/4), clay, tight. The layer above and below this one consists of red sand.
35–43 Light brownish gray (10YR6/2), fine sand, loose. At 41–43 cm is clay layer.
43–58 Light yellowish brown (2.5Y6/3), fine sand, loose.
Aeolian deposit 58–60 Light yellowish brown (10YR6/4), fine sand, loose.
Mixed layer 60–75 Pale brown (10YR6/3), mainly fine sand, loose. A few oxidation spots are visible at 60–75 cm, where there is a
0.5 cm thick thin layer of medium-fine sand at 71 cm.
Aeolian deposit 75–80 Light yellowish brown (2.5Y6/4), fine sand, loose, interspersed with 1 cm thick clay layer.
Mixed layer 80–88 Light gray (2.5Y7/2), mainly clay, tight, with light yellowish brown (10YR6/4), coarse sand at 84–85 cm, loose.
88–93 Light yellowish brown (2.5Y6/4), fine sand, loose, with 0.1–0.2 cm of medium sand at 89 cm.
93–98 Pale brown (10YR6/3), clay, tight, with a few gravels, gravel size 3–5 cm.
98–100 Pale brown (2.5Y7/4), fine sands, loose, interspersed with 1 cm thick gray (10YR5/1) clay.
100–105 Light gray (2.5Y7/2), fine sand, loose. The lower part of this layer is tight.
105–110 Pale brown (10YR6/3), clay and coarse sand alternate in deposition, tight, with iron nodules at 106–107 cm.
110–116 Light yellowish brown (10Y6/4), medium-fine sand, relatively tight, mixed with gravel approximately 1 cm in
diameter.
116–120 Dark gray (10YR4/1), mainly fine sand, relatively tight, mixed with brownish yellow (10YR6/6) of medium-fine
sand at 117–118 cm.
120–123 Pale brown (10YR6/3), medium-fine sand, relatively tight.
Flood sediment 123–125 Light gray (10YR7/2), clayey silt, tight, occasionally with 1 cm diameter gravel, and calcium spots and rust spots.
Mixed layer 125–130 Brownish yellow (10YR6/6) medium-fine sand and yellowish grayish brown (10YR5/2) clay interbedded, tight.
Flood sediment 130–134 Very pale brown (10YR7/4), mainly clay, interspersed with a small amount of sand, tight at the top and loose at
the bottom, with rust spots, calcium spots and occasionally 1 cm gravels.
Mixed layer 134–136 Yellowish (10YR8/6), medium-fine sand, interspersed with thin layers of dark gray (10YR4/1) clay, loose, contain-
ing gravels of 0.3–0.5 mm in diameter
136–138 Yellow (10YR7/6), fine sand, finer in texture than the upper layer, loose.
138–140 Very pale brown (10YR7/3), fine sand mixed with silt, and clayey silt, tight.
140–144 Grayish brown (10YR5/2), sandy clay, tight, with glassy lustrous minerals.
Flood sediment 144–146 Grayish brown (10YR5/2), silt clay, tight, with oxidized layers and calcium spots
Aeolian deposit 146–151 Yellow brown (10YR5/4), fine sand, loose
Flood sediment 151–154 Very pale brown (10YR7/4), clay mixed with fine sand, tight, with red rust spots and oxidation processes.
Mixed layer 154–158 Very pale brown (10YR7/3), fine sand, interspersed with thin layers of dark gray (10YR4/1) clay, loose.
Flood sediment 158–161 Pale brown (10YR6/3), clay, tight.
Mixed layer 161–162 Brownish yellow (10YR6/6), medium-fine sand, loose.
Flood sediment 162–169 Grayish brown (10YR5/2), clay, tight, occasionally 1 cm diameter gravels with rusty spots.
Mixed layer 169–179 Dark grayish brown (10YR4/2), clay and fine sand interbedded, tight, with red oxide layer.
Aeolian deposit 179–182 Yellow (10YR7/6), fine sand, loose.
Mixed layer 182–195 Brownish yellow (10YR6/6), clay and medium-fine sand interbedded, tight, with iron nodules and gravels of
0.5–1.5 cm in diameter.
6 The Holocene 00(0)
formation primarily in glaciofluvial environments (Figure 6a–d).
Notion that slight silica precipitation on the surfaces, and the sur-
faces with high roundness appear to have dissolution grooves and
pits, indicating that slow runoff deposition generated by increased
glacial meltwater under a relatively high-temperature environ-
ment (Wang, 1985; Xie, 1984).
The quartz grains in the aeolian layer exhibit two main mor-
phologies. One type maintains an irregular shape with high round-
ness, displaying conchoidal fractures, bamboo conchoidal fractures,
dish-shaped depressions, and upturned plates, as well as a mechani-
cal breakage surface with step-like features (Figure 6e–h). These
features reflect characteristics of wind transport over short dis-
tances. The other type is oval shape with high roundness, a frosted
surface, and dish-shaped depressions, indicating wind transport
over long distances.
Some quartz grains from the mixed layers exhibit high round-
ness, with dish-shaped depressions and pocks overlaying trian-
gular pits, conchoidal fractures, parallel steps, and scratches
(Figure 6i–l). This suggests that these grains underwent fluvial
transport initially and were subsequently deposited by aeolian
processes. On the other hand, another set of quartz grains show
high roundness with V-shaped and triangular pits, superimposed
on a frosted surface, indicating they first experienced aeolian
transport and then fluvial action. The series of aeolian transport
and fluvial action may show climate changes.
End-member modeling of grain size
The grain-size EM results are illustrated in Figure 5f–i. Assuming
that the number of EMs is 1–10, as the number of EMs increases,
Figure 3. OSL result of DQ profile. (a) Shows decay curve and the inset image shows the growth curve for the quartz signal of DQ-4. (b)
Shows decay curve and the inset image shows the growth curve for K-feldspar pIRIR signal of DQ-4. (c) Shows the dose recovery test results
of the solar simulator bleached for K-feldspar sample DQ-4. (d) Shows the preheat temperature test results for K-feldspar sample DQ-4.
(e) Shows the recycling ratios and recuperation of all aliquots for the pIR50IR170 De measurements. (f ) Shows a plot of IR50 De versus their
pIR50IR170 De values for all samples from DQ sections. The data points were fitted by a single saturating exponential function: ya ebx
1
(solid lines). The dashed lines are the solid line multiplying 0.9 and 1.1, respectively (Buylaert et al., 2013).
