Abstract

As Earth's ecological landscape continues to change, it will become increasingly important to understand how it has changed, and how it may change in the future. Freely available multispectral remote sensing datasets, such as the Landsat-8 dataset accessed through the USGS EarthExplorer tool, provide large scale, high resolution satellite imagery that can be leveraged by researchers across scientific disciplines for timeseries index analysis. LINDEX provides an extensible framework that automates Landsat-8 timeseries index analysis, resulting in an average ninety-four percent reduction in overall processing time compared to hands-on methods. Traditionally, in order to make use of this historical data, researchers must acquire the data, uncompress each scene, crop each scene to their region of interest, sort each cropped scene by date and cloud cover, and then either use GIS tools to run index analyses or code the index analyses themselves. This process is time consuming, requires extensive computational knowledge, and is prone to human error. In order to address these challenges, we developed LINDEX, a fully containerized end-to-end extensible processing framework for Landsat-8 timeseries index analysis. LINDEX is open source and leverages open source python packages to automate decompression, cropping, cloud cover detection, sorting by date, and index analysis for Landsat-8 data while also providing a customisable and growing library of eleven ecologically useful indices including NDWI, NDMI, NDSI and NDGI. LINDEX is designed to work synergistically with the EarthExplorer Bulk Download Application as well as QGIS, providing a bridge from download through analysis. The LINDEX framework uses a well defined methodology for incorporating custom indices into your workflow, making LINDEX a useful tool for researchers interested in exploring any Landsat-8 multispectral index while saving time, and reducing error.