香港白沙洲现代水成沉积样品的钾长石低温MET-pIRIR法测年研究

对香港白沙洲(连岛沙坝)采集两个现代样品，其年代范围为小于14年，并利用钾长石碎屑粗颗粒对这两个现代样品开展低温多步升温红外激发法(MET-pIRIR)测年研究，以进一步扩宽钾长石MET-pIRIR法的应用范围。经测试后(激发温度分别为50 ℃、80 ℃、110 ℃、140 ℃和170 ℃)，发现两个样品110 ℃、140 ℃和170 ℃的等效剂量(D_e)数据在平均年龄模型下均有明显的年龄-温度坪。第1个钾长石样品(BSZ-A)在110 ℃、140 ℃和170 ℃对应年代结果分别为-26±30 a、-22±30 a和6±49 a，这些年代结果在误差范围内与年龄期望值一致；第2个钾长石样品(BSZ-B)在110 ℃、140 ℃和170 ℃的对应年代结果分别为357±168 a、289±138 a和391±158 a，这些年代结果比年龄期望值偏老约300年。以上数据表明对于香港白沙洲海岸带的年轻水成沉积物，在利用钾长石低温MET-pIRIR法测试时，如果采用平均年龄模型，即使出现很好的年龄-温度坪，但结果仍可能偏老数百年。通过分析发现造成第2个样品偏老的主要原因是在D_e分析中，先假定了平均年龄模型对D_e数据进行分析；这种假设对晒退较好的风成沉积物成立，但是对于晒退不充分的水成沉积物不一定成立，其应该采用最小年龄模型。在平均年龄模型中，由于平均效应可能造成一些晒退不好的D_e数据混入最终年代计算之中，造成年龄偏老。针对以上不足，本次研究对低温MET-pIRIR法的D_e分析流程进行了改进：首先对样品建立钾长石低温MET-pIRIR标准生长曲线，利用该曲线可以快速、大量获取不同测片的D_e数据；然后对每个测片逐片分析等效剂量-温度坪，并挑选有坪的D_e数据参与最终年代计算，以筛除pIRIR信号晒退很差的测片；最后根据筛选后D_e数据的统计分布特征和离散度(over-dispersion)来决定统计年龄模型。利用低温MET-pIRIR法新的D_e分析流程对香港白沙洲采集的两个现代样品进行重新测年，发现这两个样品的测年结果均比较理想，例如BSZ-A样品基于79个测片的年龄结果为-10±10 a，而BSZ-B样品基于73个测片的年龄结果为27±27 a。低温MET-pIRIR法改进的D_e分析流程为今后测试有晒退问题的钾长石光释光样品提供一种新的数据分析思路，尤其是钾长石MET-pIRIR法在平均年龄模型中没有年龄-温度坪的样品。

Optical dating study of modern water-lain sediments from Pak Sha Chau in Hong Kong based on low temperature MET-pIRIR method using K-feldspar grains

Pak Sha Chau is a typical tombolo at Sai Kung in Hong Kong. In this study, two modern water-lain OSL samples were collected on May 4, 2019 at Pak Sha Chau in Hong Kong and the two OSL samples were termed as BSZ-A and BSZ-B, respectively. According to the Google Earth images for the Pak Sha Chau at different time, it is found that the sands of BSZ-A(22°22′34.65″N, 114°17′14.31″E) and the sands of BSZ-B(22°22′34.54″N, 114°17′13.59″E) were deposited after March 9, 2008. Thus, the OSL ages of BSZ-A and BSZ-B were controlled to be less than 14 years. The low temperature multi-elevated-temperature post-IR IRSL(MET-pIRIR)dating protocol was then applied to the two OSL samples to further test the robustness of the protocol. In the protocol, IR stimulation temperature was set at 50 ℃, 80 ℃, 110 ℃, 140 ℃ and 170 ℃, respectively and a moderate preheating of 200 ℃ for 60 s was applied before measuring the IRSL and MET-pIRIR signals. Signals from the test dose were used for sensitivity change correction. At the end of each cycle, an IR bleaching at 280 ℃ for 100 s was applied to minimize the signals left to the next cycle. As MET-pIRIR signals have some residual doses, the residual doses of 0.22±0.01 Gy, 0.39±0.02 Gy, 0.63±0.03 Gy, 0.85±0.06 Gy and 1.11±0.13 Gy were used for correction for De results of 50 ℃ IRSL, 80 ℃ IRSL, 110 ℃ IRSL, 140 ℃ IRSL and 170 ℃ IRSL, respectively. After residual dose correction, it is found that both OSL samples exhibit obvious age-temperature plateau in the MET-pIRIR method if the average age model was used. The 110 ℃ IRSL, 140 ℃ IRSL and 170 ℃ IRSL of OSL sample(BSZ-A)produced the ages of -26±30 a, -22±30 a and 6±49 a, respectively. These ages are consistent with expected age within error. For the OSL sample(BSZ-B), the 110 ℃ IRSL, 140 ℃ IRSL and 170 ℃ IRSL produced the ages of 357±168 a, 289±138 a and 391±158 a, respectively. The ages are ca. 300 a older than the expected age. The above results suggest that the average age model for low temperature MET-pIRIR De data might lead to the age overestimation of several hundred years for coast water-lain young sediment, even if the samples exhibit obvious age-temperature plateau. By further investigation, it is found that the main reason for the age overestimation of BSZ-B is the average age model assumed for De analysis. Such assumption might be valid for wind-blown sediments, but might be invalid for water-lain sediments. Some insufficient bleaching De data might be mixed and used in the finial age calculation. To address above limitations, the De determining procedure of the low temperature MET-pIRIR method was modified as following: firstly, constructing the low temperature MET-pIRIR standardized growth curve(SGC)for K-feldspar and measuring enough De for different aliquots. In this study, five SGCs for low temperature MET-pIRIR signals were constructed using least squares(LS)-normalisation method, based on the results of 23 aliquots of two samples. Secondly, analyzing the relation between De and stimulation temperature for each aliquot and selecting De value with plateau. Thirdly, using different age model to analysis the final ages according to the over-dispersion(OD)value of selected De data. In this study, the minimum age model was used when the OD value of De data was calculated to be more than 22 %. Mean age model was used when the OD value of De data was calculated to be not more than 22 %. The modified De determining procedure of the MET-pIRIR method was used to reanalyze the ages of two samples. It is found that the modified version of De determining procedure for MET-pIRIR method produce satisfactory age results for the two water-lain samples. For example, the age of BSZ-A is measured at 12±22 a with results of 30 aliquots and is measured at -10±10 a with results of 79 aliquots. The age of BSZ-B is measured at 7±70 a with results of 40 aliquots and is measured at 27±27 a with results of 73 aliquots. The modified De determining procedure of MET-pIRIR method provides a new way for dating insufficient bleaching K-feldspar samples, especially for the samples without age-temperature plateau using average age model.