Yuan et al. 7
Table 3. K-feldspar pIR50IR170 ages from the DQ section.
Sample
ID
Depth (m) Aliquots U (ppm) Th (ppm) K (%) Cosmic dose
rate (Gy/ka)
Dose rate
(Gy/ka)
OD of IR (%) OD of pIRIR (%) IR De (Gy) pIRIR De (Gy) IR Age (ka) pIRIR
Age (ka)
DQ-1 0.18 14 2.34 ± 0.13 9.40 ± 0.63 3.74 ± 0.01 0.49 ± 0.05 5.33 ± 0.21 0 ± 0 0 ± 0 10.17 ± 0.13 20.29 ± 0.18 1.84 ± 0.06 4.10 ± 0.21
DQ-2 0.59 13 2.6 ± 0.09 10.63 ± 0.51 3.69 ± 0.01 0.4 ± 0.04 5.68 ± 0.17 5.43 ± 1.43 5.19 ± 1.33 7.33 ± 0.13 23.54 ± 0.39 1.29 ± 0.04 4.42 ± 0.15
DQ-3 0.76 14 2.26 ± 0.01 9.17 ± 0.13 3.97 ± 0.01 0.39 ± 0.04 5.74 ± 0.16 9.26 ± 1.9 7.22 ± 1.23 6.72 ± 0.12 23.79 ± 0.51 1.17 ± 0.04 4.48 ± 0.16
DQ-4 0.90 13 3.01 ± 0.1 11.89 ± 0.15 3.61 ± 0.03 0.37 ± 0.04 5.56 ± 0.22 3.73 ± 1.21 0.27 ± 5.96 8.36 ± 0.11 24.08 ± 0.19 1.45 ± 0.05 4.58 ± 0.20
DQ-5 1.02 14 3.25 ± 0.07 10.87 ± 0.34 3.51 ± 0.04 0.36 ± 0.04 5.66 ± 0.17 0 ± 0 3.66 ± 1.23 9.74 ± 0.08 24.39 ± 0.33 1.79 ± 0.07 4.57 ± 0.21
DQ-6 1.39 15 3.04 ± 0.28 12.8 ± 0.84 3.80 ± 0.05 0.31 ± 0.03 5.75 ± 0.24 3.63 ± 1.61 1.97 ± 1.53 7.71 ± 0.11 26.37 ± 0.28 1.34 ± 0.06 4.79 ± 0.23
DQ-7 1.50 14 1.93 ± 0.1 7.09 ± 0.35 3.5 ± 0.03 0.33 ± 0.03 5 ± 0.15 3.12 ± 1.14 3.74 ± 1.1 7.93 ± 0.09 23.07 ± 0.29 1.65 ± 0.07 4.90 ± 0.22
DQ-8 1.73 14 2.75 ± 0 9.57 ± 0 3.46 ± 0.02 0.32 ± 0.03 5.15 ± 0.2 6.57 ± 1.56 6.57 ± 0.16 7.81 ± 0.15 23.66 ± 0.28 1.52 ± 0.07 4.85 ± 0.27
DQ-9 180 14 2.1 ± 0.06 8.04 ± 0.23 3.33 ± 0.04 0.34 ± 0.03 4.96 ± 0.15 6.66 ± 1.46 0 ± 0 8.36 ± 0.16 23.88 ± 0.18 1.75 ± 0.08 5.12 ± 0.22
the chosen number of EMs could be better when following the
constraints: angular deviation θ ⩽ 5, linear correlation R2 ⩾ 0.8,
and low EM correlation value. When the number of EMs reaches
5, the variance of interpretation reaches 97.8%, the R2 value
reaches 0.994, the θ value decreases to 4.1, and the EM correla-
tion value is 0.153, indicating that the function is properly fitted
and is representative of the original data. it can be concluded that
the description of five mixed EMs for the samples from the Daqu
section is reasonable. The grain-size EM analysis of the DQ pro-
file reveals strong correlations between each EM and lithological
changes, while correlations between different EMs are weak. This
suggests that the EM analysis effectively distinguishes various
dynamic sedimentary processes within the sediments. The EMs of
the grain size distribution for all sections are shown in Figure 5h.
Each EM corresponds to a single-peaked normal distribution. The
grain sizes of the EMs are 21.2, 98.1, 185.8, and 309.5 μm, and
454 μm. The contents of EM1, EM2, EM3, EM4, and EM5 ranged
from 2.37% to 57.82%, 0% to 51.67%, 0% to 46.19%, 6.59% to
66.71%, and 0% to 52.84%, respectively.
Discussion
Evidences of aeolian-fluvial sediment
In the Paiku Co Basin, multiple glacial streams rivers meander
and flow into lakes. Additionally, sand patches can be observed
within the braided channels and riverbanks of the Daqu river (Fig-
ure 2a and b). These landscapes can provide the most direct evi-
dence for the occurrence of small scale fluvial-aeolian interactions.
Geologically, this fluvial-aeolian interaction could be manifested
as alternating deposits between river and aeolian sediments.
In the Daqu profile, the aeolian layers light yellow fine sand.
These layers exhibit a σ ranging from 0.7 to 0.9, loose texture
(Figure 5c). The clay content (<2 μm) in the aeolian layers is less
than 0.1%, with the predominant fraction being medium and fine
sand, suggesting good sorting. The GSD is positively skewed,
with primary modal sizes around ~310 μm. Additionally, these
characteristics are consistent with aeolian sand found in the Yar-
lung Tsangpo River (Zheng et al., 2009; Zhou et al., 2021).
Besides, high roundness, frosted surface and dish-shaped depres-
sions of quartz surface texture for aeolian layers, suggest a wind-
induced genesis. Consequently, it is inferred that the aeolian
layers represent aeolian sand. Given that the thickness of these
deposits ranges from approximately 3 to 10 cm, it is speculated
that they are thin sand patches deposited along the riverbanks.
In flood layers, clay and silt increases rapidly to 5.58% and
46.7%, and the sediment color notably darkens. The sediment is
consisted of light gray-black sandy clay, compact, and consisted of
red rust spots, small amounts of gravel with gravel sizes ranging
from 3 to 5 cm. The grain size distribution exhibits a multimodal
distribution with peak sizes at 10–11, 50–70, and 300–453 μm.
