A workflow for spatio-seasonal hydro-chemical analysis using multivariate statistical techniques

Multivariate statistical techniques are powerful in data interpretation and pattern recognition, which play a vital role in pollutant source identification for water environment management. Despite of their wide application in hydro-chemical analysis, absence of a comprehensive workflow hinders the practices and further studies. The present study constructed a workflow on the application of multivariate statistical techniques in spatio-seasonal hydro-chemical analysis, which provided a basic guidance for practices and a systematic support to future exploration. Selection of the methods and work paths for spatio-seasonal analysis largely depends on the structure of data set and the requirements of specific tasks. Trial and adjustment could be repeatedly performed to optimize the analysis strategy and identify more underlying patterns. Given a multiscale dataset concerning complex spatio-seasonal variations, temporal or spatial grouping using appropriate methods to reasonably divide the complicated data set contributes to data interpretation and pattern recognition. The upper Yangtze River basin (UYRB, China) was employed for case analysis to demonstrate how the workflow guides an efficient and effective data exploration. Efforts could be made in future works to continually improve the workflow to involve more complicated analysis and techniques and the integrated application in various fields.