Pay the PIED Piper: Guidelines to Visualize Large Geochemical Datasets on Piper Diagrams. Issue 3 (11th November 2019)

• Record Type: Journal Article
• Authors: Russoniello, Christopher J.; Lautz, Laura K.
• Is Part Of: Ground water. Volume 58:Issue 3(2020)
• Page Start: 464
• Page End: 469
• Publication Date: 2019-11-11
• Subjects: Groundwater -- Periodicals; Wells -- Periodicals; Eau souterraine -- Périodiques; Puits -- Périodiques; Grondwater; Eau souterraine; Puits; Electronic journals; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 551.49
• DOI: 10.1111/gwat.12953
• Languages: English
• ISSNs: 0017-467X

Abstract

Abstract: The Piper diagram has increased in popularity since its 1944 introduction and is now one of the most familiar and effective tools in the hydrogeologist's toolbox. Within the Piper diagram, three points on three related plots fully display the major ionic species of a water sample. Recently the size and availability of datasets have increased as additional field measurements and modeling results are shared more effectively in online databases. This growth presents opportunities and challenges for data analysis and conveyance—larger and longer datasets increase the potential to identify trends and patterns, but traditional Piper diagrams are quickly overwhelmed by large datasets as dense points overlap and become obscured. We present guidelines for effectively displaying large geochemical datasets on traditional Piper diagrams and new code that adds novel functionality for following these generic guidelines. This code, plotting interesting environmental data with Piper diagrams (PIED Piper), can be run within the Matlab environment or through a stand‐alone graphical user interface, and is the first Matlab code to generate Piper diagrams. The illustrative examples herein demonstrate (1) how limitations in displays of large datasets may be overcome with translucent symbology, contours, and heatmaps to identify trends and patterns, (2) how clusters of similar points can be identified and differentiated with convex hulls, and (3) how temporal‐and‐spatial patterns may be visually diagnosed with image groups and movies. The guidelines discussed in these examples will aid PIED Piper users to achieve the two goals of effective big data visualization: analysis and communication. Abstract : Article impact statement : Matlab code generates publication‐quality Piper diagrams of large geochemistry datasets with heatmaps, contours, and automated group lassoes.