However, it shows poor sorting with a σ greater than 2. The Sk1 of
grain size distribution is less than 1, and the GSD shows the pres-
ence of fine-grained tails (Figures 4 and 5c). These characteristics
all indicate the result of either flood deposition or quiet-water
deposition (Folk and Ward, 1957). Besides, the V-shape, triangular
depression, slight silica precipitation, dissolution grooves, pits,
and visible red rust spots in the quartz microstructure of this layer
all suggest that it was formed in a river environment and under
relatively high-temperature conditions (Wang, 1985; Xie, 1984).
The mixed layers are predominantly characterized by alternat-
ing deposition of gray-black clay and yellow medium sand, along
with a mixture of gray-white and yellow fine sand, and abundant
rust spots at 187–195, 165–169, 105–110 cm. Additionally, the
thin thickness of clay and medium sand layers, at 1 cm, indicates a
aeolian-fluvial interaction deposition environment (Robins et al.,
2022; Yu et al., 2022), attributed to different sedimentary environ-
ments and processes. Thin clay layers typically form due to short
8 The Holocene 00(0)
Figure 5. The DQ profile stratigraphy and measured proxies tendency along depth. (a) Is simplified stratigraphy. (b) Is mean grain size. (c) Is
sorting index. (d) Is magnetic susceptibility (χlf ). (e) Is CaCO3. (f) Is the linear correlation between R2. (g) Is the angular deviation θ of the EMs.
(h) Is distribution of different EMs. (i) Is the variation of the content of each EM with depth.
Figure 4. Grain size distribution and probability cumulative curves from the Daqu profile.
periods of slow-moving or stagnant water flow, allowing fine par-
ticles to settle, while yellow medium sand layers may result from
wind environments (Robins et al., 2023). Overall, the mixed layers
exhibit a mixture of grain sizes, poor sorting, σ ranging from 1.2 to
2.0, loose (Figure 5c). The GSD curve presents a complex pattern,
with the majority showing a bimodal distribution and a minority
showing a multimodal distribution. The primary peak and second-
ary peak model sizes of the bimodal distribution are 210–250 μm
and 70–90 μm, respectively. The primary peak model sizes are
similar to aeolian sand layers, while the multimodal distribution
with a primary peak at 300–350 μm is similar to flood layer’s pri-
mary peak. The bimodal-multimodal distribution of mixed layers
suggests a more complex material source. The fine-grained por-
tion aligns with the grain size typical of low-energy fluvial fine-
grained deposits. (Roskin et al., 2014). Furthermore, the CaCO3
content and χlf values of the mixed layers also exhibit complexity,
with values ranging from 2. 7% to 8.2% and 4.4% to 13.2%,
encompassing both aeolian sand layers and flood layers.
Yuan et al. 9
Importantly, the surface of quartz shows features such as butterfly
pits, pitting superimposed on triangular pits, shell fractures above
V-shaped pits, triangular pits, and superimposed frost surfaces.
Further confirming that this layer represents a typical aeolian-flu-
vial mixed deposition (Wang, 1985; Xie, 1984).
Environmental significance of proxies and grain size
EMs for the aeolian-fluvial interaction sequence
In sediment, the content of χlf primarily is widely accepted as the
indicator to the precipitation changes as it depends on the concen-
tration of ferrimagnetic minerals. A high χlf value indicates strong
precipitation, while a low χlf value suggests the opposite (Fang
et al., 1999; Yang et al., 2015). The carbonate content in sedi-
ments can be categorized into primary and secondary types, with
secondary carbonates being predominantly influenced by varia-
tions in the climate environment. This includes the dissolution of
carbonate rocks and the precipitation of carbonates caused by
rainfall. It can serve as an important indicator for paleoclimate
studies (An et al., 2012; Dar et al., 2015). In the DQ profile, both
the magnetic susceptibility and the overall carbonate content are
relatively low, with high values occurring in flood layers and rela-
tively low values in aeolian sand layers. This may be related to the
relatively low temperature and precipitation in the Paiku Co
basin. Additionally, the presence of silica precipitation on the
quartz surface of the flood layers suggests that magnetic suscepti-
bility and calcium carbonate can indicate a relatively warm and
humid climate environment in this study.
According to 14C dating of shore terraces, the lake level eleva-
tion of the Paiku Co basin was 4635 m at about 12 ka (Wünnemann
et al., 2015). At ~6 ka, the lake level declined to 4610 m, while the
lake level was 4640 m when the Daqu profile formed at 5.1–4.1 ka.
The Daqu profile was formed by interactive deposition from river
and wind, with a vague relationship with lake deposition. There-
fore, the clay deposition of the Daqu profile was likely formed by
river spreading over the river floodplain during the flooding period.
The grain size composition of EM1 is dominated silt (2–
63 μm), with a modal size of 21.2 μm, which is the finest EM in
the profile. The GSD of EM1 is a positively skewed flat single-
peaked distribution with poor sorting, corresponding to the grain
size characteristics of river deposition. The high values of EM1
appear in the clay layers, indicating that EM1 in the Daqu profile
corresponds to river suspension. In terms of vertical variation
(Figure 5i), EM1 dominates from 5.1 to 4.6 ka but is relatively
low at 4.6–4.1 ka. Additionally, EM5, consisted of coarse sand,
has a modal size of 454 μm, consisting of coarse sand. The high
values of EM5 also appear in the clay layers, which exhibit a
slightly positive skewness and good sorting. Therefore, EM1 and
EM5 may represent end members related to fluvial hydrodynamic
processes, representing the transport of regional surface materials
by rivers.
According to Sun et al. (2001), three characteristics of fluvial
sediments that distinguish the suspended components from salta-
tion are large grain size difference, small grain size overlap, and
small proportion difference. EM5 and EM1 have large size differ-
ence, less overlap in grain size range, and similar grain size distri-
bution of the Daqu profile sample. Moreover, EM5 is the coarsest
component in the profile, and the main composition is coarse sand
(310–586 μm), similarly to the EM1 trend. The sample of 5.1–
4.6 ka has more EM5 than that of 4.6–4.1 ka. Therefore, EM5
could be considered as the saltation component of the DQ profile
during a high hydraulic energy period in the fluvial environment.
