LAMOST光谱质量控制和检查系统的软件实现和关键技术研究

项目名称： LAMOST光谱质量控制和检查系统的软件实现和关键技术研究

项目编号： No.11203045

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 天文学、地球科学

项目作者： 王凤飞

作者单位： 中国科学院国家天文台

项目金额： 24万元

中文摘要： LAMOST进入先导巡天阶段，目前每天的数据量很大，每天产出1-2万条光谱。光谱的质量与科学产出密切相关，为了更好的完成数据发布工作，需要对光谱质量进行分析并按照相应标准给出参照。通过对光谱进行质量分级和各个指标的界定，包括信噪比，流量定标，坏点以及低流量，低信噪比等小项，最终给出系统的质量分析，并形成每个天区的质量报告。具体方法：在巡天前期，通过人工检查光谱质量是有必要的，可以帮助了解光谱现实状况和处理程序的不足，并给出人工质量分级和人工光谱类型分类等结果，进而将人工结果提交到LAMOST光谱分析数据库。对人工结果的检查和整理之后可以为改进处理程序提供方向，并且标定需要重复处理的数据，进而提高光谱处理质量。一维光谱在处理过程中会对光谱的质量进行判断和反馈，与人工检查相结合，不断将这一过程自动化和标准化，进而代替人工检查,从而实现LAMOST数据的高质量自动化处理和发布系统。

中文关键词： LAMOST；人机交互；光谱；；

英文摘要： LAMOST has been running for its pilot survey for about 6 months. The telescope obtains a great quantity of spectra every observering night which could be up to 20,000. The quality of spectra is very important to science study. For release better data of LAMOST, we need anlaysis the quality of spectra and give out a regulation about spectra quality and analysis results. The factors which can influence final results includes signal noise ratio, flux calibration, low quality spectra. We will consider all factor here and analysis the final results of this observation and provide quality analysis reports. In the early phase of survey, human check is important and inevitable. True situation of spectra can collected from members and form useful advices to pipeline. Eyes results including classification and redshift are submitted to our datebase. Cross analysis on database would help us conclude what efforts we need to focus and guide the pipeline rewriting. Also, eyes results are useful for calibration of LAMOST spectra on many fields. As the closure of this study, we can implement a high automatic software system which could reduce human participation.

英文关键词： LAMOST；human-computer interaction；spectra；；