Independent component analysis of space-time patterns of groundwater system

Chin Tsai Hsiao, Jui Pin Tsai, Yu Wen Chen

Research output: Chapter in Book/Report/Conference proceedingChapter


This study proposed a method based on Independent Component Analysis (ICA) to understand the mechanisms that cause regional groundwater head variations. To verify the capability of the proposed method, this method is applied to an ideal numerical groundwater model, which was developed by using MODFLOW. The unconfined aquifer parameters are set as homogeneous and isotropic. The values of the two groups of pumpages (sinks) and one rainfall recharge (sources) were time-variant, and the frequencies among the three sink/sources were different. The simulated heads were sampled from 64 selected observation wells within the model boundary with a daily time step for 5 years. The simulated heads of the 64 wells were inputted to ICA. The study results show that the ICA can successfully decompose the sampled heads into three independent components (ICs) resulted from the three sink/source. To identifying the physical meanings of the three ICs, the correlation coefficients between ICs and the three sinks/sources were computed, and their values are 0.9816, 0.888 and 0.684, respectively. The separating matrix of ICA was also used to identify the pumping well locations. The study results show that the proposed method provides a novel and efficient method to understand the spatiotemporal head variations of ground-water system and can be used to locate the pumping wells, which is crucial for the regional groundwater management.

Original languageEnglish
Title of host publicationLecture Notes in Electrical Engineering
PublisherSpringer Verlag
Number of pages12
Publication statusPublished - 2016 Jan 1

Publication series

NameLecture Notes in Electrical Engineering
ISSN (Print)1876-1100
ISSN (Electronic)1876-1119

All Science Journal Classification (ASJC) codes

  • Industrial and Manufacturing Engineering


Dive into the research topics of 'Independent component analysis of space-time patterns of groundwater system'. Together they form a unique fingerprint.

Cite this