The visible 1 cm gravel layer reveals that the Daqu River had a
large runoff, strong hydrodynamics, and a high content of EM1
and EM5 during 5.1–4.6 ka, which further indicate that EM1 and
EM5 were the suspended and saltation components in the river
deposition. As the DQ is formed by glacial meltwater during
warmer periods, the rise in EM1 and EM5 content is closely asso-
ciated with fluvial processes. Therefore, EM1 and EM5 may fur-
ther indicate changes in temperature.
EM3 and EM4 consist of medium and fine sand, with modal
grain sizes of 185.8 and 309.5 μm, respectively. The grain size
distributions of EM3 and EM4 are slightly positively skewed with
better sorting, peaking in the aeolian sand layer, indicating that
EM3 and EM4 are controlled by aeolian processes. According to
the study of Pye and Tsoar (1987), where 70–500 μm grain
Figure 6. Scanning electron microscope images of quartz from the Daqu section. (a–d) Show the quartz surface texture from the flood
layers. (e–h) Show the quartz surface texture from the aeolian layers. (i–l) Show the quartz surface texture from the mixed layers.
10 The Holocene 00(0)
represents the saltation component in aeolian deposition, EM3
and EM4 are positively correlated with the saltation component
of 70–500 μm grain (p < 0.01) (Figure 7a). So, EM3 and EM4
represent aeolian transport saltation components. Moreover, EM4
content was dominant in the 4.6–4.1 ka period, while EM3 con-
tent was minimal in the 5.1–4.6 ka period. This demonstrates that
EM4 was continuously influenced by aeolian processes through-
out the entire 4.6–4.1 ka period. Furthermore, EM3 and EM4
exhibit similar characteristics to the aeolian sand EMs formed by
westerly winds and glacier winds, as suggested by Lai et al.
(2023). Thus, EM4 corresponds to short-distance transport, repre-
senting regional aeolian activity, while EM3 represents saltation
components under long-distance aeolian transport. Studies indi-
cate that the aeolian sand activity in the Yarlung Tsangpo River
signifies a cold and dry climate environment (Li et al., 2020;
Zheng et al., 2009). EM3 and EM4 together may indicate the
intensity of aeolian sand activity in the region under cold and dry
conditions.
EM2 is mainly composed of 59–127 μm well-sorted very fine
sand with a small amount of silt. The EM2 grain size distribution
is slightly positively skewed, symmetric to EM4 and is negatively
correlated with EM4 (Figure 7b). It is suggested that EM2 and
EM4 are controlled by wind dynamics. According to Figure 4,
0–109 μm is the suspended component in the probability density
curve of EM2, so EM2 is believed to be a near-source deposition
under the action of near-surface wind after a short distance of
suspension transport.
The aeolian-fluvial interactions deposition patterns
In the Paiku Co basin, several runoffs are generated by glacier
meltwater into Paiku Co. The Daqu profile is locate in the second-
ary terrace of the Daqu runoff. When EM1 content is high, it indi-
cates a relatively high temperature, while low EM1 content
suggests a relatively low temperature. So, EM1 can reflect the
temperature trend that determines glacier melt and further changes
the river level. The EM2, EM3, and EM4 are indicators of aeolian
activities in the Paiku Co basin. An increase in their content sug-
gests intensified aeolian sand activity, while a decrease in their
content indicates a weakening of such activity. These variations in
content could reflect environmental changes in the Paiku Co
basin. As shown in Figure 5i, the range of the EM4 content from
6.59% to 66.71% illustrates that local winds were always present
in the region during 5.1–4.1 ka. In this study, combined with the
sedimentary characteristics of the Daqu profile, two periods could
be studied to discuss the sediment process of the Paiku Co basin.
During 5.1–4.6 ka, the dominance and drastic changes in EM1
indicate fluctuating and unstable river levels, which correspond to
frequent temperature variations. Specifically, high EM1 content
suggests a rise in temperature, increased glacial meltwater, and
higher river levels, leading to the formation of floods and clay
deposition in the Daqu profile. Additionally, due to increased gla-
cial meltwater, the river’s sediment-carrying capacity is enhanced,
resulting in the transport of a large amounts of debris to a certain
distance from the glacier terminus (Zhou et al., 2007). Conversely,
a decrease in EM1 content and an increase in EM3 content indi-
cate a decrease in temperature. The decrease in glacial meltwater
results in a decrease in river level, exposing sand in the riverbed.
Thin layers of sand are then deposited under the influence of local
winds. Additionally, the relative increase in χlf and CaCO3 content
suggests a relatively warm and humid period during this time.
The Yarlung Zangbo River profile also suggests that interbed-
ded deposits of alluvial silt-aeolian sand layers, subsand‒sand lay-
ers, aeolian sand-paleosol layers and peat formed at 5.5–4 ka (Li
et al., 2010). River terraces and fans that rapidly developed through
debris flow and flood processes were reported for the Garhwal
Himalaya during the 4–5 ka period of deglaciation (Barnard et al.,
2004a, 2004b). The Puruogangri and Guliya ice cores recorded a
possible warm period in 5.9–4.2 ka with a small peak in δ18O
(Duan et al., 2012; Thompson et al., 1997, 2006) (Figure 8a and b).
In addition, the humus layer developed in the dunes west of the
Puruogangri glacier at 4985 and 4750 a (Li et al., 2006). The mon-
soon rainfall record obtained from Paru Co and Pumoyum Co in
the southern TP (Bird et al., 2014; Wang et al., 2016)(Figure 8c
and d). These results indicate an unstable warm and wet period
during 5.1–4.6 ka. At this time, the lake level of Paiku Co about
18–10 m above the present lake level (Wünnemann et al., 2015).
