The Yellow River, well-known for its tremendous sediment load, originates from the northeast of the Tibetan Plateau, and flows eastwards to the Bohai Sea. It runs across the Chinese Loess Plateau (CLP) and the east of the Ordos Plateau, resulting in the formation of a long large canyon called Jinshaan Canyon between the Shanxi (Jin) and Shaanxi (Shaan) provinces (
(a) DEM showing the course of the Yellow River, the Ordos Plateau and the location of study area (Heiyukou) and other localities mentioned in the text, (b) Oblique Google Earth image showing the localities of the fluvial sediment exposures (sections) numbered A to G, the localities are also defined as fluvial terraces shown in Fig. 2, and the lithology and stratigraphy of the exposures are shown in Fig. 3. The pictures of Exposures A and D are also shown in Fig. 4.
The recent development of optically stimulated luminescence (OSL) dating techniques makes it possible to accurately determine the formation ages of fluvial terraces (e.g. Gong
The Yellow River flows along the northeastern and northern borders of the Ordos Plateau and cuts through the eastern plateau at an average elevation of 1000 to 1500 m above sea level (asl) from north to south. This results in the formation of the Jinshaan Canyon, which is limited to the reach from Lamawan to Yumenkou (
In the Heiyukou area (38°32′36.34″N, 110°54′25.88″E) (
Synthetic cross section of the Yellow River terrace sequence (T1 to T7) of the Heiyuhou area (see text for details), the capital letters in brackets refer to the localities of the exposures in Fig. 1b and the section numbers in Fig. 3.
Detailed stratigraphy of the Yellow River terrace deposits in the Heyukou area, the localities of the sections are marked in Fig. 1b, and the associated terraces are displayed in Fig. 2. The positions of OSL samples and the OSL ages (ka) on fine-grained quartz and coarse-grained quartz (in bold italics) are given. Sections A and B are also illustrated by the pictures in Fig. 4.
Photographs looking north showing Exposures A (a and b) and D (c and d). The channel lag deposits (gravel) atop the strath surface are overlaid by floodplain silt underlaying loess/paleosol deposits. Arrows point to sampling positions, and numbers refer to the sample number (HK16-x).
The deposits on the terraces can be divided into two units: fluvial terrace deposits between the strath and tread, and loess/palaeosol or red clay deposits overlying the fluvial deposits (
As shown in
Results of optical dating of terrace samples from the Yellow River in the Heyoukou area.
HYK16-6 | L3343 | 0.2 | Modern fluvial sand | 125–150 | 1.15 ± 0.06 | 7.02 ± 0.22 | 2.05 ± 0.06 | 81.2 ± 5.0 | 1.71 ± 0.07 | 21 | 91 ± 23 | 1.9 ± 0.5 | 1.1 ± 0.3 |
