Simulated data from Kelly et al. 2011, A signal-to-noise index to quantify the potential for peak detection in sediment-charcoal records, Quaternary Reserach, 75: 11-17.

Simulated charcoal time series used in: Kelly, R. F., Higuera, P. E., Barrett, C. M. & Hu, F. S. (2011) A signal-to-noise index to quantify the potential for peak detection in sediment-charcoal records. Quaternary Research, 75, 11-17. Each .csv file contains "summary data" that describes the parameters used in the CharAnalysis program for threshold determination. Row 16 includes column headings for CharAnalysis input, as follows: age -- [yr BP] age of simulated record cm -- [cm] depth of simulated record count -- [#] charcoal count (not discrete) vol -- [cm3] volume of sample acc -- [# cm^-2 yr^-1] Charcoal accumulation rate bkg -- [# cm^-2 yr^-1] Background charcoal rsd -- [# cm^-2 yr^-1] Residual charcoal thr -- [# cm^-2 yr^-1] Threshold separating peaks from non-peak values pkBool -- [binary] Peaks (1) or non-peak values (0) pkMCP -- [binary] Equals 1 if a "peak" value was screened out by the minimum count test. That is, the sample surpassed the threshold, but the count was low enough to not pass the minimum count test. (These values should almost always be 0) sni -- [index] Signal-to-noise index sniSm -- [index] Smoothd SNI Detials from Kelly et al. (2011): "Simulated records were generated using the CharSim model (Higuera et al., 2007), which simulates a spatially and temporally explicit fire regime, charcoal production and dispersal processes, primary (airborne) and secondary (slope wash and withinlake redeposition) deposition, sediment mixing, and sediment sampling. The CS1 scenario is based on model parameters representing boreal-forest charcoal records, with high temporal resolution (15-yr contiguous samples) and little vertical mixing in sediments. The CS2 scenario differs from CS1 only in that additional sediment mixing is simulated (vertical mixing depth doubled from 1.0 to 2.0 cm), as might be observed in shallower lakes. In both CS1 and CS2 scenarios, fire sizes mimic those observed in Alaska from 1988 to 2003 (Alaska Fire Service, 2004). By contrast, the CS3 scenario is based on fires of constant size, equal to the mean fire size of the same dataset. This scenario creates a charcoal series with a relatively uniform distribution of CHAR values, and illustrates a record in which peaks from local fires are difficult to detect due to processes independent of charcoal taphonomy."

本数据集为用于以下研究的模拟木炭时间序列：Kelly, R. F., Higuera, P. E., Barrett, C. M. 与 Hu, F. S. (2011) 《用于量化沉积物木炭记录中峰值检测潜力的信噪比指数》（A signal-to-noise index to quantify the potential for peak detection in sediment-charcoal records），《第四纪研究》(Quaternary Research)，75卷，第11-17页。

每个.csv文件均包含用于阈值确定的CharAnalysis程序的"汇总数据"，该数据描述了其阈值确定所用的全部参数。第16行包含CharAnalysis程序输入的列标题，具体如下：
age -- [yr BP] 模拟记录的年代（单位：距今年份yr BP）
cm -- [cm] 模拟记录的深度（单位：厘米）
count -- [#] 木炭计数（非离散计数）
vol -- [cm³] 样品体积（单位：立方厘米）
acc -- [#·cm⁻²·yr⁻¹] 木炭积累率
bkg -- [#·cm⁻²·yr⁻¹] 背景木炭含量
rsd -- [#·cm⁻²·yr⁻¹] 残余木炭含量
thr -- [#·cm⁻²·yr⁻¹] 区分峰值与非峰值的阈值
pkBool -- [binary] 峰值标记（1代表峰值，0代表非峰值）
pkMCP -- [binary] 若某"峰值"因最小计数测试被筛除则标记为1。即该样品值超过阈值，但计数过低未通过最小计数测试（此类数值通常应为0）
sni -- [index] 信噪比指数
sniSm -- [index] 平滑化信噪比指数

以下为Kelly等人(2011)的详细细节："模拟记录通过CharSim模型（Higuera等，2007）生成，该模型可模拟时空显式的火况、木炭生产与扩散过程、一次（大气沉降）和二次（坡面冲刷与湖内再沉积）沉积作用、沉积物混合以及沉积物采样流程。CS1情景基于代表北方森林木炭记录的模型参数，具备高时间分辨率（15年连续采样）且沉积物垂直混合程度极低。CS2情景仅在沉积物混合设置上与CS1存在差异：模拟的垂直混合深度从1.0 cm翻倍至2.0 cm，该情况常见于较浅湖泊。CS1与CS2情景中的火灾规模均参照1988至2003年阿拉斯加观测到的火灾规模（阿拉斯加消防局，2004）。与之相反，CS3情景基于固定规模的火灾，其规模等于上述数据集的平均火灾规模。该情景生成的木炭序列具备相对均匀的CHAR值分布，展示了一类因非木炭埋藏学过程导致的本地火灾峰值难以被检测的记录。"