During 4.6–4.1 ka, EM1 content decrease and increases content
of EM3 and EM4 reveal temperature decline, river levels reduced
and strong aeolian activity during this period. Moreover, the rela-
tive decrease in χlf and CaCO3 content reveals a potential response
to cold and dry climatic conditions during this period. Considering
the uncertainty in published age models, the enhanced aeolian
activity might have been a response to the 4.2 ka events at 4.6–
4.1 ka in the Paiku Co basin (Bianchi and McCave, 1999; Bond
et al., 1997; Wang et al., 2005). However, at about 4.6 ka, with EM1
content increase and EM3, EM4 content decrease, which reveals
the dominance of river activity, a temporary rise in temperature and
aeolian activity weaken. Several studies have also verified the trend
of cooling and aridity after 4.6 ka. Firstly, Neoglacial events were
also recorded glacier of Himalayan mountain after 4.6 ka. For
example, Neoglacial events in the Khumbu Himal were reported
Figure 7. Correlation analysis. (a) Shows the correlation of the contents of EM3 and EM4 with the saltation component of 70–500 μm.
(b) Shows the correlation of the contents of EM3 and EM4.
Yuan et al. 11
during about 4.3–2.5 ka (Abramowski, 2004; Finkel et al., 2003). In
addition, the Langtang Khola and Kali Gandaki valleys of the
Annapurna Himal recorded Neoglacial events at about 4.1 and
4.0 ka. Dortch et al. (2013) revealed that the glacial advances in the
western Himalayane at 4.4–3.2 ka is concurrent with the Northern
Hemisphere climatic events.
Glacier advancement during Holocene regional glacial stages
at ~4.5–2.8 ka at the northwestern end of the Himalayan-Tibetan
orogen has been reported (Saha, 2018; Saha et al., 2019). Sec-
ondly, aeolian activities of the southern TP intensified during
about 4.7–4 ka BP (Dong et al., 2017; Li et al., 2020; Liu et al.,
2013; Pan et al., 2014) (Figure 8e). The grain size PC1 of Paru Co
and δD of Pumoyum Co record reveal a decrease in precipitation
during that period (Bird et al., 2014; Wang et al., 2016) (Figure 8c
and d). Thus, as shown in Figure 9, in the Paiku Co basin, when
the temperature declined, the glacier advanced, the river level
dropped, the sand was widely exposed in the riverbed and was
influenced by local and long distant winds. Therefore, aeolian
sand was mainly deposited after 4.6 ka in the Paiku Co basin.
Based on the analysis above, in the Paiku Co Basin, during
warm and humid periods, the river’s sediment-carrying capacity
increases (Figure 9a). In cold and dry periods, under the influ-
ence of aeolian processes, materials can be transported from the
river channel to the banks (Figure 9b). Therefore, climate change
promotes the occurrence of fluvial-aeolian interactions. In other
words, near the source of aeolian sand, fluvial-aeolian deposition
is primarily controlled by climate change.
Conclusions
The study presents the response of aeolian-fluvial interaction and
sedimentation patterns to climate change on a small regional scale
in the Paiku Basin. The results suggest that the sedimentation pat-
terns of aeolian-fluvial interaction are primarily controlled by cli-
mate change on a millennial scale.
Between approximately 5.1 and 4.6 ka, warm and humid, fluvial
activity dominates due to the increased flow of the Daqu River
caused by melting glaciers. This resulted in the river carrying addi-
tional sediment. Subsequently, from 4.6 to 4.1 ka, as temperatures
decreased and the climate transitioned to cold conditions, glaciers
advanced, leading to significant sand exposure in the riverbed.
Under the influence of winds, both riverine and aeolian sands were
deposited during this period.
Besides, on a millenium scale, the sedimentation pattern of
aeolian-fluvial interaction is characterized by alternating depo-
sition of clay layers and medium to fine sand. Grain size analy-
sis reveals a distribution pattern exhibiting bimodal-multimodal
peaks. The dominant peak is observed in the range of 210–
250 μm, with a bimodal-multimodal distribution showing a pri-
mary peak at 300–350 μm, a secondary peak at 70–90 μm, and
smaller peaks at 8–10 μm.
Overall, these findings enhance our understanding of the
complex dynamics of aeolian-fluvial interaction and shed light
on the role of climatic oscillations in shaping sedimentary pat-
terns. However, further research is needed with longer times-
cales and additional dating techniques and proxies to investigate
the response patterns of aeolian-fluvial interaction sedimenta-
tion to climate changes more comprehensively.
Author contributions
Wenjie Yuan: Data curation; Investigation; Methodology; Soft-
ware; Visualization; Writing – original draft.
Yinghua Zheng: Conceptualization; Methodology; Supervision;
Writing – review & editing.
Ping Yan: Conceptualization; Funding acquisition; Resources;
Supervision.
Zhiyong Ding: Conceptualization; Formal analysis; Methodol-
ogy; Supervision; Writing – review & editing.
Xiaoxu Wang: Data curation; Investigation; Supervision;
Visualization.
Figure 8. Comparison of proxies records from the DQ profile
with the southern Tibetan Plateau environmental records. (a and b)
Show the δ18O record from Guliya and Puruogang ri, respectively
(Thompson et al., 1997, 2006). (c) Shows the PC1 grain size from
Paru Co (Bird et al., 2014). (d) Shows the δD mid-long record from
Pumuyum Co (Wang et al., 2016). (e) Shows the probability density
distribution curve of aeolian sediment ages in the Yarlung Tsangpo
River (Li et al., 2020). (f–i) Show the EM1, EM3, CaCO3, χlf from
the DQ profile in this study. The gray dashed line represents 4.6 ka.
Figure 9. Possible sediment model in the Paiku Co basin during
the mid Holocene. (a) Represents the sedimentary pattern during
5.1–4.6 ka. (b) Represents the sedimentary pattern during 4.6–
4.1 ka. The lake variation is referred from Wünnemann et al. (2015).
12 The Holocene 00(0)
Xiao Zhang: Data curation; Investigation; Software; Supervision.
Yating Gao: Data curation; Methodology; Software; Supervision.
Jiaqi Chu: Data curation; Investigation; Supervision; Visualization.
Hua Fan: Data curation; Investigation; Software; Visualization.
Yingna Liu: Data curation; Formal analysis; Methodology; Soft-
ware; Supervision.
Funding
The author(s) disclosed receipt of the following financial sup-
port for the research, authorship, and/or publication of this ar-
ticle: This work was supported by the Second Comprehen-
sive Scientific Expedition to the Qinghai-Tibet Plateau (Grant
2019QZKK0906), the National Natural Science Foundation of
China (Grant 42371008) and Opening Foundation of Engineering
Center of Desertification and Blown-Sand Control of Ministry of
Education, Beijing Normal University (No. 2023-A3-4).
