图1 说明释光过程的能带模型示意图（修改自文献[1,2]） 1) 晶体受到放射性辐照使原子发生电离产生电子-空穴对,将电子从价带推入导带并在价带留下空穴,电子和空穴分别被T和L晶体缺陷（陷阱）捕获Ionizing radiation (α, β, γ) produces electron-hole pairs, pushing electrons into the conduction band, leaving holes in the valence band, and resulting in the trapping of electrons and holes at T and L defects (traps), respectively;2) 在电子陷阱中的电子(T)和在复合释光中心(L)的空穴的储存寿命从数秒到数百万年不等,它取决于陷阱在导带下面的能量深度(E),陷阱越深,电子越稳定,停留的时间越长Electrons in electron traps (T) and holes in the recombination center have lifetimes ranging from seconds to millions of years. The lifetime is dependent on the energy depth (E) of the traps below the conduction band. The more stable the electron and the longer it stays trapped;3) 当样品受到加热或合适波长的光照时,电子从电子陷阱中被驱逐出来,其中一些通过导带到达释光中心(L)与捕获的空穴重新复合并发出释光 By heating and shining light, electrons are released from the electron traps, some of them reach the luminescence center (L) through the conduction band, recombine with holes at luminescence centers and emit light (TL or OSL)

Fig.1 Simple energy-band-model for luminescence processes (after reference [1,2])

图2 沉积物释光测年基本原理示意图 图中PMT表示光电倍增管;(a)沉积物颗粒在搬运过程中原来的释光信号因光照晒退回零,在埋藏后接受周围中的放射性（α、β和γ射线）辐照(b)产生新的释光信号直到取样测量。

Fig.2 The basic principle of luminescence dating of sediments PMT refers to photomultiplier; (a) Sediment grains are exposed to sunlight during transportation, their luminescence signals are bleached and zeroed. The grains are irradiated by α-particles, beta and gamma rays during the burial period (b), and the signals have been accumulated until sampling for OSL measurement.

图3 用单片再生法建立的石英光释光信号生长曲线（剂量-反应曲线） a)当辐照(再生)剂量(D)达到600 Gy 左右时,光释光信号(I)达到完全饱和,样品的特征饱和剂量(D0)为112 Gy,将样品的自然光释光信号投影到曲线上可求出样品的等效剂量（De）,该样品（HS11-1）来自西班牙Huéscar-1动物化石遗址中的河流沉积物[31,32]。b)该生长曲线为双饱和指数方程拟合,样品在1500 Gy剂量后还在随再生剂量增加而增加,两个函数的特征饱和指数分别是71 Gy和828 Gy,该样品为来自贵州盘县大洞旧石器洞穴遗址的堆积物[33]。c)生长曲线斜率对计算等效剂量误差的影响（修改自文献[26]）,当光释光信号是饱和信号强度的20%、86%、96% 和100%（y轴）时对对应的等效剂量分布（x轴）,图中假设光释光信号的相对准偏差为5%,且是正态分布,当光释光信号为饱和信号强度95%和100%时,它们分别有15%和50%的信号（图中灰色部分）不能投影到曲线上,从而引起等效剂量的低估[34,35]。

Fig.3 Growth (dose-response) curves for quartz obtained using the single-aliquot regenerative-dose protocol (a) The curve is fitted using a single saturation exponential function. The OSL signal (I) is saturated when the regenerative dose (D) reaches 600 Gy, and the characteristic saturation dose (D0) of the curve is 112 Gy. The natural signal is projected onto the fitted growth curve to estimate the De value by interpolation. This sample (HS11-1) is fluvial sediment from the Huéscar-1 site in Spain[31,32]. (b) Growth curve was fitted using double saturating exponential function. The OSL signal increases with increasing dose when the dose was larger than 1500 Gy, the two characteristic saturation doses are 71 Gy and 828 Gy, respectively. This sample from the Panxian Dadong cave in Guizhou province[33]. (c) The effect of the slope of a growth curve on De error (see details in reference [26]). When the natural luminescence signals are close to the maximum level of the curve, the corresponding De obtained may be underestimated [34,35]

