Luminescence Dating Of Sediments From SE Margin Of Mu Us Sandy Land And Implications For Environment Change

Mu Us sandy land is an ideal region for the research of ancient climate and environment change and the relationship between people, as it is located in the North-west edge area of the East Asian monsoon and there are a large number of human activity sites. We collected 29 samples from 11 sections in field for an accurate and reliable framework, as it is the key to the reconstruction of palaeoenvironment evolution. Most of the luminescence dated ages ever reported for this area are based on quartz mineral, whereas feldspar minerals are less used. We found that, in experiment, the quartz signal of many Holocene samples here is too weak to be dated. In order to solve this problem, we extracted K-feldspar in the samples, utilizing a higher stimulation temperature of 150℃ after the 50℃ low temperature (post-IR IRSL, as pIRIRi150) for dating. At the same time, we selected 3 samples with strong quartz signal from lacustrine sequences and single aliquot regenerative-dose (SAR) protocol was applied to measure the equivalent dose (De). The three quartz ages can be an independent age control. In view of much more instrument time needed for dating K-feldspar samples, we use the standard growth curve (SGC), with pIRIRiso protocol, to test the Des.For 3 quartz samples, dose recovery test shows that their recovery ratios are 0.9-1.1, with a mean of 0.98±0.025 (n=16). Luminescence characteristics show that their quartz signal is very strong; most of the recycling ratios are 0.9-1.1, with a mean of 1.01±0.008 (n=78); the heat transfer effect is less than 5%; infrared detection shows that there’s no feldspar pollution in quartz. De distributions are relatively concentrated, with discrete degree less than 20%.For the same 3 samples, K-feldspar mineral is extracted. Dose recovery test shows that the recovery ratios of pIRIR150 signal are 0.9-1.1, with a mean of 1.0±0.003 (n=9); Luminescence characteristics show that the pIRIR150 signal is much stronger than that of quartz; both the recycling ratios and the heat transfer effect meet the requirements well. Residual dose test shows that the residual doses of pIRIR150 signal are between -0.2 Gy to 0.026 Gy, equivalent to causing deviation of less than 20 years for the measured samples. Because the effect is very small, the residual doses are not eliminated when calculating the ages. Anomalous fading test shows that the pIRIR150 g-values of the 3 samples are 0.55-1.71%/decade, much smaller than IRSL50 signal. De distributions are more concentrated, with discrete degree smaller than quartz.6 samples from 2 different profiles are selected to establish the SGC in this region, which is used to calculate Des of all the samples. For samples with De in 1-70 Gy, the ratios of SGC De to pIRIR150 De are 0.92-1.06, with a mean of 1.00±0.01; for young samples with De less than 1 Gy, the calculation error is bigger.In the current study, several main conclusions are as follows:(1) The experiment conditions of quartz SAR protocol are suitable, and the quartz ages are reliable; the 3 quartz ages can be used as contrast basis.(2) The experiment conditions of K-feldspar pIRIRiso protocol are suitable for the same 3 samples, and there’s no need for abnormal fading rate calibration; K-feldspar pIRIR150 ages, consistent with quartz ages, are reliable. So K-feldspar pIRIR150 protocol is available for dating samples in this region.(3) The SGC established in this paper can accurately calculate Des of samples between 1 to 70 Gy, but shouldn’t be used for De calculation of young samples less than 1 Gy.(4) During last glacial maximum to last deglacial (22-11.8 ka), dunes were active, as the climate was dry and cold; during early Holocene and preliminary late Holocene (11.8-2.4 ka), there were lacustrine sediments with different ages, it tended to be warmer in early Holocene and it was relatively the best in the middle of the Holocene; during the late Holocene (2.4-0 ka), dunes activated again and there’s a fixed trend about 1.5 ka. Human activity is the main factor of dune activation.(5) There were shallow lakes and wide meander in Sanchahe section of Salawusu river; during 2.4-1.5 ka, the river incised deeply and Sanchahe formed. This had a great influence on landscape pattern and dune activation in the study area.In this paper, we verify the reliability of K-feldspar pIRIRiso protocol dating samples in Mu Us sandy land and the accuracy of SGC method calculating De, providing strong technical support for follow-up study in this region.