ORCID iD
Ping Yan https://orcid.org/0000-0002-6031-2650
References
Abramowski U (2004) The use of 10Be surface exposure dating
of erratic boulders in the reconstruction of the late Pleisto-
cene glaciation history of mountainous regions, with exam-
ples from Nepal and Central Asia. Doctoral thesis, Bayreuth.
Available at: https://epub.uni-bayreuth.de/894/
An Z, Colman SM, Zhou W et al. (2012) Interplay between the
westerlies and Asian monsoon recorded in Lake Qinghai sedi-
ments since 32 ka. Scientific Reports 2(1): 1–7.
Barnard PL, Owen LA and Finkel RC (2004a) Style and timing
of glacial and paraglacial sedimentation in a monsoon-influ-
enced high Himalayan environment, the upper Bhagirathi
Valley, Garhwal Himalaya. Sedimentary Geology 165(3–4):
199–221.
Barnard PL, Owen LA, Sharma MC et al. (2004b) Late Quater-
nary (holocene) landscape evolution of a monsoon-influenced
high Himalayan valley, gori Ganga, Nanda Devi, NE Garh-
wal. Geomorphology 61(1–2): 91–110.
Belnap J, Munson SM and Field JP (2011) Aeolian and fluvial
processes in dryland regions: The need for integrated studies.
Ecohydrology 4(5): 615–622.
Bianchi GG and McCave IN (1999) Holocene periodicity in
North Atlantic climate and deep-ocean flow south of Iceland.
Nature 397(6719): 515–517.
Bird BW, Polisar PJ, Lei Y et al. (2014) A Tibetan lake sediment
record of holocene Indian summer monsoon variability. Earth
and Planetary Science Letters 399: 92–102.
Blott SJ and Pye K (2012) Particle size scales and classification
of sediment types based on particle size distributions: Review
and recommended procedures. Sedimentology 59(7): 2071–
2096.
Bond G, Showers W, Cheseby M et al. (1997) A pervasive mil-
lennial-scale cycle in North Atlantic holocene and glacial cli-
mates. Science 278(5341): 1257–1266.
Bullard JE and Livingstone I (2002) Interactions between aeolian
and fluvial systems in dryland environments. British Geogra-
phers 34(1): 8–16.
Bullard JE and McTainsh GH (2003) Aeolian-fluvial interac-
tions in dryland environments: Examples, concepts and
Australia case study. Progress in Physical Geography 27(4):
471–501.
Buylaert J, Jain M, Murray AS et al. (2012) A robust feldspar
luminescence dating method for Middle and Late Pleistocene
sediments: Feldspar luminescence dating of Middle and Late
Pleistocene sediments. Boreas 41(3): 435–451.
Buylaert JP, Murray AS, Gebhardt AC et al. (2013) Luminescence
dating of the PASADO core 5022-1D from Laguna Potrok
Aike (Argentina) using IRSL signals from feldspar. Quater-
nary Science Reviews 71: 70–80.
Buylaert JP, Murray AS, Thomsen KJ et al. (2009) Testing the
potential of an elevated temperature IRSL signal from K-feld-
spar. Radiation Measurements 44(5–6): 560–565.
Buylaert JP, Thiel C, Murray A et al. (2011) IRSL and post-
IR IRSL residual doses recorded in modern dust samples
from the Chinese Loess Plateau. Geochronometria 38(4):
432–440.
Clarke ML and Rendell HM (1998) Climate change impacts on sand
supply and the formation of desert sand dunes in the south-west
U.S.A. Journal of Arid Environments 39(3): 517–531.
Cohen TJ, Nanson GC, Larsen JR et al. (2010) Late Quaternary
aeolian and fluvial interactions on the Cooper Creek fan and
the association between linear and source-bordering dunes,
Strzelecki Desert, Australia. Quaternary Science Reviews
29(3–4): 455–471.
Dar RA, Chandra R, Romshoo SA et al. (2015) Micromorpho-
logical investigations of the late quaternary loess–paleosol
sequences of the Kashmir Valley, India. Journal of Asian
Earth Sciences 111: 328–338.
Dietze E, Hartmann K, Diekmann B et al. (2012) An end-member
algorithm for deciphering modern detrital processes from lake
sediments of Lake Donggi Cona, NE tibetan Plateau, China.
Sedimentary Geology 243–244: 169–180.
Ding Z, Lu R, Liu X et al. (2023) High-northern-latitude forc-
ing modulated fluvial‒aeolian interplay during the last glacial
period in the Mu Us Desert, northern central China. Quater-
nary Science Reviews 315: 108255.
Dong Z, Hu G, Qian G et al. (2017) High-altitude aeolian
research on the Tibetan Plateau. Reviews of Geophysics
55(4): 864–901.
Dortch JM, Owen LA and Caffee MW (2013) Timing and
climatic drivers for glaciation across semi-arid western
Himalayan–Tibetan orogen. Quaternary Science Reviews
78: 188–208.
Draut AE (2012) Effects of river regulation on aeolian landscapes,
Colorado River, southwestern USA. Journal of Geophysical
Research Earth Surface 117(F2): F02022.
Duan K, Yao T, Wang N et al. (2012) The unstable Holocene cli-
matic change recorded in an ice core from the central Tibetan
Plateau. Science China Press 42(9): 1441–1449.
Fang XM, Ono Y, Fukusawa H et al. (1999) Asian summer mon-
soon instability during the past 60,000 years: Magnetic sus-
ceptibility and pedogenic evidence from the western Chinese
Loess Plateau. Earth and Planetary Science Letters 168(3–4):
219–232.
Finkel RC, Owen LA, Barnard PL et al. (2003) Beryllium-10 dat-
ing of Mount Everest moraines indicates a strong monsoon
influence and glacial synchroneity throughout the Himalaya.
Geology 31(6): 561–564.
Folk RL (1966) A review of grain-size parameters. Sedimentology
6(2): 73–93.
Folk RL and Ward WC (1957) Brazos River bar [Texas]; a study
in the significance of grain size parameters. Journal of Sedi-
mentary Research 27(1): 3–26.
Fouinat L, Sabatier P, Poulenard J et al. (2017) One thousand
seven hundred years of interaction between glacial activity
and flood frequency in proglacial Lake Muzelle (western
French alps). Quaternary Research 87(3): 407–422.
