Abstract Principal Component Analysis and Multivariate Curve Resolution (MCR) chemometrics soft modelling approaches are proposed for the simultaneous analysis of the half hourly tropospheric ozone concentrations over a continuous period of six years at three different monitoring stations. The proposed approach is based on the direct multilinear modelling of experimental data without needing any prior physicochemical, geographical nor meteorological assumptions. By application of the proposed modelling approach, a summarized description of the changes of ozone concentrations due to a reduced number of contributions is achieved. Each one of these contributions is described by a set of diurnal, seasonal, year and location profiles whose interpretation is then possible using the available knowledge. The use of the proposed chemometrics approach can be generalized for the analysis and interpretation of environmental big data sets generated in continuous on-line monitoring of multiple environmental parameters over long time periods and for a large number of monitoring stations.

中文翻译：

---

不同监测站臭氧半小时浓度时间变化的化学计量学建模

摘要 提出了主成分分析和多元曲线分辨率 (MCR) 化学计量学软建模方法，用于同时分析三个不同监测站连续六年的半小时对流层臭氧浓度。所提出的方法基于实验数据的直接多线性建模，无需任何先前的物理化学、地理或气象假设。通过应用建议的建模方法，实现了对由于贡献数量减少而导致的臭氧浓度变化的概括描述。这些贡献中的每一个都由一组昼夜、季节性、年份和位置概况描述，然后可以使用可用知识对其进行解释。