4–11 | 1.91 ± 0.07 | 5 | 99.5 ± 9.6 | 52.1 ± 5.4 | |||||||||
HYK16-1 | L3338 | 4.78 | Loess | 90–125 | 2.29 ± 0.09 | 9.55 ± 0.28 | 1.75 ± 0.06 | 83.2 ± 5.0 | 2.81 ± 0.06 | 25 | 36 ± 5 | 251.8 ± 19.6 | 89.8 ± 7.2 |
4–11 | 3.33 ± 0.07 | 6 | 219.8 ± 4.0 | 66.0 ± 1.8 | |||||||||
HYK16-2 | L3339 | 5.45 | Paleosol | 90–125 | 2.43 ± 0.10 | 9.98 ± 0.29 | 1.86 ± 0.06 | 83.6 ± 5.0 | 2.68 ± 0.06 | 24 | 53 ± 8 | 193.5 ± 22.6 | 72.3 ± 8.6 |
4–11 | 3.17 ± 0.06 | 6 | 198.7 ± 9.7 | 62.7 ± 3.3 | |||||||||
HYK16-3 | L3340 | 6.3 | Floodplain silt | 125–150 | 1.66 ± 0.08 | 8.07 ± 0.24 | 1.75 ± 0.06 | 75.80 ± 4.93 | 2.68 ± 0.05 | 24 | 48 ± 7 | 147.6 ± 18.0 | 55.2 ± 6.8 |
4–11 | 2.98 ± 0.06 | 6 | 198.1 ± 6.3 | 66.4 ± 2.5 | |||||||||
HYK16-4 | L3341 | 6.8 | Floodplain silt | 90–125 | 1.91 ± 0.08 | 9.94 ± 0.29 | 1.69 ± 0.06 | 77.7 ± 5.1 | 2.82 ± 0.06 | 22 | 43 ± 8 | 171.8 ± 18.4 | 61.0 ± 6.7 |
4–11 | 3.16 ± 0.06 | 6 | 198.1 ± 13.5 | 62.6 ± 4.5 | |||||||||
HYK16-5 | L3342 | 8.57 | Sand | 150–200 | 0.96 ± 0.05 | 5.30 ± 0.18 | 2.04 ± 0.06 | 73.9 ± 4.8 | 2.70 ± 0.07 | 26 | 42 ± 6 | 124.8 ± 9.4 | 46.2 ± 3.7 |
4–11 | 2.91 ± 0.10 | 6 | 208.6 ± 6.1 | 71.7 ± 3.1 | |||||||||
HYK16-23 | L3360 | 4.5 | Paleosol | 4–11 | 2.12 ± 0.09 | 10.5 ± 0.30 | 2.06 ± 0.06 | 77.6 ± 4.89 | 3.30 ± 0.07 | 7 | 235.2 ± 3.0 | 71.2 ± 1.8 | |
HYK16-22 | L3359 | 7.5 | Floodplain silt | 90–125 | 1.85 ± 0.08 | 9.3 ± 0.27 | 1.77 ± 0.06 | 86.0 ± 5.16 | 2.68 ± 0.06 | 23 | 23 ± 4 | 170.7 ± 9.1 | 63.6 ± 3.6 |
4–11 | 3.16 ± 0.07 | 3 | 232.6 ± 11.4 | 73.6 ± 3.9 | |||||||||
HYK16-21 | L3358 | 18.4 | Floodplain silt | 90–125 | 2.26 ± 0.09 | 11.3 ± 0.32 | 1.66 ± 0.06 | 79.1 ± 4.9 | 2.74 ± 0.06 | 22 | 25 ± 4 | 193.2 ± 11.4 | 70.6 ± 4.4 |
4–11 | 3.30 ± 0.07 | 3 | 224.3 ± 3.9 | 68.0 ± 1.9 | |||||||||
HYK16-7 | L3344 | 0.8 | Paleosol | 4–11 | 2.13 ± 0.09 | 9.67 ± 0.28 | 1.96 ± 0.06 | 90.4 ± 5.2 | 3.23 ± 0.07 | 7 | 344.1 ± 5.5 | 106.4 ± 2.9 | |
HYK16-8 | L3345 | 2.26 | Floodplain silt | 90–125 | 1.79 ± 0.08 | 8.99 ± 0.27 | 1.67 ± 0.06 | 72.3 ± 4.7 | 2.63 ± 0.22 | 22 | 27 ± 5 | 254.2 ± 15.1 | 96.5 ± 9.9 |
4–11 | 3.09 ± 0.07 | 3 | 357.6 ± 1.3 | 115.6 ± 2.5 | |||||||||
HYK16-9 | L3346 | 2.86 | Floodplain silt | 90–125 | 2.31 ± 0.09 | 9.95 ± 0.29 | 1.74 ± 0.06 | 88.2 ± 5.1 | 2.86 ± 0.06 | 22 | 26 ± 4 | 224.3 ± 12.9 | 78.5 ± 4.8 |
4–11 | 3.39 ± 0.07 | 3 | 345.6 ± 13.0 | 101.8 ± 4.4 | |||||||||
HYK16-10 | L3347 | 3.41 | Floodplain silt | 90–125 | 1.89 ± 0.08 | 8.59 ± 0.26 | 1.67 ± 0.06 | 71.3 ± 4.6 | 2.61 ± 0.06 | 24 | 25 ± 1 | 238.7 ± 14.4 | 91.5 ± 5.8 |
4–11 | 3.07 ± 0.07 | 3 | 421.5 ± 22.3 | 137.5 ± 7.8 | |||||||||
HYK16-11 | L3348 | 5.5 | Sand | 90–125 | 1.70 ± 0.08 | 8.21 ± 0.25 | 1.85 ± 0.06 | 70.7 ± 4.6 | 2.82 ± 0.06 | 23 | 38 ± 6 | 224.7 ± 19.6 | 79.6 ± 7.1 |