Gao X and Tang K (1997) Study on alternate action of wind ero-
sion and water erosion in Shenfu-Dongsheng coal mining
area. Journal of Soil Erosion and Soil and Water Conserva-
tion 3(4): 1–7.
Huntley D and Baril M (1997) The K content of the K-feldspars
being measured in optical dating or in Thermoluminescence
dating. Ancient TL 15(1): 3.
Yuan et al. 13
Huntley D and Hancock R (2001) The Rb contents of the K-feld-
spar grains being measured in optical dating. Ancient TL
19(2): 43–46.
Hu X, Zhu L, Wang Y et al. (2014) Climatic significance of
n-alkanes and their compound-specific δD values from lake
surface sediments on the southwestern tibetan Plateau. Chi-
nese Science Bulletin 59(24): 3022–3033.
Immerzeel WW, Lutz AF, Andrade M et al. (2020) Impor-
tance and vulnerability of the world’s water towers. Nature
577(7790): 364–369.
Kasse C (2002) Sandy aeolian deposits and environments and
their relation to climate during the Last Glacial Maximum
and lateglacial in northwest and central Europe. Progress in
Physical Geography 26(4): 507–532.
Kasse C, Vandenberghe D, De Corte F et al. (2007) Late Weichse-
lian fluvio-aeolian sands and coversands of the type locality
Grubbenvorst (southern Netherlands): Sedimentary environ-
ments, climate record and age. Journal of Quaternary Science
22(7): 695–708.
Lai W, Liang W, Hu X et al. (2023) Grain-size and composi-
tional variability of Yarlung Tsangpo sand (Xigaze transect,
south Tibet): Implications for sediment mixing by fluvial
and aeolian processes. Journal of Palaeogeography 12(2):
195–210.
Langford RP and Chan MA (1989) Fluvial-aeolian interactions:
Part II, ancient systems. Sedimentology 36(6): 1037–1051.
Lei Y, Yao T, Yang K et al. (2018) An integrated investigation of
lake storage and water level changes in the Paiku Co basin,
central Himalayas. Journal of Hydrology 562: 599–608.
Lei Y, Yao T, Yang K et al. (2021) Contrasting hydrological and
thermal intensities determine seasonal lake-level variations
– A case study at Paiku Co on the southern tibetan Plateau.
Hydrology and Earth System Sciences 25(6): 3163–3177.
Li G, Wen L, Xia D et al. (2015) Quartz OSL and K-feldspar
pIRIR dating of a loess/paleosol sequence from arid central
Asia, Tianshan Mountains, NW China. Quaternary Geochro-
nology 28: 40–53.
Li S, Yang P, Dong Y et al. (2010) Land Desertification and Its
Prevention and Control in Tibet. Beijing: Science Press (in
Chinese).
Li S, Yao T, Tian L et al. (2011) Seasonal transition characteristics
of the westerly jet: Study based on field observations at an
altitude of 6900 m on the Mt. Xixiabangma Dasuopu glacier.
Chinese Science Bulletin 56(18): 1912–1920.
Li T, Zhang J, Wu Y et al. (2020) Holocene aeolian activities
linked to Indian summer monsoon in the middle reaches of
the Yarlung Zangbo River. Journal of Geographical Sciences
30(12): 2002–2014.
Liu X, Song L, Jin Y et al. (2013) History of aeolian deposits in
Tibetan Plateau and climate change over Holocene. Journal
of Arid Land Resources and Environment 27(6): 41–47.
Li X, Yi C, Chen F et al. (2006) Formation of proglacial dunes in
front of the Puruogangri Icefield in the central Qinghai–Tibet
Plateau: Implications for reconstructing paleoenvironmental
changes since the Lateglacial. Quaternary International 154-
155: 122–127.
Mueller D, Raith K, Bretzke K et al. (2023) Luminescence chro-
nology of fluvial and aeolian deposits from the Emirate of
Sharjah, UAE. Quaternary Research 112: 111–127.
Nanson GC, Chen XY and Price DM (1995) Aeolian and fluvial
evidence of changing climate and wind patterns during the
past 100 ka in the western Simpson Desert, Australia. Palaeo-
geography Palaeoclimatology Palaeoecology 113(1): 87–102.
Pan M, Wu Y, Zheng Y et al. (2014) Holocene aeolian activity in
the Dinggye area (Southern Tibet, China). Aeolian Research
12: 19–27.
Paterson GA and Heslop D (2015) New methods for unmixing
sediment grain size data. Geochemistry Geophysics Geosys-
tems 16(12): 4494–4506.
Pomar F, del Valle L, Fornós JJ et al. (2018) Late pleistocene
dune–sourced alluvial fans in coastal settings: Sedimentary
facies and related processes (Mallorca, western Mediterra-
nean). Sedimentary Geology 367: 48–68.
Prescott JR and Hutton JT (1994) Cosmic ray contributions to
dose rates for luminescence and ESR dating: Large depths
and long-term time variations. Radiation Measurements
23(2–3): 497–500.
Pye K and Tsoar H (1987) The mechanics and geological implica-
tions of dust transport and deposition in deserts with particu-
lar reference to loess formation and dune sand diagenesis in
the northern Negev, Israel. Geological Society London Spe-
cial Publications 35(1): 139–156.
Regional Geology of Xizang (Tibet) Autonomous Region
(1993) Bureau of Geology and Mineral Resources of Xizang
Autonomous Region. Beijing: Geological Publishing House
(in Chinese).
Reimann T, Tsukamoto S, Naumann M et al. (2011) The poten-
tial of using K-rich feldspars for optical dating of young
coastal sediments – A test case from darss-Zingst peninsula
(southern Baltic Sea coast). Quaternary Geochronology
6(2): 207–222.
Robins L, Roskin J, Marder O et al. (2023) Geomorphic, environ-
mental, and archeological significance of last glacial maxi-
mum to middle holocene dune damming, northwestern Negev
dunefield margin, Israel. Quaternary Science Reviews 308:
108098.
Robins L, Roskin J, Yu L et al. (2022) Aeolian-fluvial processes
control landscape evolution along dunefield margins of the
northwestern Negev (Israel) since the late Quaternary. Qua-
ternary Science Reviews 285: 107520.