图4 正态或高斯分布和标准差 图中的年代是一个例子; µ 和 σ 分别表示平均值和标准差

Fig.4 The normal or Gaussian distribution and standard deviation the age in the figure is an example; µ and σ refer to mean and standard deviation, respectively

图5 用来说明精密度和准确度区别的靶标 µ 和 σ 分别表示平均值和标准差

Fig.5 Targets used to illustrate the difference between precision and accuracy µ and σ refer to mean and standard error, respectively

图 6 钾长石和石英的光释光年龄与其对应的独立年龄（其他测年方法得到的年龄）的比较 图中的斜线为1:1线或比值,每个数据点的误差棒为 1σ误差。(a)来自世界各地116个样品的钾长石两步法红外后红外释光年龄和(b)来自欧亚45个样品的钾长石多步高温红外后红外（MET-pIRIR 或pMET-pIRIR）释光年龄,其中Lx、Tx和 Lx/Tx,分别表示再生（或自然）剂量和试验剂量产生的释光信号和灵敏度校正后的释光信号（修改自文献[4]）;（c）来自世界各地152个释光信号晒退回零较好样品的石英光释光年龄,图中n表示样品数,以及在1:1线±2σ误差范围内的样品百分比(修改自文献[26])。

Fig.6 Comparison of potassium feldspar and quartz OSL ages with their corresponding independent ages obtained by other dating methods The lines in the figures are 1:1 lines or ratios, and the error bars for each data point refer to 1σ error. (a) The two-step pIRIR ages of potassium feldspar for 116 samples around the world, and (b) Multi-elevated-temperature (METor pMET) pIRIR ages of potassium feldspar from 45 samples from Europe and Asia, where Lx, Tx and Lx/Tx respectively represent regeneration-dose (or natural) OSL, test-dose OSL and sensitivity-corrected OSL signals (modified from reference [4]); (c) The quartz OSL ages of the 152 samples from all over the world. The samples are fluvial, eolian, ocean, and glacial sediments, which were well bleached prior to deposition. n = the number of samples, followed by the percentage of samples within ±2σ error of the 1:1 line (modified from reference [26])

图7 已经公认的部分古人类分类群时间范围和释光测年范围 释光测年范围：双箭头表示获得可靠释光年龄的大致区间,虚线是理想环境下的范围（修改自文献[25]）

Fig.7 The time range of hominin taxa currently recognized and the age range and luminescence dating Luminescence dating range: The double arrows indicate the approximate interval where reliable luminescence ages can be obtained (modified from reference [25])

图8 石英和钾长石光释光信号生长曲线和其特征饱和剂量(D0)分布 (a)同一样品中不同矿物和释光信号的生长曲线：石英TT-OSL信号、钾长石红外后红外释光信号（pIRIR）,石英超级颗粒的OSL和一般石英的OSL信号,该样品（HS11-1）来自西班牙Huéscar-1遗址中的河流堆积物[31,32] Growth curves for different luminescence signals from quartz and potassium feldspar, the sample （HS11-1）is from the Huéscar-1 site in Spain[31,32]。(b)来自贵州观音洞遗址堆积物中样品GYD-OSL8的粗粒石英生长曲线,根据曲线的形状可将该样品中的石英颗粒分为4类,其中有近40%的颗粒的自然光释光信号已经饱和[19]Growth curve for a fine quartz aliquot from sample GYD-OSL8 from the Guanyindong cave in Guizhou province[19], based on the shapes of the curves, four groups are divided。(c) 西伯利亚3个石英样品和 (d)西班牙Cuesta de la Bajada旧石器遗址3个石英样品的D0值分布的直方图[31,32] (c) and (d) D0 distribution of the quartz grains from three samples from Siberia and three quartz samples from the Cuesta de la Bajada site, respectively[31,32]

Fig.8 Dose response curves for different luminescence signals from quartz and potassium feldspar and distribution of characteristic dose (D0)