4–11 | 3.28 ± 0.07 | 3 | 353.9 ± 11.0 | 107.7 ± 3.9 | |||||||||
HYK16-17 | L3354 | 2.3 | Paleosol | 4–11 | 2.37 ± 0.09 | 13.4 ± 0.38 | 2.32 ± 0.07 | 111.0 ± 6.1 | 3.87 ± 0.08 | 7 | 544.6 ± 6.1 | 140.8 ± 3.5 | |
HYK16-16 | L3353 | 9 | Sand | 200–250 | 0.95 ± 0.05 | 3.42 ± 0.13 | 2.26 ± 0.07 | 63.5 ± 4.4 | 2.47 ± 0.06 | 20 | 32 ± 6 | 200.9 ± 14.8 | 81.5 ± 6.3 |
4–11 | 2.80 ± 0.06 | 3 | 235.8 ± 4.5 | 84.2 ± 2.5 | |||||||||
HYK16-18 | L3355 | 32.7 | Loess | 90–125 | 2.36 ± 0.09 | 12.3 ± 0.34 | 1.85 ± 0.06 | 85.4 ± 5.1 | 2.97 ± 0.06 | 22 | 24 ± 4 | 255.0 ± 14.1 | 85.8 ± 5.0 |
4–11 | 3.57 ± 0.07 | 3 | 411.7 ± 2.2 | 115.2 ± 2.3 | |||||||||
HYK16-19 | L3356 | 39.6 | Floodplain silt | 4–11 | 2.06 ± 0.09 | 11.8 ± 0.33 | 2.09 ± 0.06 | 98.3 ± 5.5 | 3.66 ± 0.07 | 7 | 496.5 ± 4.9 | 135.6 ± 3.0 | |
HYK16-12 | L3349 | 2.2 | Loess | 4–11 | 2.34 ± 0.09 | 10.4 ± 0.30 | 1.92 ± 0.06 | 93.1 ± 5.2 | 3.63 ± 0.07 | 7 | 540.4 ± 9.0 | 149.1 ± 3.9 | |
HYK16-13 | L3350 | 22.2 | Red clay | 4–11 | 2.32 ± 0.09 | 10.5 ± 0.30 | 1.89 ± 0.06 | 92.9 ± 5.2 | 3.12 ± 0.07 | 7 | 418.1 ± 15.1 | 134.2 ± 5.7 | |
HYK16-14 | L3351 | 39 | Sand | 4–11 | 1.54 ± 0.07 | 7.14 ± 0.23 | 1.89 ± 0.06 | 112.0 ± 6.2 | 2.91 ± 0.06 | 3 | 498.7 ± 6.5 | 171.4 ± 4.2 |
Sample preparation was conducted in a dark room with a dim red light. The fine quartz grains were first extracted using the procedure in our laboratory (Zhang and Zhou, 2007), and the portions of >11 μm were used for extracting coarse-grained quartz (Zhang
In order to find suitable luminescence procedures on quartz and feldspar grains from these samples, the samples (HYK16-1 to -5) from Section A and the modern sample (HYK16-6) (
All luminescence measurements were carried out using an automated Risø TL/OSL-DA-15 reader equipped with a 90Sr/90Y beta source (Bøtter-Jensen
The U, Th and K contents of the samples were analysed using neutron-activation-analysis (NAA). The water contents are assumed to be 5 ± 1% for sand samples, and 10 ± 2% loess and floodplain silt and 20 ± 4% for palaeosol samples, respectively. This is because the sampled sections have been exposed to air for a long time before sampling, and the sediments near the surface were partly dried up. The alpha efficiency factor (
Examples of OSL decay curves and dose-response curves (DRC) for the FG quartz aliquots of the samples from the T2 to T7 terraces are shown in
Dose response curves for the fine quartz OSL signals from six samples from the T2–T7 terraces, and insets show their natural OSL shine-down curves. The solid circles represent the sensitivity-corrected natural signals (Lx/Tx). The curves were fitted using a double saturating exponential function of the form y=a(1-exp(-bx)) + c(1-exp(-dx).
