Seasonal hydrological loading from GPS observed data across contiguous USA using integrated R and Hadoop-GIS framework

Abstract

Hydrological water mass redistribution on the Earth’s surface causes elastic loading deformation. In this study, vertical crustal deformation is assessed by an integrated spatio-temporal approach based on R statistical software and Hadoop-GIS framework at both regional and local scales. Specifically, the study examines the relationship between seasonal vertical loading deformation and seasonal hydrological loading from rain and snow. The vertical loading deformation is characterized by time-series estimated from continuous Global Positioning System (GPS) network across the contiguous USA for a timeframe of 48 months (January 1, 2013 to December 31, 2016) which coincides with California’s most extreme recent drought. The data processing used custom-built R programming scripts and spatial libraries that were integrated with Geographic Resources Analysis Support System (GRASS), Geographic Information System (GIS), and Hive framework which is a data warehouse extension of Apache Hadoop used as a database query interface. The relationships of vertical displacement are explored by visualization techniques such as spatial maps and wavelet coherence plots. Additional analysis of the relationships can be performed through interactive web-based interface.