Roskin J, Katra I, Agha N et al. (2014) Rapid anthropogenic
response to short-term aeolian-fluvial palaeoenvironmental
changes during the late pleistocene–Holocene transition in the
northern Negev Desert, Israel. Quaternary Science Reviews
99: 176–192.
Saha S (2018) Reconstructing High-frequency Holocene Glacial
Chronostratigraphies in the Himalayan-Tibetan Orogen. Uni-
versity of Cincinnati. Available at: http://rave.ohiolink.edu/
etdc/view?acc_num=ucin1543920897036039
Saha S, Owen LA, Orr EN et al. (2019) High-frequency holocene
glacier fluctuations in the himalayan-tibetan orogen. Quater-
nary Science Reviews 220: 372–400.
Simpson EL, Hilbert-Wolf HL, Simpson WS et al. (2008) The
interaction of aeolian and fluvial processes during deposi-
tion of the Upper Cretaceous capping sandstone member,
Wahweap Formation, Kaiparowits Basin, Utah, U.S.A. Pal-
aeogeography Palaeoclimatology Palaeoecology 270(1–2):
19–28.
Sun D, An Z, Su R et al. (2001) Mathematical approach to sedi-
mentary component partitioning of polymodal sediments and
its applications. Progress in Natural Science 11(5): 56–64.
Sun D, Bloemendal J, Rea DK et al. (2002) Grain-size distribu-
tion function of polymodal sediments in hydraulic and aeolian
environments, and numerical partitioning of the sedimentary
components. Sedimentary Geology 152(3-4): 263–277.
Thiel C, Buylaert J-P, Murray A et al. (2011) Luminescence dating
of the stratzing loess profile (Austria) – Testing the potential
of an elevated temperature post-IR IRSL protocol. Quater-
nary International 234(1-2): 23–31.
Thomas DSG, Nash DJ, Shaw PA et al. (1993) Present day lunette
sediment cycling at Witpan in the arid southwestern Kalahari
Desert. CATENA 20(5): 515–527.
14 The Holocene 00(0)
Thompson LG, Tandong Y, Davis ME et al. (2006) Holocene cli-
mate variability archived in the Puruogangri ice cap on the
central Tibetan Plateau. Annals of Glaciology 43: 61–69.
Thompson LG, Yao T, Davis ME et al. (1997) Tropical climate
instability: The last glacial cycle from a qinghai-tibetan ice
core. Science 276(5320): 1821–1825.
Van Huissteden J, Vandenberghe J, Van der Hammen T et al.
(2000) Fluvial and aeolian interaction under permafrost con-
ditions: Weichselian late pleniglacial, Twente, eastern Nether-
lands. CATENA 40(3): 307–321.
Wang Y (1985) Model Atlas of Surface Textures of Quartz Sand.
Beijing: Science Press (in Chinese).
Wang Y, Cheng H, Edwards RL et al. (2005) The holocene Asian
Monsoon: Links to solar changes and North Atlantic climate.
Science 308(5723): 854–857.
Wang Y, Zhu L, Wang J et al. (2016) Paleohydrological processes
revealed by n-alkane δD in lacustrine sediments of Lake
Pumoyum Co, southern Tibetan Plateau, and their response to
climate changes during the past 18.5 cal ka. Journal of Paleo-
limnology 56(2-3): 223–238.
Weltje GJ (1997) End-member modeling of compositional data:
Numerical-statistical algorithms for solving the explicit mix-
ing problem. Mathematical Geology 29(4): 503–549.
Wünnemann B, Yan D and Ci R (2015) Morphodynamics and
lake level variations at Paiku Co, southern tibetan Plateau,
China. Geomorphology 246: 489–501.
Xiao J, Chang Z, Fan J et al. (2012) The link between grain-size
components and depositional processes in a modern clastic
lake. Sedimentology 59(3): 1050–1062.
Xie Y (1984) Atlas of Quartz Sand Surface Textural Features of
China Micrographs. Beijing: Ocean Press (in Chinese).
Yanchilina AG, Ryan WBF, Kenna TC et al. (2019) Meltwater
floods into the Black and Caspian Seas during Heinrich Sta-
dial 1. Earth-Science Reviews 198: 102931.
Yang S, Ding Z, Li Y et al. (2015) Warming-induced northwest-
ward migration of the East Asian monsoon rain belt from the
Last Glacial Maximum to the mid-holocene. Proceedings of
the National Academy of Sciences 112(43): 13178–13183.
Yan P, Li X, Ma Y et al. (2015) Morphological characteristics
of interactions between deserts and rivers in northern China.
Aeolian Research 19: 225–233.
Yu L, Lai Z, An P et al. (2015) Aeolian sediments evolution
controlled by fluvial processes, climate change and human
activities since LGM in the Qaidam Basin, Qinghai-Tibetan
Plateau. Quaternary International 372: 23–32.
Yu L, Sun Y, An P et al. (2022) Dunefield expansion and paleocli-
mate during MIS 3 in the Qaidam Basin, northeastern Tibetan
Plateau: Evidence from Aeolian-fluvial processes and revised
luminescence chronologies. CATENA 215: 106354.
Zhao H and Li SH (2005) Internal dose rate to K-feldspar grains
from radioactive elements other than potassium. Radiation
Measurements 40(1): 84–93.
Zheng Y, Wu Y, Li S et al. (2009) Grain-size characteristics of
sediments formed since 8600 yr BP in middle reaches of
Yarlung Zangbo River in Tibet and their paleoenvironmental
significance. Chinese Geographical Science 19(2): 113–119.
Zhou N, Li Q, Zhang CL et al. (2021) Grain size characteristics
of aeolian sands and their implications for the aeolian dynam-
ics of dunefields within a river valley on the southern Tibet
Plateau: A case study from the Yarlung Zangbo river valley.
CATENA 196: 104794.
Zhou X, Liu Y and Wang M (2007) Geomorphology and Funda-
mentals of Quaternary Geology. Beijing: Geological Publish-
ing House.
Zhu D, Shao Z, Meng X et al. (2008) Redefinition and division
of Quaternary lacustrine strata and establishment of the Paiku
Co group in the Paiku Co basin, Tibet, China. Geological Bul-
letin of China 27(7): 1035–1043.