Distributions of De values for the coarse-grained quartz fractions of the samples from Section D in the T4 terrace. Left column: plots of De value as a function of sensitivity-corrected natural OSL signals; Right column: radial plots of De values, the same data as shown in the left column are demonstrated as radial plots (Galbraith et al., 1999). The right-hand y-axis refers to De values, and the x-axis shows the precision of the individual De values. The shaded regions include all aliquots (solid circles) within 2σ errors, and open circles represent aliquots that fall outside the region.
The dating results are summarised and presented in
The degree of bleaching of fluvial sediment samples are generally evaluated by the OD values of their single-grain or aliquot
Comparison of fine-grained and coarse-grained quartz SAR-OSL ages for the samples from the Yellow River terraces and modern analogue samples. The circles, diamonds, squares and triangles refer to channel sands, floodplain silts, overlying loess/palaeosol deposits and modern fluvial samples, respectively. The solid dots indicate the data in this study, and the open dots are the published data for the Hukou area (Zhang et al, 2010, 2011; Hu et al., 2010; Guo et al., 2012). For comparison, the residual De values in Hu et al. (2010) were converted to ages in ka using the dose rates of 1.71 ± 0.07 Gy/ka for coarse grains and 1.91 ± 0.07 Gy/ka for fine grains of the modern sample (HYK16-6) in this study. Error bars represent 1 standard deviation of the mean; where not visible, they are smaller than the size of the data points.
The OSL ages on fine and coarse quartz grains are compared in
The internal stratigraphic consistency of OSL ages obtained are usually used to evaluate their potential validity, stratigraphically consistent ages may be valid (Rhodes
For Section D, the OSL ages on FG quartz for the five samples are statistically consistent except for sample HYK16-10 (
As shown in
Here terrace ages refer to the age of strath formation, which are usually determined by directly dating the over-lying terrace deposits that must be accumulated immediately after or simultaneous to strath cutting. If the terrace deposits are thick (fill terrace), ages obtained for the strath formation may be underestimated. For Sections A, B, D and F (
Based on the above discussion and associated considerations, we attempt to deduce the formation ages of the studied terraces. For the paired T2 terrace (T2-E and T2-W), seven samples from Sections A and B were dated. As shown in
The formation ages of the T2 and T4 terraces are compared with the marine oxygen-isotope stage (MIS) in
(a) Plot of terrace heights above modern river against terrace ages. The long-term incision rates derived from the T2, T4 and T5 terraces were calculated. Note that the T3, T6 and T7 terraces are not displayed because the OSL ages of the samples from these terrace are not valid(see text for details), and the calculated incision rate for the T5 terrace is maximum due to its minimum age. Stacked δ18O record of benthic foraminifera is from Lisiecki and Raymo (2005), and the marine oxygen isotope stages are displayed.
Based on the terrace ages and the elevation of the bedrock strath surface above the modern channel, the long-term incision rates of the river were calculated to be <0.36, 0.34 and 0.18 mm/a for at least the past 141 ka (T5), 108 ka (T4) and 72 ka (T2), respectively. This indicates that the incision rate generally decreases through time, similar to the situation in the Hukou area, where the river incision rates in the past 116, 84 and 29 ka are respectively 0.52, 0.44 and, 0.31 mm/a (Zhang
Seven Yellow River terraces (T1–T7) in the Heiyukou area were identified, including one fill terrace (T7) and five strath terraces (T2–T6). The deposits on the terraces consist of the channel, overbank and overlying loess facies. Twenty samples from these terraces and one modern fluvial sample were optically dated using fine- and coarse-grained quartz fractions. The fine and coarse quartz grains demonstrated different luminescence behaviours. The reliability of the ages obtained for the samples is evaluated based on bleachability, comparison of OSL ages on fine and coarse grains, stratigraphic consistency and geomorphological setting. The paired T2 terrace was dated to 72 ± 3 ka, and the formation ages of the T4, T5 terraces are 108 ± 3 and >141 ± 4 ka, respectively. The deposits on the T6 and T7 terraces are beyond the age range of luminescence dating, and the age determination of the T3 terrace needs more samples. Based on the ages and elevation above the modern river of the T2, T4 and T5 terraces, the long-term incision rates were calculated to be <0.36, 0.34 and 0.18 mm/a for at least the past 141, 108 and 72 ka, respectively, they also represent the uplift rates of the Ordos plateau.