High performance analysis of big spatial data

Abstract

Every year research institutions produce petabytes of data. Yet, only a small percent of the data is readily accessible for analysis. Terra Populus acts as the bridge between big data sources and researchers. Researchers are provided convenient web applications that allow them to access, analyze, and tabulate different datasets under a common platform. Terra Populus is developing three unique applications. The first application, Paragon, is a prototype parallel spatial database, which aims to extend the functionality of PostgreSQL and PostGIS onto multinode systems. Terra Populus' Tabulator application employs Parquet on Spark to build dynamic queries for analyzing large population survey data. The last application, Terra Explorer, is an exploratory analysis tool for visualizing the spatial datasets within the repository.